Volume 8, Issue 2 - Proceedings of the 2022 Joint Annual Conference of the Austrian (ÖGBMT), German (VDE DGBMT) and Swiss (SSBE) Societies for Biomedical Engineering; Guest Editors: Daniel Baumgarten and Johannes Vorwerk

In this paper, we present a control system theory based modeling approach for the voltage-gated ion-channel Kv10.1 at different temperatures. Kv10.1 is expressed in the central nervous system and the brain of healthy humans. The study of the activity of this ion channel is important, because its activity is associated with the occurrence of cancer in different organs or tissues. According to systems and control theory, the voltage-gated channel Kv10.1 was assumed to be a linear time-invariant system and, as such, exhibits dynamic behavior. The experimental results show that Kv10.1 operates as a first-order model based on the input voltage step protocol and the measured macroscopic ion current output.

The incidence of diabetes is increasing globally. Short and long-term effects of diabetes can negatively affect the lives of a person and their relatives. A good glycaemic control is important. Compared to conventional blood glucose measurement with capillary blood, which is obtained by finger pricking, it is now possible to continuously measure glucose using sensors that are placed in the subcutaneous tissue and remain there for several days to weeks. This means that the "tissue sugar" and not the blood sugar directly is determined. For this reason, it is important to determine the accuracy of these sensors. Furthermore, such systems can influence not only the physical but also the psychological/mental health of people with diabetes and their relatives. Reimbursement of sensors differs in Europe. This overview is based on the Austrian health care system.

Ewing sarcomas are malignant neoplasm entities typically found in children and adolescents. Early detection is crucial for therapy and prognosis. Due to the low incidence the general experience as well as according data is limited. Novel support tools for diagnosis, such as deep learning models for image interpretation, are required. While acquiring sufficient data is a common obstacle in medicine, several techniques to tackle small data sets have emerged. The general necessity of large data sets in addition to a rare disease lead to the question whether transfer learning can solve the issue of limited data and subsequently support tasks such as distinguishing Ewing sarcoma from its main differential diagnosis (acute osteomyelitis) in paediatric radiographs. 42,608 unstructured radiographs from our musculoskeletal tumour centre were retrieved from the PACS. The images were clustered with a DeepCluster, a self-supervised algorithm. 1000 clusters were used for the upstream task (pretraining). Following, the pretrained classification network was applied for the downstream task of differentiating Ewing sarcoma and acute osteomyelitis. An untrained network achieved an accuracy of 81.5%/54.2%, while an ImageNet-pretrained network resulted in 89.6%/70.8% for validation and testing, respectively. Our transfer learning approach surpassed the best result by 4.4%/17.3% percentage points. Transfer learning demonstrated to be a powerful technique to support image interpretation tasks. Even for small data sets, the impact can be significant. However, transfer learning is not a final solution to small data sets. To achieve clinically relevant results, a structured and systematic data acquisition is of paramount importance.

Superparamagnetic iron oxide nanoparticles are a promising material in nanomedicine, especially for generating magnetically controlled drug delivery systems. We developed an innovative ureidopyrimidinone (UPy) coating based on supramolecular hydrogen bonding units. The synthesized nanoparticles possess a positively charged surface with a hydrodynamic diameter of 177 nm at pH 7 and magnetization of 31 emu g -1 . The system has the potential to be modified by drugs or bioactive molecules altered with UPy units.

Craniosynostosis is a congenital disease characterized by the premature closure of one or multiple sutures of the infant’s skull. For diagnosis, 3D photogrammetric scans are a radiation-free alternative to computed tomography. However, data is only sparsely available and the role of data augmentation for the classification of craniosynostosis has not yet been analyzed. In this work, we use a 2D distance map representation of the infants’ heads with a convolutional-neural-network-based classifier and employ a generative adversarial network (GAN) for data augmentation. We simulate two data scarcity scenarios with 15% and 10% training data and test the influence of different degrees of added synthetic data and balancing underrepresented classes. We used total accuracy and F1-score as a metric to evaluate the final classifiers. For 15% training data, the GAN-augmented dataset showed an increased F1-score up to 0.1 and classification accuracy up to 3 %. For 10% training data, both metrics decreased. We present a deep convolutional GAN capable of creating synthetic data for the classification of craniosynostosis. Using a moderate amount of synthetic data using a GAN showed slightly better performance, but had little effect overall. The simulated scarcity scenario of 10% training data may have limited the model’s ability to learn the underlying data distribution.

Cold plasma disinfection is a cost-efficient and, above all, fast way of disinfecting even complex products. There are already approaches where plasma disinfection is used for wound treatment or hand disinfection. The ionization of a gas results in a number of physical and chemical processes that have a damaging effect on microorganisms. Especially in the field of medical device reprocessing, a tool that can reliably disinfect even very complex products in a short time would be a great asset. In this work, the potential of a newly developed cold plasma disinfection device was tested for the reduction of microbiological contamination and thus the disinfecting effect. In order to examine this microbiological reduction 3D printed scaffolds with contaminated test plates were used. This was done with different concentrations of the bacteria in the cold plasma disinfection process to determine the maximum germ reduction effect. In a first test run, the maximum effect of germ reduction was achieved with log 3.6. By making further changes and increasing the disinfection cycles, it was possible to increase the germ reduction to log 4.7. If these values are confirmed and can be improved by further modifications (e.g. increasing the plasma concentration), cold plasma technology represents a very good alternative to conventional disinfection methods.

Due to the corona pandemic, the demand for efficient and fast-acting disinfection measures is high. In this work, the disinfecting potential of a dry fogging technology (Apollon Biotech GmbH, Germany) was investigated for the possibility of disinfection of vehicles for passenger transport. This could represent a new approach for processing emergency vehicles or local passenger transport vehicles quickly and cost-effectively. A public bus was used as a model for a passenger transport vehicle (Ferienfahrschule Hense GbR, Germany). This vehicle was tested at predetermined points before and after dry-fog disinfection with swab and contact plate samples. In addition, contaminated sample tiles (Staphylococcus aureus ATCC 6538; 5x104 CFU) were placed at some locations in order to be able to record a standardized germ reduction. To determine the decontamination effect of the air in the vehicle, active airborne germ measurements were performed. In the airborne germ measurement, our results indicated a reduction of approx. 68% of the microbiological air contamination. Screening of the total contamination of the bus at several selected sites showed an average load of approximately 348 CFU per test site. After disinfection, the average bacterial count was reduced to approx. 18 CFU per test site. This disinfecting effect could also be shown with the standardized test tiles. An efficiency level of between log 0.8 and 3.7 was obtained at the various test sites.

passive mixers and droplet generation microfluidic chip modules were designed and manufactured using a commercial SLA 3D-printer. The mixing modules were designed specifically for 3D-printing and evaluated using FEM modeling. The co-flow droplet generator was used for cancer cells encapsulation and drug potency evaluation.

Pickering stabilisation is a manufacturing process involving the adsorption of colloidal particles at gas-liquid interfaces. It is used to create the shells of stable, long-lived ultrasound contrast agent microbubbles. The purpose of the present study is to determine whether high-amplitude sonication influences the integrity of Pickering-stabilised shells. To this purpose, Pickering-stabilised microbubbles were subjected to high-speed photography at 10 million frames per second during 1-MHz, 1-MPa sonication. In addition, radial excursions as a function of time were simulated using the Rayleigh-Plesset equation for free gas microbubbles and microbubbles encapsulated by Pickering-stabilised shells of 7.6-Nm −1 stiffness. The maximum expansions observed from camera recordings were either agreeing with those computed for Pickering-stabilised microbubbles or corresponding to greater values. The results indicate that optically identical microbubbles may undergo shell disruption of different severity. We conclude that the disruption occurs during sonication and not prior to it. These findings may aid in the development of Pickering-stabilised agents that facilitate ultrasoundtriggered release.

Gellan gum is a hydrogel with several applications in ultrasonic imaging, novel drug delivery, and tissue regeneration. As hydrogels are dynamic entities, their viscocelastic and therefore their acoustic properties change over time, which is of interest to monitor. To determine the speed of sound from brightness-mode images, however, rather large quantities of hydrogel are needed. In this study, we investigated torsion rheometry as a means to determine acoustic properties. Perceived speeds of sound were derived and computed from torsion rheometry measurements of gelating gellan gum mixed with spermidine trihydrochloride crosslinker. For comparison, brightness-mode ultrasonic images were recorded of the same material inside a phantom well. The rheometry data converged to a speed of sound within a standard devitation of the speed of sound measured from the brightness-mode images.We have shown that dynamic acoustic properties of gelating gellan gum can be simulated and experimentally determined using torsion rheometry.

In current endoscopes, the estimation of the proximity to the surrounding tissue is based on the experience of the user. To provide precise proximity meassurement, a suitable sensor is needed. This paper presents a fiber optic proximity sensor that is inserted through the working channel of an endoscope. The impact of the fiber cutting angle at the distal end of the optical fiber is investigated. If the fiber is cut at a 3° angle to the optical axis, the measurement signal in axial direction can be increased by 126mV at the peak (measurement accuray 4.9 mV). A 45° fiber cutting angle allows radial measurement but with a weak measurement signal compared to an axial measurement. A distance between endoscope tip and intestine wall up to 15 mm can be measured. In conclusion, optical fibers can be used to insert a low-cost sensor through a working channel of an endoscope. By modifying the fiber cutting angle, an increased sensor signal can be obtained in the axial direction and a measurement in radial direction is possible.

Controlled disintegration of pharmaceutical tablets has been of interest for consistency checks and drug delivery. Under sonication, tablet disintegration is accelerated. This acceleration has been attributed to the existence of microscopic air pockets inside compacted tablets. In this study, we investigated the existence of such pockets by subjecting tablets to micro-computed tomography. In addition, we subjected tablets from the same batch to sonography. The tablets were measured to have an ultrasonic swelling rate of 162±16 μms−1. The micro-computed tomography images showed air pockets of up to 9 μm in diameter, some of which were visibly connected to each other. The brightness-mode images showed scattering from inside the tablets, suggesting cavitation activity. We conclude that, inside the tablets investigated, microscopic air entrapments exist, whose pulsations are detectable with brightness-mode ultrasound.

Malignant melanoma is a very aggressive tumour with the ability to metastasize at an early stage. Therefore, early detection is of great importance. Multiphoton tomography is a new non-invasive examination method in the clinical diagnosis of skin alterations that can be used for such early diagnosis. In this paper, a method for automated evaluation of multiphoton images of the skin is presented. The following features at the cellular and subcellular level were extracted to differentiate between malignant melanomas, lesions, and healthy skin: cell symmetry, cell distance, cell density, cell and nucleus contrast, nucleus cell ratio, and homogeneity of cytoplasm. The extracted features formed the basis for the subsequent classification. Two feature sets were used. The first feature set included all the above-mentioned features, while the second feature set included the significantly different features between the three classes resulting from a multivariate analysis of variance. The classification was performed by a Support Vector Machine, the k-Nearest Neighbour algorithm, and Ensemble Learning. The best classification results were obtained with the Support Vector Machine using the first feature set with an accuracy of 52 % and 79.6 % for malignant melanoma and healthy skin, respectively. Despite the small number of subjects investigated our results indicate that the proposed automatic method can differentiate malignant melanoma, lesions, and healthy skin. For future clinical application, an extended study with more multiphoton images is needed.

Computed tomography (CT) is widely used as the imaging modality for the treatment of tumors in Microwave Ablation (MWA) therapy. In order to accurately perform ablation of liver tumors and prevent tumor recurrence it is necessary to segment both the living tumor and the ablated tissue on the CT images. The U-Net model has outperformed other methods in biomedical image segmentation. However, because of the low contrast between tumor and liver tissue texture, the traditional U-net network cannot perform an accurate segmentation of the CT images of liver during MWA therapy. The aim of this study is to improve the U-net model network to achieve a higher segmentation performance on the CT images of liver tumor inMWA therapy. To achieve this, residual block is added in the first steps of up-sampling to deepen the network depth and enhance the segmentation result. We compare the proposed method named as ‘ResLU-Net’ with a conventional U-Net model. The results show that the ResLU-Net method has a good performance in tumor segmentation with a structure similarity index (SSIM) value of 0.97. This new method can help physicians in the MWA therapy process.

To overcome limitations of currently used blood pressure measurement devices in accuracy, continuity and comfort, we propose an approach for blood pressure estimation from electrocardiogram (ECG) signals only. Thereby, statistical signal features are extracted from the ECG which, eventually, serve as input to a random forest regression. The method is trained and tested on MIMIC III waveform data with a large range of blood pressure values. It obtains a mean absolute error ± standard deviation of 3.73 ± 5.19 mmHg for diastolic blood pressure (DBP) and 5.92 ± 7.23 mmHg for systolic blood pressure (SBP), with Pearson coefficients Υ DBP =0.92 and Υ SBP =0.91 respectively.

In conventional laparoscopy, the field of view of the surgeon is limited. Angled laparoscopes enable the observation of surrounding structures, however, to change the viewing perspective on a specific object, permanent repositioning and rotation of the shaft is required. In this paper, a demonstrator of a steerable flexible laparoscope is presented, enabling the user to intuitively adjust the perspective onto an object by means of a single control element. The laparoscope provides a viewing angle of ± 50° at 50 mm working distance. In first tests, the presented laparoscope showed advantages regarding intuitiveness of the control, easier handling, and improved depth perception.

Motivation: Within the field of bio-implants one topic deals with the in-vitro production of 3D tissues from cell cultures. A major challenge in the cultivation of biopreparations is the imitation of the vascular system with its surrounding tissue. Additiv manufacturing of such vascularisation could be supported with an adequate template from high-resolution 3D radiographs. Goals: Evaluation of highresolution 3D-imaging methods and 3D-analysis approaches to extract the vascularisation from tissue samples. Methods: To provide such template, vascular structure of small tissues samples (partially treated with contrast medium) were recorded and compared using high-resolution CT and MRI imaging modalities. Optimal measurement parameters were selected for the acquisition of very small vessels. Interactive and semi-automated segmentation of the vessel system were investigated and compared. Results: MRI scans yield a higher contrast than CT scans, but are much slower. Duration of interactive segmentation ranges between 1 to 12 hours. Runtime of the (semi) automatic method was between 5 and 20 minutes, not counting manual adjustment of the parameters. Correlation between manual and automatic segmentation yield Hausdorff distances of 0.024 (CT) and 0.74 (MRI) and Dice coefficients of 0.7 (CT) and 0.39 (MRI). Conclusion: Both imaging methods are appropriate for high-resolution vessel detection and segmentation, nevertheless, MRI with no contrast agent seems preferable if the imaging time can be reduced.

Robotic assistance in rehabilitation procedures has gained increasing popularity for both scientists and patients. However, aside from the practicality and safety issues, the ability to motivate patients to pursue rehabilitation therapy is a hindrance and is still limiting the domestic usage of Ankle Rehabilitation Robotic Systems (ARRS). Most of the available ARRS are either applicable under clinical setups, or are dull to operate. New advancements have led to the fusion of virtual realities (VR) and augmented realities (AR) with ARRS. This paper presents a classification of contemporary ARRS based on their usability and practicality along with required modifications to make them befitting for domestic use. A systemic literature search is applied for these systems on Web of Science and PubMed databases and the search is confined to the last twelve years’ research from 2011 to 2022. Among the 219 found systems, 4 use VR, AR, or gamification. Generally, having the rehabilitation therapy equipped with novel interfaces, patients improve their motor function.

Bone tumours are a rare and often highly malignant entity. Early clinical diagnosis is the most important step, but the difficulty of detecting and assessing bone malignancies is in its radiological peculiarity and limited experience of non-experts. Since X-ray imaging is the first imaging method of bone tumour diagnostics, the purpose of this study is to develop an artificial intelligence (AI) model to detect and segment the tumorous tissue in a radiograph. We investigated which methods are necessary to cope with limited and heterogeneous data. We collected 531 anonymised radiographs from our musculoskeletal tumour centre. In order to adapt to the complexity of recognizing the malignant tissue and cope with limited data, transfer learning, data augmentation as well as several architectures, some of which were initially designed for medical images, were implemented. Furthermore, dataset size was varied by adding another bone tumour entity. We applied a data split of 72%, 18%, 10% for training, validation and testing, respectively. To provide statistical significance and robustness, we applied a cross-validation and image stratification with respect to tumour pixels present. We achieved an accuracy of 99.72% and an intersection over union of 87.43% for hold-out test data by applying several methods to tackle limited data. Transfer learning and additional data brought the greatest performance increase. In conclusion, our model was able to detect and segment tumorous tissue in radiographs with good performance, although it was trained on a very limited amount of data. Transfer Learning and data augmentation proved to significantly mitigate the issue of limited data samples. However, to accomplish clinical significance, more data has to be acquired in the future. Through minor adjustments, the model could be adapted to other musculoskeletal tumour entities and become a general support tool for orthopaedic surgeons and radiologists.

Antibubbles are small gas bubbles comprising one or multiple liquid or solid cores, typically surrounded by stabilising shells. Acoustically active microscopic antibubbles have been proposed for use as theranostic agents. For clinical applications such as ultrasound-guided drug delivery and flash-echo, it is relevant to know the fragmentation threshold of antibubbles and the influence of the stabilising shells thereon. For antibubbles with an infinitesimal frictionless elastic shell of constant surface tension, we simulated ultrasoundassisted fragmentation by computing radial pulsation as a function of time using an adapted Rayleigh-Plesset equation, and converting the solutions to time-variant kinetic energy of the shell and time-variant surface energy deficit. By repetition over a range of pressure amplitudes, fragmentation thresholds were found for antibubbles of varying size, core volume, shell stiffness, and driving frequency. As backscattering increases with scatterer size, and as drug delivery would require vehicles just small enough to pass through capillaries with a relatively large payload, we chose to present typical results for antibubbles of resting diameter 6 μ m with a 90% incompressible core. At a driving frequency of 13 MHz, the fragmentation threshold was found to correspond to a mechanical indices less than 0.4, irrespective of shell stiffness. This mechanical index is not considered unsafe in diagnosis. That means that antibubbles acting as drug-carrying vehicles could release their payload under diagnostic conditions.

QRS detection is a pre-processing step to detect the heartbeat in an electrocardiogram (ECG) for subsequent rhythm classification. However, measured ECG waveforms may differ as a result of intrinsic variability or due to artefacts or noise. If the signals are distorted, then this often leads to difficulties in QRS detection. Of course, a high QRS detection performance is an important part of an ECG analysis algorithm, and furthermore, it must work even for highly noisy signals. Singular value decompositon (SVD) is the factorization of a matrix into the product three matrices. SVD allows us to find important components of data and, thus, can be used for dimension reduction or denoising. We introduce SVD based methods for QRS detection and ECG denoising, especially for short unknown signal segments, and show application results.

Acquisition of medical data such as blood pressure, ECG, pulse, and other values is currently often carried out in doctor’s offices by physicians or medical staff. This requires valuable time of patients as well as of practices’ personel and can deliver specific data only of the time of the visit. In this paper we describe a different approach. Patients measure their biomedical vital data at home. Therefore a technical infrastructure together with a workflow were developed and applied within a current project with patients in a rural area. Vital data is transferred via smartphone apps to a cloud environment where doctors can easily access and assess the data.

Cancer cells typically exhibit altered ion channel expression and membrane potential, wich plays an important role in various processes such as proliferation or apoptosis Hence, therapeutic strategies that modulate ion channel activity or shift the membrane potential prove to be promising for cancer therapy. Since temperature significantly affects ion channel function, deviation from normal body temperture might provide akey factor in electrophysiological progression. In this pilot study, we focused on a first insight into the temperature-induced modulation of A549 cells using an automated patch clamp system. Measurements of 14 cells at hypo-, normo- and hypertermia were accomplished to investigate the temperature-dependent electophysiological behavior. The results revealed that both, temperatures below and above normal body temperature affect the current and potential of A549 cells. In hyperthermia the potential tends to more hyperpolarized values, while conversely, in hypothermic conditions we observer mainly a lower current reponse and thus a depolarization of the membrane potential.

If a long ultrasound driving pulse is chosen such, that its frequency is greater than the resonance frequency of one cell type and less than the resonance frequency of another cell type, the cells are separated in different nodes and antinodes of the standing sound field. The purpose of this study was to build an experimental setup for the ultrasonic fractionation of whole blood in flow conditions. Blood fractionation is the separation of whole blood into its separate components, commonly used for analysis and transfusions. The experimental setup comprised a container with singleelement ultrasound transducers customisable in elevation and azimuth, focussed on a capillary through which blood was flowing. The acoustic focus coincided with the optical focus of a microscopic system to which a high-speed camera was attached. Ultrasound transducers were used with centre frequencies between 1MHz and 20MHz and with pressure amplitudes less than 1MPa. The duty cycle was greater than 10% in all experiments. The field of view was 0.2 × 0.2mm2. The optical resolution and camera speed allowed for visibility of red and white blood cells during flow conditions. Preliminary results show that red and white blood cells could be forced to move in opposite directions. The acoustic frequencies and pressures used were representative for those used in clinical diagnosis. We have shown that ultrasonic whole blood fractionation is feasible using ultrasound. Sonic fractionation speed might be an indicator for diseases such as malaria.

Self-care and improving one’s well-being has been growing rapidly in recent years for manifold reasons (e.g. higher workload, corona pandemic). Consumer-grade noninvasive stimulation devices are therefore on the rise to counteract the occurrence of mood disorders and burn-out symptoms. Here, we aim at investigating the impact of dynamically varying auditory-visual stimulation patterns on neural entrainment patterns and resonance phenomena. Twenty-two healthy volunteers (11 female, 25.4 ± 5.1 years, one dropout, seven in control group) participated in the study. EEG data (64 channel; equidistant layout) were acquired preand during stimulation for each volunteer. Visual and auditory stimuli were presented via a headset (ATUM, Neuro- Bright; https://www.neurobright.co.uk/). Presentation patterns (frequency, intensity, spatial distribution) varied within a presentation session but were kept constant across all volunteers. Stimulus intensity was adjusted to individual comfort levels. Individual alpha peak frequencies (iAPF) were calculated via the power spectral density with 50% overlapping 10s epochs from pre-stimulation segments. For both, the study and the control group, a time-frequency representation was calculated for the pre- and during-stimulation segments. From this, power values were determined for different frequency-bands (iAPF, stimulation frequencies and second harmonics of the latter). Statistical analyses focused on contrasting the power values between pre- and during stimulation. Mean iAPF values were 10.25 ± 0.99Hz for the study and 10.63 ± 1.21Hz for the control group respectively. Both, power values at the stimulation frequencies and their second harmonics differed significantly between pre- and during stimulation (p stim =0.001; p harm =0.001) in the study group. No such difference was found for the control group (pstim=0.352; pharm=0.237). Further, neither the study nor the control group showed significant iAPF power differences (p study =0.035; p control =0.352; alpha*=0.008). Our results suggest that lightweight, portable auditory-visual presentation devices represent an effective tool for generating entrainment and resonance effects at home. Further analyses will focus on the investigation of individual differences driving such modulatory effects.

Interatrial conduction block refers to a disturbance in the propagation of electrical impulses in the conduction pathways between the right and the left atrium. It is a risk factor for atrial fibrillation, stroke, and premature death. Clinical diagnostic criteria comprise an increased P wave duration and biphasic P waves in lead II, III and aVF due to retrograde activation of the left atrium. Machine learning algorithms could improve the diagnosis but require a large-scale, well-controlled and balanced dataset. In silico electrocardiogram (ECG) signals, optimally obtained from a statistical shape model to cover anatomical variability, carry the potential to produce an extensive database meeting the requirements for successful machine learning application. We generated the first in silico dataset including interatrial conduction block of 9,800 simulated ECG signals based on a bi-atrial statistical shape model. Automated feature analysis was performed to evaluate P wave morphology, duration and P wave terminal force in lead V1. Increased P wave duration and P wave terminal force in lead V1 were found for models with interatrial conduction block compared to healthy models. A wide variability of P wave morphology was detected for models with interatrial conduction block. Contrary to previous assumptions, our results suggest that a biphasic P wave morphology seems to be neither necessary nor sufficient for the diagnosis of interatrial conduction block. The presented dataset is ready for a classification with machine learning algorithms and can be easily extended.

Bioelectrical signals are often pulse-shaped with superimposed interference signals. In this context, accurate identification of features such as pulse onsets, peaks, amplitudes, and duration is a frequent problem. In this paper, we present a versatile method of rather low computational complexity to robustly identify such features in real-world signals. For that, we take use of two straight-line models fit to the observations by minimizing a quadratic cost term, and then identify desired features by tweaked likelihood measures. To demonstrate the idea and facilitate access to the method, we provide examples from the field of cardiology.

INTRODUCTION: Deep brain stimulation (DBS) is widely used to alleviate symptoms of movement disorders. During intraoperative stimulation the influence of active or passive movements on the neuronal activity is often evaluated but the evaluation remains mostly subjective. The objective of this paper is to investigate the potential of a previously developed Weighted-frequency Fourier Linear combiner and Kalman filter-based recursive algorithm to identify tremor phases and types. METHODS: Ten accelerometer recordings from eight patients were acquired during DBS from which 186 phases were manually annotated into: rest, postural and kinetic phase without tremor, and rest, postural and kinetic phase with tremor. The method first estimates the instantaneous tremor frequency and then decomposes the motion signal into voluntary and tremorous parts. The tremorous part is used to quantify tremor and the voluntary part to differentiate rest, postural and kinetic phases. RESULTS: Instantaneous tremor frequency and amplitude are successfully tracked online. The overall accuracy for tremorous phases only is 89.1% and 76.3% when also non-tremorous phases are considered. Two main misclassification cases are identified and further discussed. CONCLUSION: The results demonstrate the potential of the developed algorithm as an online tremorous movement classifier. It would benefit from a more advanced tremor detector but nevertheless the obtained digital biomarkers offer an evidence-based analysis and could optimize the efficacy of DBS treatment.

A patient’s respiratory rate is one of the critical vital signs that is a determinant of patient well-being. However, it is all too often neglected or misreported by health care professionals. This study presents the design, build and testing of a low-cost, portable monitor to facilitate accurate reporting of respiratory rate. The monitor comprised a thermistor-based transducer to capture the breath cycle of patients based on the temperature differential created across the thermistor. The signal was conditioned and processed such that the signal could be analysed to identify the peaks and ultimately determine the respiratory rate. For a total cost at the time of development of less than C40, the integrated system demonstrated a modest average error of 5.6% across a range of different ambient temperatures, rate and depth of breathing, and orifice of breathing. This is comparable with other commercial and custom devices. The presented monitor may be of interest for use in an emergency room or clinical setting, especially in severely resource-constrained countries.

For treatment decision of Eustachian tube (ET) dysfunction, a reliable diagnosis of the reason for this dysfunction is of interest. Imaging methods like CT or MRI do not show enough details today. Endoscopy is challenging but can show at least superficial structures at the pharyngeal orifice. To look into the deeper tissue layers vascular imaging methods like IVUS and OCT were evaluated already. But these procedures are not made and certified for this application and come with risks for the patient. To overcome this issue, we propose the use of an additional closed end guide catheter that can be advanced into the ET. An OCT catheter can be placed inside without touching the patient’s tissue. In this paper we evaluate this approach on a sheep cadaver head and a human cadaver head.

Artificial Intelligence (AI) methods, which are often based on Machine Learning (ML) algorithms, are also applied in the healthcare domain to provide predictions to physicians and patients based on electronic health records (EHRs), such as history of laboratory values, applied procedures and diagnoses. The question about these predictions “Why Should I Trust You?” encapsulates the issue with ML black boxes. Therefore, explaining the reasons for these ML predictions to physicians and patients is crucial to allow them to decide whether the prediction is applicable or not. In this paper, we explained and evaluated two prediction explanation methods for healthcare professionals (physicians and nurses). We compared two model-agnostic explanation methods based on global feature importance and local feature importance. We evaluated the user trust and reliance (UTR) for the explanation results of each method in a user study based on real patients’ electronic health records (EHR) and the feedback of healthcare professionals. Based on the user study, we observed that both methods have strengths and weaknesses according to the patients’ data, especially based on the data size of the patient. When the amount of data is small, global feature importance is enough to use. However, when the patient’s data size is big, using a local feature importance method makes more sense. As future work, we will develop a hybrid explanation method (by combining these methods automatically with a smart setting) to obtain higher and more stable performance results in terms of user trust and reliance.

Optical fibres enable to design of new textile-based sensors for an unobtrusive, continuous, and long-term acquisition of photoplethysmography (PPG) signals. However, no research has been done about the accuracy of measuring blood oxygen saturation (SpO 2 ) and the feasibility of PPG-based blood pressure measurement. This proof-ofconcept study examined these aspects for a woven sensing patch applied on the forehead in five healthy participants. During a controlled desaturation study, SpO 2 estimation revealed an amplitude of the root-mean-squared error (ARMS) of 3.6%, and acceptable signal quality for blood pressure estimation was achieved for 40% of all data. These results indicate that textile-based sensors can reach the required PPG signal quality to simultaneously estimate multiple vital signs, including blood pressure.

DC potential shifts are the shifts observed in the EEG baseline which can last from seconds to minutes. The significance of these low-frequency components in healthy as well as pathological states of human physiology is getting more and more attention not only in scientific research but also in clinical applications. In this paper, we present our novel multimodal measurement setup for simultaneously investigating DC potential shifts in EEG (DC-EEG) and the changes in noninvasive transcutaneous pCO2 measurements. We present preliminary results of our measurements during hyperventilation and apnea, which are two commonly used activation methods for changes in pCO2.

Introduction: Osteosynthesis of the equine femur is still a challenge for veterinary medicine. Even though intramedullary fracture fixation is possible nowadays, the varying geometry of the medullary cavity along the bone axis is a critical factor. Limited contact area between implant and bone can cause insufficient primary stability. In this study, it was investigated whether the osteosynthesis stability can be improved with a form-adaptive reinforcement for the diaphyseal part of the proximal fragment. Material and Methods: Eight equine femora were fitted with intramedullary nail osteosynthesis and analyzed by 4-point bending. Virtual position planning of the ex-vivo implantation using CT-data increased comparability. For five femora the proximal fragment was reinforced with a flexible polymer mixture. Longterm stability was tested via cyclic loading. Bending stiffness and its development due to cyclic loading was evaluated before and after reinforcement procedure. Finally, load-to-failure was tested in the same setup. Results and Discussion: The application of the polymer reinforcement increased the maximum torque in the load-tofailure measurement by 26%. Bending stiffness was not affected in the measured loading range by the reinforcement. Cyclic loading increased bending stiffness for a conditioned state but showed to be reversible for the most part. Conclusion: The fracture adjacent reinforcement showed to be beneficial to the osteosynthesis stability, but further investigation is necessary for surgical application.

Atrial fibrillation is responsible for a significant and steadily rising burden. Simultaneously, the treatment options for atrial fibrillation are far from optimal. Personalized simulations of cardiac electrophysiology could assist clinicians in the risk stratification and therapy planning for atrial fibrillation. However, the use of personalized simulations in clinics is currently not possible due to either too high computational costs or non-sufficient accuracy. Eikonal simulations come with low computational costs but cannot replicate the influence of cardiac tissue geometry on the conduction velocity of the wave propagation. Consequently, they currently lack the required accuracy to be applied in clinics. Biophysically detailed simulations on the other hand are accurate but associated with too high computational costs. To tackle this issue, a regression model is created based on biophysically detailed bidomain simulation data. This regression formula calculates the conduction velocity dependent on the thickness and curvature of the heart wall. Afterwards the formula was implemented into the eikonal model with the goal to increase the accuracy of the eikonal model without losing its advantage of computational efficiency. The results of the modified eikonal simulations demonstrate that (i) the local activation times become significantly closer to those of the biophysically detailed bidomain simulations, (ii) the advantage of the eikonal model of a low sensitivity to the resolution of the mesh was reduced further, and (iii) the unrealistic occurrence of endo-epicardial dissociation in simulations was remedied. The results suggest that the accuracy of the eikonal model was significantly increased. At the same time, the additional computational costs caused by the implementation of the regression formula are neglectable. In conclusion, a successful step towards a more accurate and fast computational model of cardiac electrophysiology was achieved.

Motivated by a high demand, the research interest in personalized artificial tissues is steadily increasing. Combining knowledge of additive manufacturing and tissue engineering, the research field of 3D bioprinting emerged. This work presents a six-degree-of-freedom mechanically actuated extrusion bioprinter within a sterile working environment. The system is based on an off-the-shelf robot arm and a custom modular printhead end-effector. Advanced dexterity is achieved by the six degrees of freedom, enabling printing on non-planar surfaces. The printhead is designed for co-axial extrusion of three fluids but can easily be adapted for different number of fluids or different extrusion flows. The custom controller of the system is implemented within the Robot Operating System (ROS) framework and plans the trajectory based on a path given in a custom GCode dialect. Since the robot is clean-room-certified, can be sterilized using hydrogen peroxide steam, and is placed within a sterile hood, the setup enables working under sterile conditions.

Patients receiving parenteral nutrition at home suffer from a lower quality of life as they receive continuous therapy for up to 16 hours a day. It is not possible to interrupt nutriton therapy for everyday activities, as any interruption poses a health risk to the patient. Therefore, a ambulatory care services comes twice a day to connect and disconnect. The aim of this project is to develop a system that makes it manageable for patients to interrupt their feeding therapy in a self-determined way. In order to achieve this, enginners and social scientists collaborated to conduct an ergonomics study using the Design of Experiments(DoE) method to compare three commonly used connection systems. Analysis of variance(ANOVA) was used to secure the tests. In parallel, the probands were asked about their subiective feelings during the closure process. The results provided indications for the redesign of a home parenteral nutrition system. As safety is a crucial issue, the opening and closing process of the system was virtually simulated with the min. and max. finger force of the test persons in order to ensure a safe connection and disconnection with the system at all times. Overall, this participatory approach enable the design of the closure system to take account patients'fears wishes.

The formation of thrombi is still a challenge when it comes to ventricular assist devices (VADs). This can be caused by the subsequent deposition of proteins and platelets on foreign surfaces. In VADs, the blood contacting components are mainly made of titanium alloys due to hemocompatible properties and corrosion resistance. However, surfaces where mechanical contact of the rotor and the stator occur like bearing areas of axial VADs require high wear resistance due to the greater potential of scratches. Therefore, an in vitro test bench to optically investigate the deposition of fluorescent labelled platelets on the protein layer of different surfaces is hereby presented, aiming to investigate the potential for thrombi formation of five different inorganic hard material coatings on titanium. The coating materials included Titanium Nitride (TiN), Titanium Niobium Nitride (TiNbN), Diamond-Like Carbon (Ionbond-Medthin-43 (DLCI) and Oerlikon Balimed A (DLCO)) and Wolfram Carbid Oerlikon Balimed C (WC). Uncoated titanium (Ti) was investigated as reference. Heparinized human whole blood was incubated with Mepacrine (fluorescent dye) and then pumped through a flow chamber with defined shear rate conditions over the samples. The adhered platelets were visualized via inverted fluorescence microscopy. The analyzation of the green values and the binary image of the generated fluorescent picture offers conclusions about the platelet accumulation and the percentage of the covered surface area, respectively. Statistical analysis showed a significant lower potential for platelet deposition for TiN compared to Ti and no significant differences for the rest. It can be concluded that none of these coatings have a higher potential for platelet deposition than Ti and therefore point towards suitability for blood contacting components.

Amyotrophic lateral sclerosis is a progressive neuromuscular disease. Mechanical ventilation delays death and use at home can reduce costs and increase quality of life. The aim of the study is comparison of home mechanical ventilation with mechanical ventilation in a healthcare facility for adult patients with amyotrophic lateral sclerosis from the societal perspective in the Czech Republic. The Markov model from a previous Czech study was used and updated with data revelant to the societal perspective, which was obtained from the literature and opinions of experts from companies dealing with home mechanical ventilation. The cost-utility analysis was carried out over a time horizont of 10 years and results were presented in incremental cost-utility ratio. Accuracy and validity were verified by probabilistic sensitivity analysis and scenario analysis. The cumulative costs of home mechanical ventilation are CZK 6,116,394 and the cumulative utilities of home mechanical ventilation are CZK 8,195,848. The cumulative utilities of home mechanical ventilation are 3.55 QALY and the cumulative utilities of mechanical ventilation are 3.78 QALY. The incremental cost-utility ratio is CZK 9,094,042. Probabilistic sensitivity analysis and scenario analysis do not show significant changes. Home mechanical ventilation may be considered cost-effective for patients with amyotrophic lateral sclerosis from the societal perspective.

Avatars are often used to represent people in the virtual world. In the process, they can also have numerous psychological influences on the users. Therefore, this paper discusses a multiplayer application in virtual reality and how a appealing avatar could be designed in it. In a small study with six persons the application was tested competitively in teams of two, the users filled out a questionnaire and answered some additional questions. While an increase in motivation was observed in all participants, the choice of the preferred avatar was not uniform. Thus, this showed that an appealing avatar for everyone could not simply be determined, but personal preferences and experience must be taken into account.

A new approach that offers the potential for local drug delivery to the inner ear is a 3D printed, patient individualized, drug-loaded implant that precisely fits into the round window niche (RWN). Anatomically correct digital light processing (DLP) 3D printed implant prototypes are beneficial for preoperative planning and rehearsal of implantation techniques due to tactile feedback. The aim is to define desired mechanical material properties for future RWN implants. For this purpose, RWN implant prototypes (RWN-IPs) were DLP 3D printed using commercially available E-Shell 500 and E-Shell 600 materials (Envisiontec GmbH, Gladbeck, Germany) and a selfestablished PEGDA700 composition. These photopolymers are suitable for 3D printing RWN-IPs that feature different mechanical characteristics. The (1) mechanical properties (tensile test) were investigated, (2) the implantation feasibility and (3) fitting accuracy in human cadaver RWN were evaluated. As a result, E-Shell 500 has relatively high stretchability (ɛm ~ 60%) while E-Shell 600 and PEGDA700 are brittle and PEGDA700 has low strength. The E-Shell 500 material performs by far the best at handling and insertion. EShell 600 has adequate strength but is hard to handle because of rigid material behavior. PEGDA700 enables high 3D printing accuracy but lacks adequate mechanical behavior for adequate insertion of implant prototypes in RWN.

Deep learning with multiple hidden layers denoising autoencoders (MHL-DAE) is commonly used to denoise images and signals through dimension reduction. Here, we explore the potential of multiple parallel hidden layers denoising autoencoder (MPHL-DAE) to denoise complex bio-signals, like electrocardiogram (ECG). A merge layer, e.g., average layer is considered as the output of the proposed model by combining the outputs of the parallel hidden layers. The parallel hidden layers in the coding layer with activation function of different scale a, e.g., are considered to capture distinct features of the input. The lower/upper number of the required hidden neurons of the coding layer are estimated using data driven approach via singular values decomposition (SVD). The results show that the proposed MPHL-DAE model achieve better/similar SNR improvement compared to MHL-DAE with suitable scale for various noise levels respectively.

Patient satisfaction in total knee arthroplasty is still limited and lower compared to patient satisfaction in total hip arthroplasty. While there are multiple influence factors for patient satisfaction, implant design is one aspect to consider. In clinical practice, the patient-specific assessment of implant fit is limited to a matching of the anteroposterior and mediolateral size, based on X-rays. Hence, the goal of this review was to identify other potential criteria for a patientspecific implant fit evaluation. A literature review was performed in PubMed. Inclusion criteria were a patientspecific assessment of knee implant components, a comprehensible motivation or a statistically proven relationship with an outcome measure, and a clear method description. Implant fit was considered not only as a matching of size, shape and interface, but also to include the implant’s (rotational) alignment, as proposed by Dai et al. (2014). Several studies reported a statistically significant relationship of implant fit criteria with outcome measures. Many more studies proposed fit criteria, motivated by a potential relationship with outcome measures. Most criteria assessed the bone-implant interface, second the morphology of the articulating surfaces and third the component and resulting bony alignment. Obstacles for the assessment of the criteria identified is the limited availability of 3D surface models in TKA. However, with new imaging technologies this availability is likely to increase. The decision process regarding the implant in TKA could be supported by a comprehensive automated fit analysis of several designs and/or sizes. An adequate matching of the implant to the respective patient may improve patient satisfaction.

Exercise therapy is able to reduce symptom burden in advanced cancer patients (ACP). However, ACP daily form differs between days e.g. tumor and therapy induced. We included five ACP to classify individual exercise capacity due to cardiovascular parameters. Features are extracted from the electrocardiogram and then processed with a neural network after feature selection. Results indicate a high classification quality with an F1 score up to 0.95 ± 0.05. Including neuronal networks for training control can potentially help to manage exercise intensity ideal.

The usefulness of error-related potentials (ErrPs) for control in non-invasive Brain-Computer interface (BCI) research has been established over the last decades. To continuously correct for erroneous action of an end effector (e.g., robot arm) in a BCI however, these neural correlates relating only to the discrete perception of errors remain problematic. Using a pre-recorded dataset offering feedback in a 2D tracking task in different correct or erroneous conditions, we analyzed whether error processing during continuous feedback can be observed from the electroencephalogram (EEG). Within this dataset comprising 30 sessions of recordings, we were able to detect significant differences between correct and erroneous conditions. Furthermore, minimal significant difference between two erroneous conditions is reported, confirming the direct connection between error and cognitive response.

Patient-specific osteosynthesis plates can be used to reduce complications related to bone fracture treatment, such as infection, malocclusion and fatigue fractures of plates and screws. However, the implant design process is tedious. We propose a semi-automatic workflow to computationally design patient-specific titanium osteosynthesis plates for mandibular angle fractures. In this process, the plate stiffness is maximized while the mass is reduced. Two plate designs with different numbers of screw holes (implant #1 with four holes, implant #2 with eight holes) were generated with identical topology optimization settings and compared in a finite element model simulating various biomechanical masticatory loads. Differences in von Mises stresses in the implants and screws were observed. The load case of clenching the jaw on the opposite side of the fracture showed the highest stress distribution in implant #1 and higher peak stresses in implant #2. Stress concentrations were observed in sharp corners of the implant and could be reduced using local stress-based topology optimization. We conclude that the design process is an effective method to generate patientspecific implants.

Hydrogels are 3D polymeric networks, which exhibit properties such as softness, viscoelasticity and their ability to absorb large amounts of water. These characteristics make them exceptionally suitable in biomedicine as e.g. tissue scaffolds, drug delivery systems, wound dressings or contact lenses. One of these hydrogels is the biocompatible, hydrophilic and photopolymerizable poly(ethylene glycol) diacrylate (PEGDA). It is used in different biomedical applications due to its tunable mechanical characteristics. In our study, the mechanical properties of different PEGDA hydrogel compositions with variyng molecular masses and contents of water/methanol, were investigated. Different compositions containing 20 m%, 30 m% or 40 m% of PEGDA 4K (4,000 g/mol), PEGDA 10K (10,000 g/mol) or PEGDA 20K (20,000 g/mol) in ultrapure water/methanol (1:2) were produced. Dumbbell-shaped samples were prepared in molds via photopolymerization in a UV chamber. Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) was used as photoinitiator (0.5% w/w). The mechanical testing was performed using a uniaxial testing system. The obtained results showed averagely 78% higher tensile strength (σmax) values for 30 m% and 40 m% samples in comparison with 20 m% samples for all of the tested polymers. PEGDA 20K 30 m% samples showed the highest σmax among all of the samples with 12.8 MPa. All of the PEGDA20K samples exhibited the highest elongation at break (εB) values (up to 958%), whereas the lowest values were found for PEGDA 4K (up to 105%). The obtained stress-strain curves for most of the samples were typical for deformable, amorphous polymers with a deformation upon reaching a critical stress point. The PEGDA materials showed variable mechanical characteristics according to changing molecular mass or polymer concentration. These promising results showed that it should be possible to compose scaffolds with desired mechanical stability according to the needed application.

Aspiration catheters play an important role in mechanical thrombectomy in interventional neuroradiology as they provide support for stent retrievers or are increasingly used separately for contact aspiration. To evaluate the performance of different aspiration catheters in terms of suction force, most current studies rely on manual measurement methods or the suction force is not measured close to the aspiration catheter tip. Moreover, there is currently no standardized method that records measurements digitally and with high temporal resolution. To overcome these challenges, we propose a new test bench that simulates a dedicated aspiration scenario and takes measurements close to the aspiration catheter tip. To meet the requirements, a simplified vascular phantom, a pressure sensor and a simplified clot model were implemented in the prototype. To demonstrate the applicability of the prototype, two commonly used aspiration catheters were evaluated. The obtained measurements show that it is possible to perform standardized comparisons between aspiration catheters from different manufacturers as well as different aspiration pumps and aspiration tubings and their combinations. The developed prototype overcomes the still widely used manual and semimanual and thus error prone measurement methods for the evaluation of vacuum based thrombectomy systems.

The development of a non-invasive, cuffless, continuous, wearable device for the measurement of blood pressure is a complex endeavour due to the high specificity at each measuring site and the need for high accuracy. Proof-ofconcept and validation of a prototype should be performed at an early stage for functionality assessment. Additionally, the emergence of biological computer models allows for in-silico research, which results should be verified in a practical experiment. To grant an optimal preliminary assessment of a prototype, this work aimed to develop and validate accurate in-vitro and ex-vivo arterial models, with simple construction and easily available components. The comparison between a silicone tube and a porcine artery as a mimicked human radial artery was based on the stiffness parameter. Flow pressure is controlled by a centrifugal heart-like pump. Pressure values are extracted with ultrasound and a commercial piezoresistive pressure sensor is used for pressure validation. The porcine artery showed much more realistic stiffness values (·=15.360) than the silicon tube (·=543.420), which was very stiff in comparison to the typical in-vivo radial artery stiffness (·=9.5). The decrease in stiffness of 97.173 % (from the silicone tube to the porcine artery) led to an acute decrease in the derived pressure error. This work serves as guidelines for the development of a low-budget arm phantom, as the simple setup allowed for a primary validation of a proof-of-concept ultrasound-based sensor for the measurement of pressure.

Surgical navigation-assisted interventions, providing augmented previously acquired image data to support the physician, are routinely used. Visualization techniques like extended reality (XR) promise improvements because the desired information can be shown where most suitable. The accuracy of such systems is essential for usability in clinical interventions but lacks research. This work shows a system to project thermal data for use in reconstructive surgery and investigates the accuracy of the XR technology. The system consists of a thermal camera (PI450, Optris GmbH, Germany), a depth camera system (D415, Intel Inc., USA), and a HoloLens2 (Microsoft Corp., USA). 5 probands annotated the locations of 9 marker points on a calibration pattern of the real object and the projection in the XR headset. The mean and max residual projection errors were 2.8 and 6.2 mm, respectively. An XR system to superimpose thermal images was shown. The mean projection error of 2.8 mm is sufficient for further exploration of medical applications.

Deep learning is often used for automated diagnosis support in biomedical image processing scenarios. Annotated datasets are essential for the supervised training of deep neural networks. The problem of consistent and noise-free annotation remains for experts such as pathologists. The variability within an annotator (intra) and the variability between annotators (inter) are current challenges. In clinical practice or biology, instance segmentation is a common task, but a comprehensive and quantitative study regarding the impact of noisy annotations lacks. In this paper, we present a concept to categorize and simulate various types of annotation noise as well as an evaluation of the impact on deep learning pipelines. Thereby, we use the multi-organ histology image dataset MoNuSeg to discuss the influence of annotator variability. We provide annotation recommendations for clinicians to achieve high-quality automated diagnostic algorithms.

Introduction: Muscle fatigue is often experienced by athletes and in work settings. Excessive fatigue can lead to injury and musculoskeletal disorders. Surface electromyography (EMG) is typically used to detect and ultimately prevent fatigue during isometric movement. The application of EMG to fatigue detection in dynamic movement requires, however, a secondary confirmation of fatigue based on physiological measures. Our objective was to determine if muscle oxygenation derived via near-infrared spectroscopy (NIRS) was correlated with relevant EMG indicators of neuromuscular fatigue and whether observed correlations were related to the fatigue process. Methods: Bilateral electromyograms from three upper leg muscles and the tissue oxygenation index (TOI) of the vastus lateralis muscle were recorded in sixteen non-disabled individuals during cycle ergometry to volitional exhaustion. Six EMG activity features were extracted and the Pearson correlation coefficient between each feature and TOI was determined. Results: The EMG root mean square, spectral standard deviation, second spectral moment, and zero-crossing rate (ZC) were strongly correlated with TOI. The time course of ZC and the correlation of this feature with TOI suggest that there could be a relation between muscle oxygenation and fatigue. Conclusion: Future work should use the knowledge gained in this study to investigate whether NIRS can be used to verify the onset of fatigue as detected by EMG.

Introduction: Low back pain(LBP) affects many individuals and is known to be associated with impaired trunk control. While LBP can be associated by treating underlying trunk impairments, a better understandinf of the mechanistic origin of the disorder is required. Trunk control has been commonly studied via an unstable sitting paradigm. Knowledge on how the base of support and body segments work together to complete the unstable sitting task, and how this is modified in individuals with LBP, could be utilized when designing interventions for this population. Our obiective was to characterize the segmental coordination in non-impaired unstable sitting as elicited via a wobble board(WB) paradigm. Methods: WB, pelvis, and trunk motion were recorded in fifteen non-disabled participants sitting on a wobble board. We used cross-spectral analysis to quantify the coordination of the anterior-posterior angular kinematics of the wobble board, pelvis, and trunk. Results: During unstable sitting, the motion of the pelvis was followed by that of the trunk(one-eighth-cycle delay) and wobble board(half-cyclw delay) at frequenties between 1 and 2 Hz. Conclusion: Future work should utilize the knowledge gained in this study when creating rehabilitation interventions for individuals with LBP.

Introduction: gait-related falls account for most falls in the elderly. The identification of individuals at risk of falling due to unstable gait requires clinically feasible measures that can detect impairments in gait. Our obiective was to : (1) assess the feasibility for clinical implementation of existing gait stability measures and selected measures to detect gait impairments. Methods: Nine measures were assessed for clinical feasibility, with only the margin of stability(MOS) being selected for evaluation. Ground reactions and motion of the lower body were recorded in fifteen non-disabled and three disabled participants during treadmil walking. Media-lateral MOS was calculated and compared between the non-disabled and disabled participants to evaluate its ability to detect gait impairments. Results: MOS values at heel strike(HS) and at the point between HS and contralateral toe-off with minimum MOS deviated between participants with lower limb impaiments and non-disabled participants. Conclusion: MOS at HS demonstrated the greatest potential to assess fall risk. Additional work is required before MOS can be recommanded for clinical use.

Blood sampling typically involves puncturing a superficial vein with a needle and then creating a vacuum (e. g., with a syringe) to transfer the blood from the venous system to the collection container. In some cases, the venous walls collapse due to excessive vacuum generated. In this work, a device is presented that prevents venous collapse and can be combined with any blood collection system available on the market. The mechanism of preventing venous collapse is based on a tubular segment interconnected between the needle and the syringe, which collapses at a lower vacuum pressure than the vein. The prototype was tested and evaluated using a test rig that simulates venous blood draw. The experiments proved the functionality of the device on a vein phantom.

The amount of the toxic waste generated in the hospitals and medical facilities is steadily increasing. Although, the centralized disposal or waste management service solutions for the removal of toxic waste are available in highly developed industrialized countries; this does not, however, apply to the developing and emerging countries. For this reason, a compact sterilization and shredding machine was developed. The steam sterilization used (temperatures of 134 °C + 4 K and pressures of 480 to 3500 mbar) complies with the recognized procedure of the Robert Koch Institute (without chemical substances). Thus, it was possible to transform the infectious or toxic waste into non-hazardous household waste. The microbiological and physical test results show that the machine meets the requirements for sterilization of class C solid and liquid clinical wastes. The capacity of the machine is maximum 400 liters of waste per cycle, with an average cycle time of about 1 hour. It is possible to process contaminated materials both loose and in sealed disposal containers (C-boxes).

Heart failure (HF) is a chronic disease with high hospitalizations rates and increased mortality. The six-minute walk test (6MWT) is a standardized method to assess the functional health of HF patients by measuring the distance walked in six minutes [1]. The improved 6MWT with new filter algorithms, based on the development of smartphone application [9], can be used everywhere with a good global navigation satellite system (GNSS) signal. This paper tests multiple filtering approaches to reduce the measured distance error in the application.We were able to decrease the average relative error of the application from 4.5% to 2.0%. Additionally, the number of experiments with relative error over our chosen clinically significant difference of 5% is reduced by 90%. Keeping these results in mind, the system can be considered clinically acceptable.

In robot-assisted surgeries, the surgeon focuses on the surgical tool and its pose, and not on the complete robot’s shape. However, the joints of redundant robots (robots that have more degrees of freedom (DoF) than needed for the positioning of surgical tools) might move in unexpected/undesired ways. Joint motions that lead to patient or collisions are safety critical. We assume that the medical personnel in the operating room can best decide if a planned robot motions come too close to the patient or not. Therefore, we propose an augmented reality-based solution to interact with the robot during surgery planning, and intervention. The tool can be used to command a robot by drawing a trajectory in augmented reality (AR), visualizing the robot movement to check if it is safe before execution. The proposed solution allows surgeons to plan safe robot motion paths before-hand and adapt them when necessary in situ. As a proof-of-concept, we implemented and demonstrated the proposed solution on a 7-DoF redundant robot by commanding different trajectories. The control architecture to plan and execute motion for a surgical robot using AR is a key result of this work.

Deep Learning-based tissue classification may support pathologists in analyzing digitized whole slide images. However, in such critical tasks, only approaches that can be validated by medical experts in advance to deployment, are suitable. We present an approach that contributes to making automated tissue classification more transparent. We step beyond broadly used visualizations for last layers of a convolutional neural network by identifying most relevant intermediate layers applying Grad-CAM. A visual evaluation by a pathologist shows that these layers assign relevance, where important morphological structures are present in case of correct class decisions. We introduce a tool that can be easily used by medical experts for such validation purposes for any convolutional neural network and any layer. Visual explanations for intermediate layers provide insights into a neural network’s decision for histopathological tissue classification. In future research also the context of the input data must be considered.

Introduction: In the past years, some groups established methods to use promontory stimulation and subjective responses as a tool for preoperative stimulation of the cochlea before cochlear implant (CI) surgery by a temporary transtympanic needle placed on the middle ear. Our research group tested CI candidates when the presence of the auditory nerve could not be confirmed by other pre-operative tests and analyzed objective results of electrical auditory brainstem response recorded with trans-tympanic promontory stimulation in local anesthesia (LA-TT-EABR) with a “golfclub” electrode placed on the round window niche (Polterauer et al. 2018). However, EABR recording suffers from an electrical artifact from the stimulation and focuses on the excitability of the solely auditory nerve. We hypothesize that late evoked potential response (EALR) can be used for the assessment of the auditory pathway up to the cortical area with less electrical artifact interference. Methods: The trans-tympanic electrically evoked auditory late response in local anesthesia (LA-TT-EALR) was added to our protocol right after the LA-TT-EABR in a unique recording session. In this study, we investigated the feasibility of LA-TT-EALR using MED-EL hardware and software. These two pre-operative trans-tympanic measurements, LATT- EABR and LA-TT-EALR, were compared. Results: In a total n=18 tested subjects, it could be possible to record LA-TT-EABR in 10 and LA-TT-EALR in 13. In contrast to the fact that acoustically evoked auditory late responses are often hard to detect, and an electrical artifact heavily affects LA-TT-EABR, we did not experience these problems on LA-TT-EALR recordings. The matching between two measurements was present in 67% of the cases. Conclusion: LA-TT-EALR may be a complement to the existing state-of-the-art LA-TT-EABR for pre-operative assessment before cochlear implantation offering information about auditory pathways in the auditory brainstem and cortex with a small increase of recording time. The equivalency of these two methods is encouraging.

In-vitro models for the transurethral resection of the prostate (TUR-P) require prostate phantoms whose mechanical, thermal, and electrical properties resemble those of real tissue. This work focused on the latter. Phantoms were fabricated with gelatin, pectin, and corn syrup solved in an aqueous solution of 0.4 wt% NaCl. The effect of each component on the phantom’s electrical properties could be described in a multiple nonlinear regression model. As a result, the electrical current drawn during resection of the phantom could be predicted with an uncertainty of ± 10 %. This allows for purposive adjustment of the phantom’s composition to regulate the Joule heating during resection. Particularly the pectin concentration appeared to greatly affect the electrical properties as well as the mechanical stability of the artificial tissue at higher temperatures caused by Joule heating of the system. Until now, this research focused on the development of tissue that resembles porcine musculus longissimus dorsi (MLD). In the future, the phantom’s composition will be adjusted based on data obtained through in-vivo resection of the prostate.

With a growing number of people in need of care and the shortage of skilled workers, caregivers express the desire for improved support to reduce physical strain and psychological stress. One task for burden relieve is documentation of nursing actions and treatment processes, as this can be done in parallel with the action. Hence tools capturing information in real time could improve information sharing. Also, contactless assessment of patients’ vital signs and emotions is desirable. Thus, the idea is to enhance an existing mobile platform, which can follow the nurse, includes a voice-controlled assistant, and be able for wireless (e.g., via Bluetooth) assessment of vital data from patients in its direct vicinity. Using the Temi telepresence robot as an open technical platform, we provide and discus requirements for this platform to potentially aid and relief to medical staff within caring and elderly homes, with respect to voice-control, Bluetooth-based vital data collection, interfaces to a digital patient manager and automated emotion recognition. The presented concept and the rating of the associated technologies leads to the conclusion that - if adequately implemented and installed - the caregivers could potentially receive noticeable relief from them. Nevertheless, as next steps, the examined technologies and ideas must be completely implemented, thoroughly fused with each other, and evaluated to provide a real support system for clinical care and assisted documentation.

Congenital Central Hypoventilation Syndrome (CCHS) leads to insufficient breathing especially during sleep. While patients can be ventilated and monitored during the night, adolescents undergo changes in their biorhythm and lifestyle and are at higher risk to inadvertently fall asleep and stop breathing during the day. In this paper, we propose a prototype mobile wake-up device worn behind both ears that constantly measures oxygen saturation and heart rate noninvasively. If a dangerously low level of saturation is detected, it is assumed that the user fell asleep. To wake them, the device issues an acoustic and haptic alarm. The device settings are accessed via a smartphone app. To evaluate the sensor performance, a subject simultaneously wore the prototype behind his ears and a commercial finger pulse oximeter on his hand. Heart rate and blood oxygen saturation were taken for three measurements of few minutes each. Mean absolute error of oxygen saturation was less than 2 % for two data sets and 6.1 %for the third. It is assumed that during this measurement, the device lost contact to the skin. The heart rate measurement showed satisfactory agreement of < 4 % mean average error. In the future, the measurement quality will be improved by calibration and more reliable attachment to the ear and a user study will be conducted in order to evaluate the prototype usability.

Due to the direct contact between electrode and scalp, dry EEG electrodes are exposed to increased mechanical wear compared to conventional gel-based electrodes. However, state-of-the-art commercial cap systems commonly use permanently fixated electrodes which can lead to downtime of the EEG cap during professional repair and replacement as well as reduced overall lifetime. An easily replaceable EEG electrode would furthermore improve hygiene, especially for newborn and infant applications. We propose a novel replaceable electrode system, consisting of an electrode holder, a snap top, a contact ring fixated inside the electrode holder, and a replaceable electrode. The production process consists of 3D printing, silicone molding, resin casting, and electroless plating. The replaceable electrode system is integrated into a multichannel EEG cap system. A verification study is conducted with 30 volunteers. The operators experienced that the new electrode holder eases adjustment of the electrode to have proper contact with the scalp. During the study, defective electrodes can be replaced without a soldering process. Furthermore, all electrodes stayed in the holder and did not fall off the cap for the whole session. In conclusion, the novel replaceable electrode system is suitable for EEG measurements.

Computational fluid dynamics can be used to analyze the flow pattern and the anatomical predisposition of the airway collapse in patient-specific simulations of sleep apnea patients. Especially the nasal geometry is a major challenge, because of the complex structures and high calculation effort. Due to this limitations the nasal cavity is often excluded. To investigate the influence of the exclusion of the nasal geometry on the flow in the pharynx, four digital models were created. One, which includes the complete nasal structure, and three simplified models, neglecting the nasal geometry using different cutting planes. Airflow simulations using the k-ε turbulence model are performed with an inlet flow rate of 10 l/min. The results show a good agreement of the average velocities. Deviations of the velocity and pressure values occur particularly in the region of the oropharynx and indicate that the type of simplification has a significant influence on the pressure distribution.

Electroencephalography (EEG) is widely used in clinical applications and basic research. Dry EEG opened the application area to new fields like self-application during gaming and neurofeedback. While recording, the signals are always affected by artefacts. Manual detection of bad channels is the gold standard in both gel-based and dry EEG but is timeconsuming. We propose a simple and robust method for automatic bad channel detection in EEG. Our method is based on the iterative calculation of standard deviations for each channel. Statistical measures of these standard deviations serve as indications for bad channel detection. We compare the new method to the results obtained from the manually identified bad channels for EEG recordings. We analysed EEG signals during resting state with eyes closed and datasets with head movement. The results showed an accuracy of 99.69 % for both gel-based and dry EEG for resting state EEG. The accuracy of our new method is 99.38 % for datasets with the head movement for both setups. There was no significant difference between the manual gold standard of bad channel identification and our iterative standard deviation method. Therefore, the proposed iterative standard deviation method can be used for bad channel detection in resting state and movement EEG recordings.

Introduction: The human breath is an accurate but complex read-out of many physiological processes in the organism that can be monitored via volatile organic compounds (VOCs) in the exhaled air. However, there are many confounding variables that limit the transfer and application of breath analysis to become a clinical procedure. Method: This work aims to establish a systematic procedure for sampling and characterization of various everyday influences of healthy subjects using proton transfer reaction-mass spectrometry (PTR-MS). In order to limit the influencing factors on the breath profile, a standard analysis procedure for sampling and evaluation of the exhaled breath samples was developed. The correlations between the selected experimental conditions and the resulting VOC profiles were investigated using a non-parametric Wilcoxon rank sum test. Results: In addition to the relevant influence of methodological experimental parameters, interesting insights into the effect of everyday factors on the exhalat gas were obtained and discussed. Furthermore, subject and condition-specific differences were found in the exhaled air of male and female subjects. Conclusion: With a more robust, standardized and reproducible breath sampling protocol, breath analysis is a promising non-invasive tool towards a system-wide understanding and personalized diagnosis and treatment of a wide range of diseases.

During laparoscopic procedures, the surgeon's view of the situs, and thus her or his performance, is dependent on the skills of the camera assistant. The surgeon lacks control over his own field of view and there is a high potential for reducing mental load and workflow issues. In this paper, a research setup of a 360° laparoscopic imaging system is presented and evaluated. The system consists of a 360° camera and a head-mounted display, through which the surgeon can inspect the situs. In a user test, the system showed advantages over a conventional laparoscope regarding orientation in the situs, intuitiveness of operation, and faster task completion.

Surface electromyography (sEMG) records the electrical potentials on the skin surface generated by the electrical activity of muscle fibers. To better understand physiology and to assess signal processing algorithms, sEMG models have been developed in the past. However, numerical models required for modeling accurate geometries very often have a high computational complexity rendering realistic simulations challenging. In order to cope with this challenge, a pipeline is investigated to accelerate the calculation of the single fiber action potential (SFAP) based on the principle of reciprocity. To enable a comparison to an analytical solution, the investigation is carried out in a highly simplified muscle model. The results show that the pipeline is suitable for calculating SFAP with a low mean absolute error sufficient for application in sEMG models. In addition, there is a significant reduction of the calculation time allowing the simulation of even highly complex geometries and a large number of muscle fibers.

Radial compressors, Blowers, which are often used in modern devices for medical ventilation, are known to have non-linear dynamics regarding the generated pressure. To make them accessible for the large field of linear control theory their model needs to be linearized. This paper uses a gray box model of a blower driven by a brushless direct current (BLDC) motor to design a controller based on feedback linearization. The controller is designed to achieve approximate linear closed loop behavior of the system. The controller is then evaluated concerning reference tracking and disturbance rejection in simulation and on a demonstrator system. The evaluation showed that an approximate linear closed loop behavior was achieved.

Introduction: Early detection of adenomas and polyps is one central goal of colonoscopic screening programs. As the adenoma detection rate (ADR) depends on the experience of the endoscopist, AI-based polyp detection systems can be used for real-time assistance. Hence, to support the physicians such AI-based systems using deep-convolutional neural networks (DCNNS) have been introduced in the past years. One disadvantage of these techniques is the need of a huge amount of labeled training data. Method: We investigate a "human-in-the-loop approach" to minimize the required time to generate labeled training data. The approach is evaluated within the training a YOLOv4 neural network to detect polyps in colonoscopic image data. The performance metrics of the neural network are evaluated on three public datasets. Results: The performance of the YOLO network increased from a precision of 0.88, recall of 0.83, F 1 score of 0.86, and a F 2 score of 0.86 to a precision of 0.91, recall of 0.87, F 1 = of 0.89, and F 2 = of 0.88. The interactive labelling of 1,000 images only takes one hours. Conclusion: The proposed "human-in-theloop approach" is capable of generating labelled image data in a minimum of time while increasing the performance metrics as well. For higher performance increase more data can now be labeled within this new approach.

Additive manufacturing of dental prostheses e.g. composite materials is continuously finding broader application in dentistry. In the oral environment, the materials are subjected to hydrolysis and accelerated hydrolysis by enzymes. With the aim to compare an additively manufactured composite with a self-curing and a CAD/CAM material of the same composition, the effect of enzymatic degradation and hydrolysis on the materials were investigated. The experimental composites consisted of 50 wt. % inorganic filler and 50 wt. % monomer matrix (BisEMA, BisGMA, TEGDMA). The influence of hydrolysis via a buffer system was investigated in comparison to enzymatic degradation by cholesterol esterase over 22 days. The detection of degradation products over time was conducted by HPLC-DAD analysis. Degradation products and monomers calibrated were: TEGDMA, BisGMA, BisEMA, MA, TEGMA, Bis-HPPP and E-bis-PA. Surface roughness of polished specimens was measured by AFM and hardness with a Vickers micro hardness tester. After enzymatic degradation and hydrolysis, BisGMA was detected in the eluates of the additively manufactured composites. The samples fabricated by CAD/CAM behaved similar during hydrolysis, where additionally MA was detected. After enzymatic degradation, furthermore TEGMA and BisEMA were detected qualitatively. In contrast, the self-curing material exhibited traces of MA, TEGMA, TEGDMA and BisGMA after enzymatic degradation. Hereof detectable upon hydrolysis were MA, TEGMA and BisGMA. Surface roughness was comparable for all manufactured samples whereas the hardness was lowest of the 3D printed material. Enzymatic degradation and hydrolysis had no effect on surface roughness and hardness. The additively manufactured composite appeared to have low susceptibility to enzymatic degradation and hydrolysis. Due to the manufacturing process, the polishability and initial hardness are poorer.

Introduction: Carbohydrate malabsorptions symptoms include intestinal fluid retention, causing diarrhea and abdominal distention. The aim of this work is to create a machine learning model that predicts carbohydrate malabsorption using H 2 measurements from lactose and fructose tolerance tests. Methods: We compare the predictive ability of popular classifiers with classifiers that are specifically designed for time series data. Our approach was implemented using sklearn and sktime Python machine learning libraries. Results: The highest predictive ability for the fructose dataset was achieved using a Random Forest Classifier (balanced accuracy = 0.91). In contrast, the highest predictive ability (balanced accuracy = 0.81) for the lactose dataset was obtained using an IndividualTDE time classifier. Conclusion: Our results indicate a high predictive ability for distinguishing between carbohydrate malabsorptions. However, the detection of SIBO is challenging but adapted time classifier models could reach higher performances compared to standard methods. Our results could establish the basis of an expert system for diagnosing carbohydrate malabsorptions and SIBO, respectively.

Nowadays, needle artifacts are an essential restriction for MRI-guided interventions, as they influence the visually perceived needle size and needle destination. Standard MRI needles made of Nickel-Titanium (NiTi) alloys still produce massive artifacts in MRI due to materials’ interactions with the magnetic environment. The use of non-metallic materials can reduce these artifacts. Therefore, we propose a nonmetallic concept of a coaxial needle design concept with a fiber-enforced inner core and an outer polymeric hollow sheet. This work aims to evaluate the artifact performance of the proposed needles in a custom-made three-dimensional (3D) tumor model with a relevant size and thickness for medical interventions under MRI guidance using a 3T field strength with a T1-weighted gradient-echo sequence. Three coaxial MRcompatible needles were inserted in the custom-made 3D tissue phantom, one standard needle from NiTi, and two proposed non-metallic needles. Artifact’s width and length were measured for each needle. Overall, the non-metallic needles showed significantly lower artifacts than the standard NiTi needle inside the 3D tumor model.

Diabetic foot syndrome is a secondary disease of diabetes mellitus that can lead to amputations, reduced mobility and in some cases death if treatment is delayed. The objective of this study was to collect user requirements for the development of a shoe insole accompanied by a medical app for continuous pressure measurement, applying a usercentered Design Thinking process.We conducted a systematic competitor analysis and performed 10 context-interviews with both health care professionals (HCPs) (n=7) and patients with diabetic foot syndrome (n=3) at the University Hospital Graz (Austria) and a medical supply store. Structured interview guides, tailored to user groups were used for qualitative interviews. Questionnaires on the acceptance of eHealth applications were designed for quantitative analysis (n=14 patients, n=7 HCPs). The competitor analysis identified 12 inshoe- systems for plantar pressure measurement for patients with diabetes. No current system on the EU market meets the requirements of the users interviewed in this study. Quantitative evaluation showed that 95% of future users would support the use of a medical app to monitor plantar pressure in patients with diabetic foot syndrome. Based on the identified user requirements, a first user interface was developed iteratively. Usability tests of the prototype resulted in a score of 71.7 on the System Usability Scale. The medical app will be further developed and tested in clinical studies.

Wound healing after implant surgery represents a spontaneous repair process that, when disturbed, can lead to a pathological situation known as fibrosis, characterized by excessive proliferation, increased extracellular matrix deposition, and scarring of the tissue. Tissue-specific cell culture models simulating physiological or pathophysiological cell responses are valuable research tools in the field of fibrosis, among others. In such in vitro systems, endpoint measurements are typically used to assess cell viability and/or proliferation. However, this approach is usually time consuming, costly and often leads to the destruction of the sample. Additionally, these conventional in vitro systems provide only a single snapshot of the respective sample at the end of the experiment. Therefore, we aimed to evolve the fibrotic disease in vitro model established by Stahnke et al. To this end, we chose a time-resolved, non-invasive approach based on cell impedance measurements and bright-field microscopy that allows continuous monitoring of the cells. Such measurements can be directly conducted using the xCELLigence RTCA eSight system. This system was used to characterize the cellular responses of the fibrosarcoma cell line HT-1080 and primary fibroblasts isolated from the Tenon’s capsule of human donors to TGF-β1, a key cytokine involved in the pathogenesis of fibrosis. The data shows that compared to Tenon’s fibroblasts the response of HT-1080 cells to TGF- β1 is less pronounced and only detectable within a narrow time frame. Thus, HT-1080 do not appear to be a suitable cell culture model for the assessment of antifibrotic implant coatings.

Minimally invasive procedures such as Natural Orifice Transluminal Endoscopic Surgery (NOTES) require powerful, small, and flexible instruments. A cable-driven instrument was developed, which is able to retract tissue to create sufficient space for an actual operation (e.g. cholecystectomy). In this paper, the work space of a developed instrument is presented. The work space is calculated using direct kinematics equations and verified by measurement using an electromagnetic (EM) tracking system. The angular orientation of the instrument can be up to 85° with a length of the active section of 60 mm. However, a longitudinal rotation up to 17° becomes apparent. This is due to the characteristics of the steel cable used for actuation. Nevertheless, the instrument reaches the intended work space. Further measurements are necessary to evaluate the instrument’s behavior under payload and whether this affects the work space.

3D cell culture models are important tools for the development and testing of new therapeutics. In combination with immunoassays and 3D confocal microscopy, crucial information like morphological or metabolic changes can be examined during drug testing. However, a common limitation of immunostainings is the number of dyes that can be imaged simultaneously, as overlaps in the spectral profiles of the different dyes may result in cross talk. We therefore present a 3D deep learning method, able to predict fluorescent stainings of specific antigens on the basis of a nuclei staining. Using the proliferation marker Ki67, we showed that the presented model was able to predict the Ki67 staining with a strong correlation to the real signal. Additional analysis showed, that the model was not relying on signal cross talk. This approach, based on staining of the cell nuclei and subsequent prediction of the target antigen, could reduce the number of parallel antigen stains to a minimum and incompatible staining panels could be circumvented in the future.

Some magnetoelectric sensors require predefined external magnetic fields to satisfy optimal operation depending on their resonance frequency. While coils commonly generate this external magnetic field, a microelectromechanical systems (MEMS) resonator integrated with permanent magnets could be a possible replacement. In this proof-of-concept study, the interaction of a MEMS resonator and the ME sensor is investigated and compared with the standard approach to achieve the best possible sensor operation in terms of sensitivity. The achievable sensor sensitivity was evaluated experimentally by generating the magnetic excitation signal by a coil or a small-sized MEMS resonator. Moreover, the possibility of using both approaches simultaneously was also analysed. The MEMS resonator operated with 20V pp at 1.377 kHz has achieved a sensor sensitivity of 221.21mV/T. This sensitivity is comparable with the standard approach, where only a coil for sensor excitation is used. The enhanced sensitivity of 277.0mV/T could be identified by generating the excitation signal simultaneously by a coil and the MEMS resonator in parallel. In conclusion, these MEMS resonator methods can potentially increase the sensitivity of the ME sensor even further. The unequal excitation frequency of the MEMS resonator and the resonance frequency of the ME sensor currently limit the performance. Furthermore, the MEMS resonator as a coil replacement also enables the complete sensor system to be scaled down. Therefore, optimizations to match both frequencies even better are under investigation.

Acoustic or visual warning signals for workers in hazardous situations might fail under loud and/or lowvisibility work situations. A warning system that uses electrocutaneous stimulation can overcome this problem. The aim of this pilot study was to find spatio-temporal stimulation patterns for appropriate electrical warning. Eight electrode pairs were attached to the upper right arm of 16 participants. The stimulation was conducted with bi-phasic rectangular pulses of 150 μs and an amplitude of up to 25 mA. Pulse intervals that generate a single pulse, pulsating, vibrating, and continuous perception as well as varying spatial patterns (e.g. alternating between electrode pairs or circumferentially around the arm) were investigated and evaluated with regard to alertness, discomfort, and urgency. The pilot study revealed that a stimulation signal that generates a vibrating perception and is applied as a circumferential signal around the arm showed the highest values of alertness and is therefore considered a potential warning pattern for future studies with larger study groups.

Ultrasound measurements are a widely used instrument in clinical practice. For later traceability of the images, the position (and orientation) of the ultrasound probe must be recorded during the measurement. Until now this has to be done manually by the physician. An easier and more accurate approach would be the automatic tracking of the ultrasound probe. This contribution shows a first approach for automatically localizing the ultrasonic head during measurement. The proposed method is based on coils surrounding the patient bed and a 3D magnetic sensor placed on the ultrasound head. Besides some pre- and postprocessing steps, the proposed localization algorithm is based on trilateration followed by a least mean squares approach for refinement of the estimation. In a first proof-of-concept measurement with fixed positions and orientations of the ultrasound head a mean accuracy of 2 . 85 cm and 8 . 94 ◦ was achieved. Additionally, a measurement with a moving ultrasound head is presented to demonstrate the real-time capability of the system. Finally, future steps for improving the automatic measurement are discussed, including a graphical user interface for the physician and the use of magnetoelectric sensors for measurement.

Introduction: In the field of colonoscopy results of automated or supporting procedures using medical image processing are often compromised by contaminants such as blood or stool, but also due to medical instruments. Such objects or contaminations often obscure relevant areas or cause disturbances in various algorithms and also prevent physicians from recognizing relevant information in the video-stream. Hence, automatic detection and segmentation of such contaminations is desirable, by obtaining proper segmentation masks to an algorithm yields the possibility to exclude such areas from analysis. Method: We created a training and validation dataset by collecting (and annotating, where necessary) endoscopic video sequences from various public available datasets and a privat collection. An encoder-decoder deep convolutional neural network was trained with this data to predict segmentation masks for blood, stool, medical instruments and other objects. Results: Depending on the class (blood, stool, ... ), we were able to obtain a mean Dice score in the range between 0.88 to 0.93 on the evaluation dataset. Conclusion: The proposed approach hence allows us providing segmentation masks that are able to reduce problems in the subsequent image analysis and also find regions that were not properly observed.

In this work, an attempt has been made to analyze the shape variations in mediastinum for differentiation of Coronavirus Disease-2019 (COVID-19) and normal conditions in chest X-ray images. For this, the images are obtained from a publicly available dataset. Segmentation of mediastinum from the raw images is performed using Reaction Diffusion Level Set (RDLS) method. Shape-based features are extracted from the delineated mediastinum masks and are statistically analyzed. Further, the features are fed to two classifiers, namely, multi-layer perceptron and support vector machine for differentiation of normal and COVID-19 images. From the results, it is observed that the employed RDLS method is able to delineate mediastinum from the raw chest Xray images. Eight shape features are observed to be statistically significant. The mean values of these features are found to be distinctly higher for COVID-19 images as compared to normal images. Area under the curve of greater than 76.9% is achieved for both the classifiers. It appears that mediastinum could be used as a region of interest for computerized detection and mass screening of the disease.

Modern medical technology offers potential for the automatic generation of datasets that can be fed into deep learning systems. However, even though raw data for supporting diagnostics can be obtained with manageable effort, generating annotations is burdensome and time-consuming. Since annotating images for semantic segmentation is particularly exhausting, methods to reduce the human effort are especially valuable. We propose a combined framework that utilizes unsupervised machine learning to automatically generate segmentation masks. Experiments on two biomedical datasets show that our approach generates noticeably better annotations than Otsu thresholding and k-means clustering without needing any additional manual effort. Using our framework, unannotated datasets can be amended with pre-annotations fully unsupervised thus reducing the human effort to a minimum.

In collaboration with the Institute of Virology at the University of Marburg, a tracking algorithm has been developed in recent years. Thereby, for the tracking of subviral particles and their intracellular transport through the cell, the detection of these particles is the first major step. In this work a deep-learning based approach for automatic processing of fluorescence video sequences of virus infected cells is presented. The focus of this work was to generate synthetic data for training neuronal networks and to represent a particle using only a single delta peak. It was shown that a simple convolutional network is able to predict the particle positions with an almost identical accuracy as the algorithm developed by Kienzle, Rausch and colleagues [1-3], and that even a small dataset is sufficient for that purpose.

Chronic pain is a global health problem, having a tremendous effect on the quality-of-life (QoL) of patients. Monitoring the outcome of a specific therapy and the condition of the patient is of utter importance for effective pain management. A smartphone application was developed for comprehensive pain management in chronic pain patients. It allows patients to enter and keep track of their health related QoL, pain, and medication intake in an easy, fast, and regular daily routine. The app was built based on standardized questionnaires and the needs of pain specialists and patients assessed via a brief survey in a small target group. The survey showed an unmet need for applications in clinical practice, which enable comprehensive monitoring of QoL, pain, therapeutic progress, and thus allow for patient engagement and overall improved pain management. Beneficially, patients can share their reported outcome measures gathered via the app with their pain specialist via PDF exports for remote monitoring. The smartphone app has been realized and is currently evaluated in beta tests. The easy, comprehensible, and routine assessment of the patient’s condition, combined with a future integration into an online therapy management system, may further advance pain management in chronic pain patients.

In this work, an electrode was developed to determine the dielectric properties of tissue phantoms using impedance measurements. For this purpose, a coaxial electrode design was selected as the most suitable electrode design. Three tissue phantoms (blood, fat, muscle) were prepared from distilled water, sodium chloride, 1,2-propanediol, agar, and gelatine, and then characterized. Measurements were performed using an impedance analyzer in a range between 1 kHz and 1 MHz. A comparison with literature values showed good correspondance of the permittivity values determined, thus proving the applicability of the electrode design. However, the conductivity values measured were different from literature values. In the future, alteration and minimization of the electrode can be done to further improve the characterization.

We explore the task of tissue classification for colon cancer histology in a low label regime comparing a semi-supervised and a supervised learning strategy in a series of experiments. Further, we investigate the model robustness w.r.t. distribution shifts in the unlabeled data and domain shifts across different scanners to prove their practicality in a histology context. By utilizing unlabeled data in addition to nl = 1000 labeled tiles per class, we yield a substantial increase in accuracy from 89.9% to 91.4%.

This paper presents a robotic wrist rehabilitation device that facilitates both pronation/supination and flexion/ extension movements. Further, it provides technologyaided assessment capabilities through measurement of joint stiffness, range of motion (ROM) and detection of the catch that is often associated with spasticity. The mechanical and electrical system design, allowing for aforementioned training and assessment capabilities, is presented. An indirect torque measurement via the actuator currents is proposed, which can determine the angle of catch, and compared to the already established methods of using inertial measurement units. The approach is tested using a simplified, mechanical model of the joint catch on the presented device. The results show that the indirect measurement has similar performance of assessing the angle of catch as the inertial data. Therefore, the presented device shows a potential to assess not only stiffness but also the angle of catch without additional sensors.

Treating patients in an intensive care unit (ICU) isolation room is full of challenges for the caregivers and the patients. While this is known for years, the COVID-19 pandemic has highlighted these challenges as if under a burning glass - intensified by a high number of patients at the same time. The risk of infection for the caregivers is very high inside the isolation room and the time spent for putting on and taking off personal protective equipment (PPE) is enormous. Thus, every treatment that can be performed from the anteroom of the isolation room at the same quality level as if being inside the isolation room is a potential building block for solving the challenges. However, data like vital signs have to be provided to the anteroom, alarms have to be transferred to the right caregiver at the right place, and controlling medical device parameters from the anteroom are prerequisites. The new IEEE 11073 Service-oriented Device Connectivity (SDC) series of standards is a key technology to realize this. Therefore, this paper presents an example implementation of IEEE 11073 SDC-based ICU isolation rooms by means of a demonstrator. Such systems can significantly reduce the amount of time the caregivers have to spend in the isolation room and reduce the number of clothing (PPE) changes while providing care at a level of a non-isolation room. As a midterm goal these are steps for the vision of a “silent/quiet ICU”.

In this study, DSC experiments were used to elaborate an annealing protocol for injection molded very high molecular weight (VHMW) PLLA with two molecular masses, 320,000 g/mol and 700,000 g/mol. The initial material obtained from the injection molding process was found to be highly amorphous and not in thermodynamic equilibrium state, as distinct cold crystallization and enthalpy relaxation were observed in DSC data. Thermal annealing to 85 °C for 60 - 90 min under controlled conditions using a DSC device showed to be sufficient to stabilize the polymer and increase the crystallinity to up to 50 %. Annealing did not result in any signs of thermal decomposition. The elaborated thermal treatment has been transferred to manual annealing of large specimen geometries for subsequent DMA analysis.

Poly(N-isopropylacrylamide) (PNIPAM) hydrogels are one of the most commonly studied thermoresponsive biomaterials. In this work, we investigate ball milling as a method of producing PNIPAM nanogels and confirm that the temperature responsiveness is maintained in the nanogel form based on hydrodynamic diameter and zeta potential measurements.

Electrospinning, a standard technique for producing nonwoven fabrics, is widely used in the biomedical field. Due to local electric fields, fiber deposition is influenced by 3D surface structure of metallic collectors. For complex shaped structures, predicting the deposition and orientation of the fibers remains challenging. Investigations on well-defined structures permit a more detailed insight. In this study, metal blocks with round, slit and square holes of various sizes and depths were fabricated and covered with a thin layer of polymer fibers. The fiber arrangement of the layer was investigated by different microscopic methods. Analysis of the results showed less deposition on round holes and well-bridged aligned fibers on gaps which were less than four times wide as deep. Experimental results confirmed the theoretical predictions. Considering these findings, further work can be addressed to specific collector design to attain targeted fiber deposition.

The characterization of mechanical properties of materials used in biomedical applications is essential for performance evaluation. In addition to quasi-static tests, dynamic tests extend the range of methods and allow predictions of failure, as well as information on durability. Appropriate specimen grips according to the test sample geometry are crucial for a reliable examination of mechanical testing and therefore valid experimental data. In particular, the investigation of polymers is challenging, as properties show major differences depending on temperature and applied loading rate. This could result in slipping or tearing of samples in the specimen grip area. Numerical simulations of reference grips, as well as alternative custom designs, were performed evaluating damage due to the clamping process and to provide appropriate specimen grips for future dynamic-mechanical investigations of materials with variable properties. Both the results of the numerical simulation and preliminary tests with 3D-printed prototypes show a distinct improvement in specimen clamping. Plastic deformation and local stress peaks were reduced while maintaining the same tightening torque.

The effect of molecular weight M w on the polymorphous crystallization and melting behavior of poly(Llactide) PLLA (L206, L210 and L214) were systemically studied by differential scanning calorimetry (DSC). Melting and reorganization of conformationally disordered crystals (α′-phase) of PLLA are analyzed with respect to the cooling rate q c in a range between about 4 and 50 K/min and the heating rate q h in a range between about 5 and 200 K/min. It was shown that the α′- to α-crystalline phase transition prior to the dominant melting depends on cooling rate q c and M w . Unlike very high-M w PLLA L214, for low-M w PLLA L206 crystallized at high q c , the α′-form crystals only partially transformed into the α-modification, and certain amounts of α′-form crystals melted directly without α′- to α-transition during the heating process. With increasing q c , the melting of PLLA L210 changed from phase-transition- and -melting mechanism to the common melt-recrystallization mechanism. It was also found that the polymorphism of PLLA significantly affected the crystallization kinetics. Consequently, the reorganization of α′-crystals into stable α- crystals could be inhibited by fast heating (q h > 150 K/min).

In this work we present a four compartment mathematical lung model to simulate the local gas exchange based on electrical impedance tomography (EIT) input. The model simulates the local and global time course of the volume, the gas fraction in the alveoli and the arterial partial pressure. The regional information about the distribution of ventilation and perfusion is extracted from EIT data. In order to illustrate the suitability of this approach, an ARDS patient is simulated. The results show good agreement with measurement data for large portions of the time and seem promising e.g. for the use in medical decision support systems.

This paper presents the proof-of-concept study of an adaptor allowing the combination of a gamma probe with ultrasound (US) imaging, intending to improve the detectability of sentinel lymph nodes (SLNs). The performance of the adaptor in US imaging, in terms of depth of penetration and distance accuracy, and gamma scanning, in terms of sensitivity and spatial resolution, was investigated. We observed that the quality of the US imaging through the adaptor was promising and close to that of normal US imaging. However, the performance of the gamma probe through the adaptor was fairly poor, necessitating the improvement in the design of the adaptor for better gamma scanning. This study shall provide a basis for the development of a handheld gamma-US scanner for interventional procedures and small field-of-view (FOV) imaging in the future.

Introduction: The design and development of a novel dynamic hand orthosis that allows individual and progressive adaptation to the patient’s condition require knowledge about the maximum voluntary contraction forces of the fingers. The following work presents a test setup and procedure for determination of isometric finger contraction forces during flexion depending on various finger positions and finger tasks. Methods: The test setup places hand and fingers in certain positions (position 0: grasping of a plate; position 1-3: circular grasping of cylinders of three different diameters). The force measurement is carried out using pressure sensing: A capacitive force sensor is placed on each fingertip. To perform various finger tasks, the subjects are asked for isometric contraction of defined finger groups and single fingers against the boundary system. For proof of the concept, measurements are performed with six subjects and compared with reference values from the literature. Results: The separated finger force of each finger can be captured during the finger tasks in varying positions. Furthermore, the total finger force of all active fingers can be determined. It is shown that the total finger force during contraction of all fingers (IMRLtask; 78.4N ± 25.8N, d2) is higher than during contraction of the index finger (25.8N ± 3.4N, d2). Furthermore, the single finger force of the index finger is higher than the index finger activity in IMRL-task. The results also show a tendency for the finger contraction force to increase with increasing flexion. Further investigations may require improvements to the method. In this proof of concept, this potential is revealed and discussed. However, the functionality of the test setup and procedure can be demonstrated.

Skin brachytherapy is a successful practice for treating skin cancer patients. Although a gamma source like an Iridium-192 (Ir-192) is commonly used in the skin brachytherapy, the use of a beta source like Yttrium-90 (Y-90) has also been investigated for this purpose. In the current study, we simulated the dose distribution of a proposed Y-90 applicator and an Ir-192 Leipzig applicator in a skin phantom, and compared the performance of each of the methods in treating three tumors in three hypothetical scenarios. Results of this simulation study showed that because of the sharp dose falloff of beta radiation in tissue, the Y-90 applicator may lead to better protection of the bone and/or cartilage under thin tumors. However, it may lead to excessive skin surface dose if thick skin tumors are treated. To choose between the Ir-192 and Y- 90 applicators, one may need to consider the thickness of the tumor and the thickness of the adipose layer between the tumor and the bone.

One of the greatest challenges in the development of novel implants is to find materials that optimally interact with the tissue and support the function of the implant over a long period of time. Especially considering the fact that biomaterials are used in multiple applications in modern implants, a better understanding of the physiology of the tissue in contact with the material becomes more and more important. To better understand the cellular responses to biomaterials, it is furthermore necessary that the models used reflect the physiological conditions at the implant site. However, in most cases, immortalized cell lines instead of tissue-specific primary cells are used to study the biocompatibility of implant materials. When interpreting these results, it is important to consider that cell lines differ significantly from primary cells found in their native environment. In this study, we examined the morphological differences between a cell line and primary cells upon contact with various implant structures resembling implant prototypes. Using a mouse fibroblast cell line in comparison with primary porcine fibroblasts, we discuss the results with respect to the use of cell lines for biocompatibility testing of biomaterials. Our results suggest that to better estimate meaningful results regarding the compatibility of implant structures, key control experiments should always be performed with specific primary cells.

Injuries and the associated surgery to the anterior cruciate ligament (ACL) can often trigger unpredictable effects, such as the so-called giving way effect, which is an uncontrolled buckling of the knee joint. For this purpose, the KneTex project has developed a smart textile-integrated sensor and actuator bandage system to record the movement of such patients and to monitor and support the rehabilitation process. Long-term monitoring and analysis of the movement data will identify patterns or gait types that can lead to a giving way effect. This paper describes the recent developments of the random forest model-based motion classification system developed within the project. Improvements have been achieved by reducing the number of features needed by 25% using feature importance analysis, speeding up the computation time by 14%, and increasing the classification efficiency. Feature elimination is a useful tool to improve classification systems in settings where feature count is high and feature importance analysis contributed by improving our understanding which sensor of our system are important for the motion classification task.

Temporal-bone (TB) experiments are useful for the investigation of middle-ear and inner-ear transfer properties, and the analysis and development of active or passive middleear prostheses or related devices. To acquire the mechanical middle-ear (ME) response, the Laser-Doppler vibrometry (LDV) is utilized as a preferred method. In most setups, the exact measurement point chosen on the biological structure of interest as well as the direction of the corresponding vibration measurement is not precisely defined and of unknown reproducibility. Hence, an automated setup controlling coordinates of TB and measurement direction is developed and tested. To test the reproducibility, TBs were measured with ER7C and ER10C probe microphones employing a small industrial robot to position the TB to the ME points of interest and align it to the laser beam. The position and vibration data acquired at the same locations at different points of time were processed and compared. Measured transfer functions indicate noticeable differences at some frequency ranges. Nonetheless, the reproducibility of positioning to repeatedly measured points proved itself relatively accurate, if no manipulations in the setup are made. In conclusion, the current study shows that automated LDV TB measurement system is in principle capable to adapt to the changes in setup during intermediate preparation steps. With more acquired data, the setup stability and reproducibility can be more precisely verified, and additional improvements can be made.

This paper introduces a pressure sensor catheter with two different measurement principles. A Fabry-Perot- Interferometer at the fiber tip is formed by a reflective multilayer membrane of a pressure sensor chip. The cavity provides absolute pressure sensing. 20 millimeter before the fiber tip a Fiber Bragg Grating is integrated for additional temperature sensing. The sensing components are completely coated with a silicone mantle of 2 mm diameter. The signal evaluation is both done with a Fiber Bragg Grating interrogation device. First, mechanical structure aspects of the manufactured fiber-optic hybrid sensor as well as the sensing principle is described. Then, experimental results, including quantification and separation of pressure and temperature sensitivity are presented. Out of these characteristics the possibility for absolute pressure sensing with temperature compensation for the application as a pressure sensor catheter is confirmed.

The retinal disease Diabetic retinopathy (DR) is one of the most probable causes of blindness. Automatic detection of DR is mostly done using convolutional neural networks (CNNs) on colour retinal images. This work in contrast uses stochastic variational deep kernel learning (SVDKL) for DR grading, combining a deep CNN with Gaussian processes (GPs) into a single end-to-end trainable model, which promises to provide predictions with a reliable uncertainty estimate exploiting approximate Bayesian inference. Evaluating the performance and uncertainty calibration of SVDKL on DR grading compared to a plain CNN, the EfficientNet-B0, preliminary results on a subset of the Kaggle DR dataset show a naturally enhanced uncertainty calibration for SVDKL over the plain CNN as well as a good diagnostic performance. Despite SVDKL achieving a slightly reduced accuracy, incorrect predictions were in closer proximity to the target stages, which is beneficial for clinical diagnosis due to minimizing the cost related to severe misclassifications.

Stationary devices for continuous blood pressure (BP) with user-friendly finger sensors are getting common in various clinical fields and new wearable BP-monitors are about to enter the medical world. Apart from the commonly used parameters systolic and diastolic BP, beat-to-beat as well as pulse shape information of these devices is not easily accessible. This article is a review of existing pattern recognition analysis and an outlook to new biomarkers giving a deep insight into the cardiovascular system and its control.

The teaching of medicine is in a state of flux. New technologies enable students to learn faster and more efficiently than ever before. Especially the possibilities that augmented reality brings with it can be applied very well in this field. The possibility to project additional information directly into the user’s field of vision allows him to perceive additional knowledge without being distracted from the actual work. In this paper, a possibility of training Mamma Palpation is presented, which teaches students the course of action and the detection of possible tumors through virtual points. Through guideline-based interviews, added value for teaching could be highlighted.

The current information era provides the population with new innovative technologies, including extended reality (XR). These innovations enhance their scope with new opportunities. While extended reality found its way into medical teaching, its application is limmited. Causal is the lack of haptic feedback, which limits the immersive experience. While simulation devices exist, haptic gloves are oftentimes only applicable for enterprise use and not available for everyone, because of their high prices. Therefore Lucas VRTech developed the LucidGloves to allow a widespread use of cheap haptic gloves. In this paper, a suitability test is performed to determine whether the Prototype 4 of the LucidGloves is suitable for use in medical teaching, based on expert interviews with four physicians.

Pain assessment in clinical settings can be a challenging task for clinic staff and patients. An automatic and contactless pain estimation system via facial expression is a decisive advantage. In the following article we present a method for pain classification based on mimic descriptors. To automatically process facial expressions from video data we applied the Facial Action Coding System (FACS). First, we employed a logistic regression model to predict pain using a publicly available database. Second, we utilized this procedure to predict pain in individual intraprocedural patient data. We reached a mean accuracy of 80 . 94% for the publicly available data and 73 . 14% for our individual patient data by classifying data as pain and non-pain.

Photoplethysmography signal (PPG) provides information about the cardiovascular, respiratory, or nervous system activity. Recent studies proposed methods of apnea identification from PPG using amplitude or baseline modulation, or area under the curve. We hypothesized that analysis of the dicrotic notch could improve the sensitivity and specificity of apnea detection. We applied four methods to detect apnea from PPG records experimentally obtained from 10 volunteers, where each volunteer simulated three apneic pauses. We adopted the Area under the PPG and Pulse wave amplitude parameters, and we proposed the Dicrotic notch area parameter. For the methods, we determined the delay in apnea detection and the sensitivity and specificity of apnea detection. The average delay in detecting an apneic pause from the PPG using the Dicrotic notch area was 6.1 s. Combining the parameters of the Area under the PPG with the Dicrotic notch area increased the specificity of apnea detection from 73.8% to 80.0%, although sensitivity decreased from 91.1% to 85.6%. The newly proposed parameter Dicrotic notch area showed its potential for automatic apnea detection.

Introduction: Children with cerebral palsy (CCP) benefit from intensive arm training. Exergames that can be played at home offer the possibility to increase the frequency of therapy but require reliable and accurate real-time motion tracking via easy-to-use sensors in unsupervised settings and magnetically disturbed environments. Method: We propose an inertial-sensor-based method with a single sensor on the wrist for real-time tracking of the inclination of the forearm. The control parameter of the game was validated with an optical marker-based ground truth system. Results: First experiments with a therapist performing training movements in a healthy and simulated spastic manner show that the forearm inclination well captures the motion dynamics. The accuracy of the inertial-sensor-based measurement is validated with respect to the reference system in three healthy subjects. Orientation offsets between the inertial sensor and the forearm marker set in the range of 2° ̶ 6° and dynamic measurement errors about 3.1° were obtained. Conclusion: This work demonstrates that the proposed method is suitable for real-time control of exergames of CCP. The validation with an optical reference system showed that the forearm inclination can be used as for feedback and therapeutic progress monitoring.

Encapsulation is essential for mechanically flexible and electrically active implants as it protects them from the harsh environment inside the body. Since the performance and longevity of implants depends on the quality of encapsulation, research continues into new and better encapsulation strategies. Chemical vapour deposition can be used to deposit Parylene-C, a biocompatible polymer, at room temperature. Plasma enhanced atomic layer deposition can be used to deposit ultra-thin, conformal, and hermetic metal oxide layers. These strategies are both used for the encapsulation of implants. In this work, these two methods are combined to create a polymer/metal oxide hybrid system in which a layer system of metal oxides is sandwiched between two Parylene- C layers. DC leakage-current measurement and elevated temperature are used to qualify the barrier property of the layer-system. After more than one year, the barrier properties still show no significant decrease.

Contactless blood pressure monitoring displays a chance for detecting hypertension and increasing the awareness of cardiovascular diseases. Camera-based methods show high potential for replacing inconvenient cuff-based technology. Blood pressure estimation through pulse transit time (PTT) and other parameters obtained from photoplethysmography (PPG) signal present a promising alternative to conventional blood pressure measurement. In this paper, recent approaches for contactless blood pressure monitoring were reviewed. All of the articles implemented systems based on PPG, whereas most of them used PTT as measuring parameter. The results are reviewed and summarized while pinpointing key challenges and discussing expectations for future research.

Electrospun fibers based on synthetic and natural polymers represent a promising approach in order to mimic the mechanical stability and the required cell-polymer interaction of the natural extracellular matrix (ECM), . In this biomedical context, the biodegradable medical grade poly-llactide (PLLA) and the natural gelatin are already well established. However, to obtain a stabilized structured PLLA/gelatin blend, crosslinking of gelatin is required. In this study, the influence of the post processed crosslinking via Glutaraldehyde (GTA) vapor and GTA solution on the morphological and physicochemical properties of the blend, as well as on the in vitro release of previously incorporated dexamethasone (DMS), was investigated. The addition of gelatin decreased the fiber diameter and increased the surface hydrophilicity, which is furthermore unaffected by the crosslinking process. The in vitro release of DMS was decisively changed by both, the addition of gelatin and the subsequent crosslinking processes. For all modified samples, a very rapid burst release of DMS was detected within the first 24 h, which is crucial for biomedical applications such as tumor therapy.

Patients with cardiac arrhythmias are currently treated with conventional transvenous rhythm implants. Complications are frequently associated with intracardiac implanted leads primary insulated by silicone and polyurethane. However, experiences show that polyurethanes in particular are susceptible to various degradation mechanisms including hydrolysis, environmental stress cracking (ESC) and metal ion induced oxidation (MIO). In vivo, pacemaker leads are exposed to a complex thermal, chemical, mechanical and biological loading. The current study focusses on in vitro analyses to assess the biostability of Pellethane 2363 55D and the silicone MED 4765 as cardiac pacemaker lead insulations. Degradation processes were simulated in vitro in a load-free state and under static mechanical loading by subjecting a coaxial lead design to bending radii from 3 mm to 19 mm. Physiological environmental conditions were mimicked using a physiological saline solution and a tempered oxidative solution. Surface morphological and thermal analyses were performed before and after in vitro testing by means of scanning electron microscopy (SEM) and differential scanning calorimetry. Melting temperatures of silicones around 40°C were measured, before and after in vitro testing, respectively. Pellethane insulation layers had two endothermal melting regions at 100°C and 170°C before and a third melting region at 45°C after in vitro testing. The additional melting peak may indicate a change of thermal material properties due to degradation. SEM images showed degradation phenomena similar to in vivo studies, varying in severity and depending on the bending radius. Thus, the relevance of mechanical loading for in vitro replication of clinically relevant lead insulation degradation was demonstrated.

Active magnetic motion sensing relies on a combination of magnetic sensors and actuators to indirectly sense a time-varying change in position and orientation. Based on our previous work on a 1D setup with magnetoelectric (ME) sensors, we introduce a concept for a full 6D sensing system, which is positioned as a modular setup for human movement pattern capture with regard to neurodegenerative diseases. In this paper, we implemented an inverse model based on ideal triaxial sensors and coils. We then included results from the previous experiments to estimate the prospective performance in a physical setup. We obtained a spatial error of a few cm and approx. 10° with focus on superimposed sinusoidal movements of 1, 5, and 10 Hz.

Nitinol became one of the most prominent materials for biomedical devices. The unique properties such as the shape memory effect and the superelasticity enabled new applications, particularly in the field of stent technologies. Due to their thermosensitivity, manufacturing processes are required that control the heat accumulation in the material. Here we investigate the femtosecond laser ablation of Nitinol supported by a water flow inside the processed tube. We show that laser absorption in the water prevents annealing of the Nitinol, even for high pulse energies. Furthermore, we investigate the groove profiles by scanning electron microscopy and observe minor improvements in the processed structure when using water flow inside the tube. The results show that water-supported femtosecond laser ablation can be used to optimize the fabrication of medical devices such as cardiovascular stents and to reduce processing times.

A passive cavitation detection setup is used to assess and verify the capability of frequency- and amplitudemodulated ultrasound burst signals on the generation of inertial cavitation. The measurements were conducted with a flowthrough tissue-mimicking phantom with a canal of dc = 1mm in diameter simulating a fine blood vessel. By means of a flow velocity of v f = 50mm/s circulating blood in the cardiovascular system is imitated. Applying signal frequencies from f = 550 − 950 kHz and peak rarefaction pressures in a range from ˆpPRFP ≈ 0.17 − 1.81 MPa, various ultrasound stimuli with alternating signal parameters were investigated in the absence and presence of a talcum-water mixture acting as cavitation nuclei. By evaluating the broadband emissions in the frequency domain and calculating the voltage spectral density Srm, the results can be compared. The findings of the present study demonstrate that each ultrasound stimulus triggers a similar strength of cavitation noise at the same Mechanical Index, representing a certain pressure. Over the whole frequency and Mechanical Index (pressure) range, each ultrasound signal type generates a voltage spectral density from S rm = 0.25 − 2.25 V/ √ Hz.

In surgical refractive interventions over-correction or under-correction still occur in 10-15 % of cases. Improved outcomes could be achieved by incorporating patient-specific corneal biomechanics into surgical planning. As a first step, this study mechanically characterizes and numerically simulates corneal lenticules harvested during a refractive intervention called Corneal Lenticule Extraction for Advanced Refractive Correction (CLEAR). These human corneal lenticules were mechanically tested in uniaxial extension in nasal-temporal and 45 deg orientations. An HGO material model that accounts for the unique collagen arrangement of the human cornea was developed and used to identify the material parameters using a finite element model of the experimental setup. The material parameters that best represent the experimental force/displacement were obtained by parameter estimation using Bayesian optimization. The uniaxial experiments performed on 3 patients did not show distinctive differences in orientation. The estimated material parameters fit well with the experimental data, representing a promising material model. In the future, Brillouin measurements will be performed on the CLEAR patients before lenticule extraction so that material parameters can be estimated in vivo, providing the surgeon with a patient specific-surgical planning tool.

The use of high-frequency (HF) generators for HF surgical treatment of biological tissue has become an indispensable part of today's surgical applications of all kinds. Generally, HF alternating electric currents between 300 kHz and 1 MHz are used to induce hemostasis by heating at the cellular level. This effect can be attributed to Joule heating, in which a current dissipated conductor converts the electrical energy into thermal energy. However, sound evidence on the reliability and effectiveness of application-specific HF generator modes is not sufficiently available. Usually, the evidence takes place empirically by means of preclinical or clinical studies. Nevertheless, a corresponding empirical data collection is time- and cost-intensive. Therefore, physiological and statistical modeling provides the opportunity to relate tissue response to the applied electrical energy to obtain a prediction of the tissue reaction. In this contribution, we establish a monopolar coagulation model of an already well-known model approach based on Pennes bioheat equation. Additionally, the vaporization of tissue water at the water boiling point is considered. Furthermore, a variation of temperature-dependent tissue parameters was performed to analyze their impact on the model output. The simulation results demonstrate that the initial electrical conductivity has the greatest influence on the temperature distribution as well as on the time until the tissue temperature reached the boiling point of water. In contrast, the tissue water content has the least impact on the model output. Depending on the desired coagulation effect, HF power control as a function of electrical conductivity or its reciprocal, tissue resistance, must be added next in an improved model.

To simulate the specific material properties of thermoplastic polymers a suitable constitutive model is essential. The parallel rheological framework (PRF) model was calibrated and evaluated in this study as potential constitutive model for polymer stent application. Tensile as well as recovery tests with different loading rates were performed using PLLA specimens. In order to calibrate the constitutive model, the conducted material tests were simulated accordingly. The parameters of the model were iteratively varied to obtain good accordance of the simulation with the material tests. In contrast to elastic plastic material models, viscoelastic material behavior can be represented with the nonlinear viscoelastic-plastic PRF model. The generated and possibly further refined model can be used for the simulation of polymer stents.

High-frequency (HF) surgery has been established in medicine for decades and is very popular in transurethral (minimally invasive) surgery. However, the technical settings of generators are almost exclusively based on empirical data. So far, making a well-founded statement about a possible impact of technical parameters in HF surgery is nearly impossible. Therefore, the aim of this research was to systematically review the published data for transurethral resection of bladder tumor (TURBT) to find out, whether the publications consider both sufficiently, technical parameters and physical parameters. It is shown that although technical parameters are reported, they have been insufficiently considered in analyses so far. Further analyses must be performed.

Craniofacial deformities such as positional skull deformities are widespread in infants. The early detection of these deformities is crucial for the effective treatment and the associated minimization or prevention of visual and pathological abnormalities. This paper presents a solution for the early detection of craniofacial deformities using easy available materials in both the domestic and clinical environments. For this purpose, suitable web technologies as well as webserver systems are used in the context of app and server development. Users can use the app to take a bird's eye view of the child's head and use a 50ct coin as scale to then automatically obtain calculated clinical parameters which they can compare with clinical standard values to rule in or out a skull deformity. The concept represents a way to compensate for the lack of specialists in craniofacial surgery as well as knowledge gaps and to digitize care concepts in the healthcare system.

Knee arthroplasty follow-up involves only a limited number of X-ray-based implant condition assessments at extended intervals. Polyethylene (PE) wear of knee endoprosthesis often progresses unnoticed until symptomatic adverse effects become apparent or implant failure occurs. A PE wear measuring concept, consisting of an implanted passive sensor and an extracorporeal readout unit, was previously presented by our group. This system is intended to enable patients to regularly check their knee implants’ condition at home. In this context, an appropriate product design for the readout unit is essential to enable the user to record correct measurements. An experimental investigation of the optimized positioning of the system's readout unit was conducted and the measurement data were analyzed. The findings were then implemented in a prototype for readout unit attachment. A sleeve with Velcro and a sewn-in readout unit has been fabricated. A cutout for the patella ensures intuitive correct alignment of the measurement system according to the experimental results. A hardware housing incorporates the electronics.

Filter membranes (pore size 0.2 μm) are frequently used to perform a sterile filtration. To characterize the pore size of filter membranes, a method to determine the particle retention of nanoparticles was defined. The Dynamic Light Scattering method was first analysed and calibrated to correlate the particle concentration with absolute intensity of scattered light. Afterwards, standard particle suspensions with particle sizes of 0.150 μm and 0.239 μm were lead through an infusion filter and the particle retention of the filter was evaluated by the determination of the original and the filtrated particle concentration. The test filter retained 91% of the smaller and 98% of the larger particles. Evaluating the particle retention with Dynamic Light Scattering could be a time saving method with lesser effort compared to existing methods.

Besides visual assessment of the intra-vascular appearance, intravascular optical coherence tomography (IVOCT) enables precise diameter measurements within the coronary arteries. The current study presents an upgraded test setup allowing the investigation of the elastic properties of porcine coronary arteries ex vivo under pulsatile flow conditions. IV-OCT imaging was performed within the left circumflex artery of a porcine heart applying a normotensive and a hypertensive pressure regime at different pulse rates. Radial compliance was derived from the luminal diameter as well as intra-arterial pressure measurements. Test results show a significant reduction of radial compliance under pulsatile conditions, when compared to reference measurements under steady flow. The mean radial compliance decreased with increasing pulse rate, which can be attributed to the viscoelastic properties of the arterial wall.

This paper deals with the analysis of the relationship between fetal hemoglobin (HbF) and adult hemoglobin (HbA) in preterm infants as a function of the gestational week and newborn weight based on the available literature. The role of other parameters on the shift and shape of the OHDC curve is also discussed. The function could be useful when dynamic development of OHDC in a neonate occurs in emergencies such as blood transfusions. The main idea is to incorporate the dynamic evolution of the hemoglobin conversion as a mathematical function into existing models of oxygenation in preterm neonates or into automatic feedback control algorithms of oxygenation. This could help further individualize and optimize the oxygenation settings of preterm neonates.

Convolutional neural networks (CNNs) have proved to be successful in many applications such as image processing. However, even imperceptible perturbations applied to the images can make the neural network performance unreliable. To guarantee an accurate performance in safety critical fields, it is necessary to assess the robustness of CNN solutions before launching. Adversarial attack is a machine learning approach to generate perturbations on real samples to detect the vulnerability of CNN. In this paper, we will use an adversarial attack technique to evaluate a CNN at different training states. The model was trained to perform surgical tool classification task, which was applied to recognize surgical tool in Cholecystectomy to further analyze surgical process. The experiments demonstrate the relation between training states and robustness, i.e. the robustness improved at higher training states, especially for some particular classes. In future work, additional training with generated adversarial images may improve the robustness of the model.

Antibody-mediated rejection after kidney transplantation is a major limitation for long-term survival of donated kidneys within the recipient allograft. With the existing HLA cross-matching test only complement activating donor specific antibodies are evaluated. Other measurements are more sensitive, but cannot assert the clinical relevance of the donor specific antibodies. Thus, novel methods that address the more complex nature of cellular interactions with serum components are needed to improve antibody-mediated rejection prediction. Recent studies suggest, that the interaction of natural killer cells of the recipient with endothelial cells of the donor is of major interest. Here we present a microfluidic cell culture device and a corresponding assay to test for the relevance of natural killer cell activated antibody-mediated rejection in vitro. Therefore, a protocol to co-cultivate both cell types and track natural killer cell activated killing within a fluorescence-based assay was established. To automate this approach, the prototype of a well plate sized microfluidic device was developed, that enables staining of the two cell types and repeatable and timedependent co-cultivation of the cells with serum. This automated approach enables further evaluation of the clinical relevance of the established assay paving the way to additional information on HLA cross-matching that helps to quickly estimate the probability of Antibody-mediated rejection and the clinical relevance of donor specific antibodies.

This paper investigates the extent to which the IEEE 11073 SDC standard family loads a communication network. An SDC provider/consumer based on the SDCLib/J from SurgiTAIX AG was implemented as a test environment and several different test scenarios were run in order to gather data sets for different use cases. The analysis of the thus acquired measurements shows a linear, predictable growth in network utilization for almost all combinations of active metrics and value update rates. In the case of very high metric update rates, a stagnation of network utilization can be observed. This behaviour can be explained by the use’s implementation and an unfavourable test setup. Overall, the network load generated by SDC is justifiable.

The increasing use of mobile digital technologies in our daily life is also affecting the healthcare sector. Digital health applications have the potential to improve disease monitoring and management, avoid hospitalizations, and more. A main task for the health care system is to create a framework that facilitates the identification of safe and effective apps for health care professionals and patients to generate the most health benefit as well as guide payer coverage decisions. Currently some European countries have already developed a national framework for market access approval, nevertheless only Belgium and Germany have additionally created a reimbursement system. The aim of this investigation is to compare the two different approaches of Belgium and Germany, based on the framework itself, the access requirements, and the experience so far. The overall approach seems to be similar, but scrutinizing the details, there are several significant differences. The evaluation procedure in Belgium consists of a 3-level validation pyramid while in Germany there is a normal procedure as well as a DiGA-Fast- Track. Only applications in stage 3 are reimbursable in Belgium whereas in Germany applications from provisional and permanent listing are reimbursable. The comparative overview shows that the two forerunners, Belgium, and Germany, use similar but not identical approaches, with partly relevant differences such as the evaluation procedure reimbursement and the formulation of the target group. The current situation (status 11.04.22) shows that the focus of the app indications differs between both countries. Furthermore, not a single app is listed in Belgium and Germany at the same time. From our perspective it should be a goal for manufacturers of CE marked medical devices to achieve reimbursement in as many member states as possible.

The implantation of a transcatheter aortic valve replacement (TAVR) in patients with severe aortic stenosis improves pathologic blood flow through the aortic valve, but still alters hemodynamics in comparison to a healthy native aortic valve. The hemodynamic characteristics of TAVR are associated with hypo-attenuating leaflet thickening and prosthetic leaflet thrombosis, which may reduce the durability of TAVR. For this reason we developed a numerical model to identify pro-thrombotic regions of TAVR based on the velocity field. In silico models have already been proven to be an effective tool for device optimization in other medical device applications. However, when using an in silico model, it is important to validate the model assumptions with experimental data or analytical solutions to confirm the accuracy of the model. In this study, we present an approach to validate a complex numerical Fluid Structure Interaction (FSI) model used to simulate leaflet kinematics and flow in the vicinity of a TAVR. Based on the recommendation of ISO 5840 (2021), a combined validation of the kinematic and fluid dynamic parameters was performed for this purpose. The leaflet kinematic was investigated via high-speed recordings and evaluated based on the Geometric Orifice Area (GOA). The velocity field of the TAVR was determined experimentally using Particle Image Velocimetry (PIV). The evaluation of the GOA showed a good agreement of the in silico model with the in vitro data for the systolic duration. The occurring velocities were qualitatively compared during peak systole for three planes and showed similar flow characteristics. The maximum velocity was 1.1 m/s in the in vitro and in silico model. Based on the results, it can be assumed that the numerical FSI model of the TAVR can be used for thrombosis risk assessment.

The positive end-expiratory pressure (PEEP) valve is important in the regulation of the pressure at the patient airways in devices for mechanical ventilation. The goal of this work is to derive a model that sufficiently captures the dynamics of the PEEP valve to be used for model-based control design. This paper describes the development of a model for common PEEP valve designs, thus allowing an adoption for different devices. A simplified first principle model with one mechanical degree of freedom is derived, under the assumption that the valve’s membrane behaves similar to a massspring- damper system. Following a grey-box model approach, physical parameters were lumped together and experimentally determined or identified from acquired data. Therefore, a valve test bench was developed to record various series of measurements. It was possible to determine a suitable flow factor which describes the correlation between flow, valve opening and upstream pressure. Additionally, a damping coefficient was estimated, with which the model achieves a good fit to the measured data at low dynamics and flows.

Computational fluid dynamics (CFD) has enormous potential for the development of novel cardiovascular implants. In particular, the evaluation of the thrombogenic potential of implants is gaining importance. For this purpose, the numerical model must represent the essential properties of the real system, i.e., an implant in the human blood circulation, in a suitable form. Due to its complexity, the model building process is crucial for the successful application of the model. Currently, there is no structured guideline for the application of numerical flow simulation to which model developers can refer. Based on NASA and ASTM, we present key aspects of the model building process using CFD for stent flow as an example.

Artificial Intelligence has been blending into daily life by means of many useful applications from voice command to facial recognition. One therapeutic application to be supported by AI solutions is treatment of people with Autism Spectrum Disorder. A closed loop feedback system is planned in conjunction with a novel reward system that will encourage the user to express emotions and be rewarded for it in a virtual environment. In this work five popular neural network architectures of VGG16, ResNet50, GoogleNet, ShuffleNet and EfficientNetb0 are studied and compared, with the aim of finding a relation between accuracy and developed features based on the architecture, for the application in Facial Emotion Recognition (FER). Three datasets were used, the OULU-CASIA for training and validation, alongside FACES and JAFFE for robustness analysis. The images were first preprocessed to eliminate background noise. The performance of the model was based on the true positive predictions with Grad-CAM prediction visualizations to visualize the focus of the networks in making decisions for classification. Results showed that deep network architectures with high parameter space performed best, with architecture design showing more influence on the region of focus than on classification results. This is attributed to the different layer combinations as well as parameters used for feature extraction. Shallow depth networks with high parameter space performed better than deep networks with low parameter space for FER application.

Introduction: We apply deep learning to emulate Finite Element simulations by exploiting neural networks as universal function approximators. A novel processing pipeline automates the incorporation of three-dimensional biomechanical simulation data into an image-based deep learning task where spatially resolved deformations are predicted using pixelwise regression. We show that a model trained in this way can quickly reproduce simulation results with reasonable preliminary performance. Methods: We trained a customized U-Net (32 input features) to predict displacement models based on the input force vectors. The pipeline was applied to a basic indentation process for an idealized reference geometry (puck) and a 3D liver model from the LiTS dataset. The generated 3D simulation data is normalized and projected onto 2D images using the midpoint circle algorithm such that x, y, and z coordinates are mapped to pixel values of RGB image channels. This way, 1500 simulations of the reference geometry and 500 simulations of the liver deformation with force-controlled, randomly initialized load distributions were performed. Simulations took about 2-5 minutes of computation time with FE models of 6000 and 10000 elements. The neural network was trained on the simulation database using pixelwise regression. Results: We successfully trained a model on the pixelwise regression of simulation results. The model optimization reached an f1- score of 0.69 and 0.61 for the puck and the liver. The displacement images were reprojected and compared to the 3D ground truth. The trained models reached an average surface distance of 3.1 ± 1.2 and 4.2 ± 2.2, as well as 4.3 ± 0.4 and 6.4 ± 0.2 for the Hausdorff distances. Conclusion: The model performance is still impractical for clinical decisions. Still, with mean computation times of under 250 ms, the approach's potential with a more extensive database and a more sophisticated deep learning architecture, e.g., GNN, is recognizable.

Radiopharmaceuticals can be used for targetspecific functional diagnostics, such as PET or SPECT imaging, or radionuclide therapy of diseased tissue, depending on the incorporated radionuclide. Following initial in vitro testing, radiopharmaceutical candidates are usually further characterized in small animals. Since reduction, replacement and refinement (3R) of animal testing is a central precept in preclinical research it would be beneficial to replace at least some of these tests by alternative methods. Using micro physiological system technology, various organ-on-chip models can be created with human cell systems/organoids, which are operated in a circulatory system under defined physiological conditions. Here we present first attempts to introduce micro physiological systems for evaluating radiopharmaceuticals using the radiolabeled anti-EGFR antibody cetuximab as reference compound. In a micro physiological system equipped with six 96-well plate-like microwells in a flow chamber, binding of 64Cu and 68Ga-labeled cetuximab to cells and spheroids grown from A431 (EGFR-positive) and MDA-MB435S (EGFR-negative) cells was measured and compared to conventional microplates. Specific saturation binding of radiolabeled cetuximab at increasing concentrations was analyzed using a phosphor imaging system. The affinity of radiolabeled cetuximab towards A431 spheroids measured in the micro physiological system was in the same range as that of the spheroids in conventional microplates. Within the assays in micro physiological systems, the results showed a trend towards increased affinity for A431 monolayers compared to the spheroids. The values of binding capacity for radiolabeled cetuximab on 2D and 3D A431 cell culture models were in the same order of magnitude when measured in micro physiological systems or in microplates. Building on these first promising results, the work will continue on MPS modules containing advanced human spheroid/ organoid models.

In this work, an attempt has been made to analyze the influence of pathological changes in eye globe dimensions towards the mechanical responses of optic nerve head tissues during eye adductions. For this study, a 3D baseline model geometry of posterior ocular tissues has been constructed. The eye globe diameters of the model are modified to mimic the changes in ocular globe morphology in control, glaucoma, myopia and glaucoma with myopia conditions. Adductions are simulated for each modified model as the rotation of the globe from 1o to 10o in steps of 1o. von Mises strain in lamina cribrosa (LC) and posterior displacement of LC are estimated. Results show that strains developed in LC and its posterior displacement are higher in diseased eyes compared to healthy eyes. It appears that eyes with higher axial length and globe anisotropy are more susceptible to optic nerve head damage. This study might be extended to assess the progression of glaucomatous optic neuropathy.

In previous work, we demonstrated the potential of blood flow sounds for biometric authentication acquired by a custom-built auscultation device. For this purpose, we calculated the frequency spectrum for each cardiac cycle represented within the measurements based on continuous wavelet transform. The resulting spectral images were used to train a convolutional neural network based on measurements from seven users. In this work, we investigate which areas of those images are relevant for the network to correctly identify a user. Since they describe the frequencies’ energy within a cardiac cycle, this information can be used to gain knowledge on biometric properties within the signal. Therefore, we calculate the saliency maps for each input image and investigate their mean for each user, opening perspectives for further investigation of the spectral information that was found to be potentially relevant.

Cerebrovascular diseases like atherosclerosis pose a great threat to health and wellbeing of people worldwide. To enable early diagnosis and treatment of a gradually progressing occlusion of the carotid arteries, we propose the BODYTUNE system. Consisting of a custom-built auscultation device and a mobile application, it aims at enabling the monitoring of the blood flow within the carotid arteries on a regular basis at home. In this work, we present the results of a survey with 65 participants from the system target group to investigate aspects like technical affinity and experience with comparable systems. These results provide the basis on how the BODYTUNE system should be designed to be user-centered, concerns.

Image-based surgical tool localisation and detection play an important role in developing intelligent systems for operating theatres of the future. In literature, proposed approaches require large amounts of data that are fully annotated with tool positions in the image. In this paper, a deep learning framework, trained on binary tool presence, was evaluated for surgical tool localisation in laparoscopic images. Gradient class activation maps (Grad-CAMs) were extracted from an attention-CNN model. The Grad-CAMs were then processed to generate bounding boxes over the surgical tools. Experimental results showed better performance of the attention-CNN compared to the base CNN model with mean tool localisation precision of 72.4% and 28.3%, respectively. These results show the potential of using attention modules to improve tool localisation in laparoscopic images.

This paper presents an exoskeleton for upper extremity rehabilitation of stroke patients. In order to make rehabilitation widely available, the goal is an inexpensive, portable, and easy-to-use therapy for paralysis of the arm. The exoskeleton is actuated by twisted string actuators (TSA). To reduce the length of the TSA a motor with hollow shaft was chosen, as this allows the string to run through the motor. The control of the exoskeleton is designed for predefined elbow flexion exercises. The start of the movement is triggered by a threshold value in the sEMG of the m. biceps brachii. To evaluate the controller, the exoskeleton performed a flexion movement (5° to 115°) in 2 s and then an extension back to 5° in the same time after a 1 s wait. To simulate different force requirements, additional weights up to 1500 g were attached. During both, the flexion and extension movement, that angle tracking error was sufficiently small and the steady state error was zero. The angular velocity of the elbow joint showed pronounced oscillations with increasing weights. The results show that TSAs can be used to actuate elbow exoskeletons. Thus lightweight, ergonomic, simple and small rehabilitation devices can now be built on this basis. The elasticity of the system must be considered in future work. sEMG must be further implemented to control more sophisticated functions of the rehabilitation device.

Polycarbonate-urethane-co-silicone (PCU-Sil) are gaining increasing interest in biomedical application, such as drug delivery systems, coatings and combination products, due to their adjustable mechanical properties. They can be easily processed using a wide range of methods. Biocompatibility can be improved by surface modifications, such as coupling drugs, proteins, antibodies or polymer layers, which extends the range of biomedical applications. To generate additional coupling, the first step would be the generation of functional groups on the polymer surface using surface activations, such as plasma activation. In this context, the assignment of the functional groups is important, which can be done by XPS measurements. Within this study the analysis of the generated functional groups after oxygen (O2) plasma activation of two PCU-Sils was performed. In addition, the stability of the generated surface modification was evaluated after 24 hours. It was surprising that the generated groups were differently stable and also different groups were generated or rearranged over time, which would not have been recognizable if only contact angle measurements were used.

Magnetic Drug Targeting is a new approach in cancer therapy, where magnetic nanoparticles are used as carriers for cancer drugs. Commonly, external magnets are employed for steering the particles inside the blood vessels towards a desired direction. However, an unwanted side effect of the steering is the accumulation of the particles underneath the steering magnet. Many researchers address the number of accumulated particles, but, to the best of the authors’ knowledge, the retroaction of the accumulation profile on the generated magnetic field and, therefore, on the magnetic steering force, has not been investigated so far. Thus, in the proposed study, the influence of the accumulation profile on the magnetic force was numerically investigated. Therefore, a 2D model of a blood vessel with particles assumed as an accumulation profile and a nearby magnet was examined. Moreover, the length, thickness, and effective susceptibility of the approximated accumulation profile and the magnet size were varied. The results reveal that the field distribution is significantly affected, especially for high effective susceptibilities. The initially applied profile amplified the magnetic force; however, when the profile accumulated, it reduced the force up to 50 %. Overall, the results reveal that the retroaction of the particle distribution on the magnetic field must be considered in a simulation model.

Lyapunov exponent is a promising parameter to ascertain the stability of the human gait. In this work, we use a time-series model based on a second-order delay-system with inertial measurement units placed on the foot and wrist. Stability is analyzed in a localized sense, with the Lyapunov exponent computed in the temporal region between two heel-strike points, which are determined using a peak-detection algorithm. We have attempted to show correlations between variations in the stride time and stability of the gait under normal and abnormal conditions. In the latter case, we attach a weight on foot to emulate weakness. On comparison between both cases, we observe a statistical significance of p=0.0039 using Wilcoxon’s rank-sum test. Moreover, on observing the correlations between Lyapunov Exponent and Stride Time Variability, we notice a left-shift in the abnormal case, indicating a lower threshold for instability, with the Stride Time Variability being 0.07 as compared to 0.11 in the normal case.The results indicate that by exploiting the correlation between stride time variability and Lyapunov exponents, one can establish a threshold for gait stability.

Magnetic nanoparticles guided by an external magnetic field can be used to deliver drugs (e.g., chemotherapeutics) to malignant regions (e.g., tumor tissue). Some nonmagnetic, sonosensitive nanoparticles are known to generate cavitation induced by focused ultrasound. This can can be used to monitor spatial/temporal particle distribution and to stimulate local drug release in target regions. Little to nothing is known about the sonosensitivity of magnetic nanoparticles. Therefore, we investigated this physical phenomenon in the present contribution. We showed that certain magnetic nanoparticles have high cavitation potential and are promising for further research in this regard.

Gait analysis is a systematic study of human movement. Combining wearable foot pressure sensors and machine learning (ML) solutions for a high-fidelity body pose tracking from RGB video frames could reveal more insights into gait abnormalities. However, accurate detection of heel strike (HS) and toe-off (TO) events is crucial to compute interpretable gait parameters. In this work, we present an experimental platform to study the timing of gait events using a new wearable foot pressure sensor (ActiSense System, IEE S.A., Luxembourg), and Google’s open-source ML solution MediaPipe Pose. For this purpose, two StereoPi systems were built to capture stereoscopic videos and images in real time. MediaPipe Pose was applied to the synchronized StereoPi cameras, and two algorithms (ALs) were developed to detect HS and TO events for gait and analysis. Preliminary results from a healthy subject walking on a treadmill show a mean relative deviation across all time spans of less than 4% for the ActiSense device and less than 16% for AL2 (33% for AL1) employing MediaPipe Pose on StereoPi videos. Finally, this work offers a platform for the development of sensor- and video-based ALs to automatically identify the timing of gait events in healthy individuals and those with gait disorders.

3D printing has been integrated in various applications in biomedical and medical imaging. A prominent new field is that of radiation protection and shielding, where the introduction of infused filaments makes it possible to manufacture customized 3D printed shielding. In this study we examine the use of a Polylactic Acid (PLA) infused with tungsten to shield a Positron Emission Tomography / Computer Tomography (PET/CT) scanner. The goal is to reduce the effects of the radiation emitted from outside the field-of-view (FOV) on the PET image and to protect the PET electronics against the X-rays from the CT. The shielding properties were examined using the radioactive sources Ba-133, Na-22, and Cs-137, as well as two X-ray sources with energy ranging from 20 to 125 kV. For X-rays below 40 kV, less than 1 cm of the tungsten infused shielding was sufficient to block more than 95% of the X-ray beam intensity. However, only around 30% of the 511-keV rays from Na-22 was attenuated by 3.5 cm of the 3D printed material. Our results support the application of tungsten infused PLA materials to shield the PET electronics. However, further studies are required to determine if shielding the PET crystals with tungsten infused PLA can improve image quality by reducing the contribution of scattered rays from outside the FOV. Additionally, our results show new possibilities for the use of infused filament in the field of Xray imaging. For example, where lead coverings or skirts are excessively bulky or rigid, it could be employed to optimize shielding for radiation-sensitive organs.

This study aimed to optimize the ablation depth and duration at a fixed energy using a Ho:YAG laser, considering fiber delivery applications. According to the measurements performed with an optical coherence tomography system, we obtained depths of about 1.4 mm with 500 mJ/pulse applied energy when moving the laser along a 1 cm line for 60 s (corresponding to ~50 pulses per position). Coupling this laser into a flexible fiber (θ= 250 μm) will be the next step for future minimal-invasive bone ablation trials. Furthermore, deeper ablation (~1.5 mm) was also achieved with different energy levels on bone samples.

For the use of polymers as biomaterials in the production of implants, several processing methods are needed. Besides injection molding or Fused Deposition Modeling as temperature based processes, also solvent based processes were performed such as solvent-casting and electrospinning. As fabrication of medical devices makes use of potentially hazardous or toxic solvents, they have to be extracted from the final medical device. For analyses of residual solvents different methods can be used e.g. gas chromatography (GC), coulometric elemental analysis or thermogravimetrical analyses (TGA). Here we present first results for the use of X-ray photoelectron spectroscopy (XPS) to yield elemental as well as spatial information on residual solvents in polymeric materials. Therefore we analysed PLLA as biodegradable electrospun polymer, TPC-ET as biostable polymer and a PLLA-iron composite as example of a polymer metal composite via XPS and etching processes. It was possible to determine residual halogen containing solvents in all tested polymers processed from polymeric solution. As discrimination between polymer and solvent was made possible by specific elements, XPS measurements may be a suitable tool to identify even small amounts of solvent. Thus XPS analyses will help to optimize polymer processing in order to remove residual solvents.

Ultrasound (US) imaging is used for the diagnosis and also evaluation of thyroid nodules. A Thyroid Imaging Reporting and Data System (TIRADS) is used for the risk stratification of thyroid nodules through US images. The composition of thyroid nodules plays an important role in the risk-stratification process. The percentages of cystic and solid components in a thyroid nodule are one of the features that are can be indicative of the risk of malignancy. In this work, we attempt to classify and estimate solid and cystic regions within nodules. 20x20 texture patches were extracted from solid and cystic regions and converted into signals. These signals are decomposed into low, mid, and high-frequency bands using Continuous Wavelet Transform (CWT). A total of 36 features were extracted from the decomposed signals using Auto- Regressive Modeling. The features were fed into three different Machine Learning (ML) algorithms (Artificial Neural Networks, K-Nearest Neighbors, and Random Forest Classifier) to provide us with a classification of solid versus cystic regions in thyroid nodule US images. The Random Forest Classifier obtained an Accuracy, Sensitivity, and Specificity of 90.41%, 99% and 91% respectively which was the highest among the three chosen ML algorithms. Additionally, the output from the classification phase was also be used to determine the percentage of cystic and solid regions with a given thyroid nodule US image.

The increasing trend of minimally invasive therapy methods for widespread diseases in the cardiovascular system, in the field of ophthalmology, ear, nose, and throat medicine, or gynecology, leads to an extensive demand for innovative microdevice manufacturing technologies. Femtosecond (fs)-laser technology offers a wide range of applications in the field of cutting, milling, and surface structuring on the micro- and nanometer scale. Ophthalmic devices for glaucoma therapy have dimensions in the submillimeter range. Irrespective of this, there are high demands on the safety and efficacy of these implants. Within the current study, fs-laser-based manufacturing technologies for functional elements of a novel microstent for glaucoma therapy were developed, considering potential industrial scalability. We successfully developed a mounting mechanism for quick and reproducible manufacturing of a micromechanical valve mechanism for safe regulation of intraocular pressure. Furthermore, we successfully developed a fs-laser-based process for manufacturing grooves as part of a fixation mechanism for a stable device positioning inside the eye. The results show that fs-laser manufacturing represents a promising technology for manufacturers of medical microdevices, such as ophthalmic microstents.

Introduction: In endoscopic procedures of the nasal sinuses, a critical issue for classification tasks is the ambiguous anatomical representations due to their complex composition. We investigated the potential of multi-label image-based classifications of sinus landmark combinations together with explainability methods for machine learning in an assistance function at the application level. By combining image classification and pixel attribution in a navigation function, we provide the surgeon with label predictions and additional localization cues of important pixels relevant to the model output with regard to the input image. Methods: We used 3500 label annotated video sequences from 30 recorded sinus surgeries to fine-tune a pretrained ResNet50 as the feature extractor and a classification head using binary cross-entropy on one-hot encoded target vectors of landmark classes with a RAdam optimizer over 28-32 epochs. Image augmentation and a focal loss function were added to counter over-fitting. An explainability function used the trained model to produce pixel attribution maps for predicted classes regarding individual input images. These gradient maps were summed over all classes, and pixel values above 0.20 were clustered using weighted k-means based on the gradient value of each pixel coordinate. The resulting cluster centroids were then overlapped into the endoscopic image with the predicted landmark classes. Three surgeons investigated three different overlay scenarios in a validation study. Results: The top-1 predictions reached a mean f1-score of 0.47, with the highest values of 0.71 and the lowest with 0.28. Despite over-fitting mitigation, prediction results largely depended on over- and underrepresented classes. Conclusion: The provided explainability function at the application level showed the strong potential of delivering visual cues from prediction results to surgeons at runtime to support human-machine interaction.

Applications and requirements for biomedical test benches are often quite different, but there is always a need to monitor and usually to control processes. Commercially available development systems can be used to build simple laboratory setups up to industrial systems but are limited with respect to high costs and information for required validation. A test bench development kit is presented which ensures high compatibility for prospective developments, adaptations and embedding of sensors and actuators. The development kit is based on a printed circuit board built around the singleboard computer Raspberry PI. Supporting circuitry for power control, logics and special functions are implemented. Several types of peripheral devices can be connected via two serial bus standards (I2C, or RS-485). Current applications demonstrate the achieved level of automation and process control.

The number of electrodes used to acquire neonatal EEG signals varies between institutions. Therefore, tools for automatic EEG analysis, such as neonatal seizure detection algorithms, need to be able to handle different electrode montages in order to find widespread use. The aim of this study was to analyse the effect of montage on neonatal seizure detector performance. A full 18-channel montage was compared to reduced 3- and 8-channel montages using a convolutional neural network for seizure detection. Sensitivity decreased by 10 - 18 % for the reduced montages while specificity was mostly unaffected. Electrode artefacts and artefacts associated with biological rhythms caused incorrect classification of nonseizure activity in some cases, but these artefacts were filtered out in the 3-channel montage. Other types of artefacts had little effect. Reduced montages result in some reduction in classifier accuracy, but the performance may still be acceptable. Recording artefacts had a limited effect on detection accuracy.

The funded research Project HealthCONNECT implements measure to promote the term eHealth in teaching and means to ease the exchange of lecturers and students to perceive different views of this topic. This report introduces the activities that have been implemented, stresses challenges that were observed and derives suggestions how to establish a co-operation between different institutes of higher education. This includes the mutual participation and inclusion of lecturers in teaching and evaluation activities; spanning from assessing students’ projects till chairing master examinations. The co-operative organization and hosting of eHealth Summer Schools is another mean to share knowledge in the domain of eHealth with students and colleagues abroad. Such activities require a stable co-operation between the involved institutions that only can be build upon a trustworthy relationship between the lecturers as they are paving the way.

Residual fragments can remain after kidney stone extraction which may necessitate another intervention. Magnetic nanoparticles (MNP) can be applied due to their properties of being able to bind to residual fragments and to be extracted by an external magnetic field. Calcium oxalate crystals and magnetic nanoparticles have been synthesized, characterized and used for binding and magnetic separation studies. The separation is validated by simulation and experimentally. MNP bind covalently to the fractals and a magnetic extraction of calcium oxalate fractals is possible. The agglomeration of MNP can be induced with the addition of salt which improves the extraction process. This proof-of-concept study is the fundament for a new way of stone extraction and can pave the way for new procedures in urology.

Tiffeneau manoeuvres are an important method in pulmonary function testing of the human lungs and can help to diagnose respiratory obstructions. Data from a motion capture system and a spirometer was used to evaluate Tiffeneau indexes which can theoretically be derived via a Smart-Shirt that incorporates three circumference measurements of the upper body. The mean error was 4.5% regarding the Tiffeneau indexes gained by the spirometer, indicating that clinical diagnosis of obstruction is potentially possible using a Smart-Shirt.

The development of three dimensional geometries by electrospinning of nonwoven, bioresorbable PDLLA nanoand microfibers is a prerequisite for multimodal growth factor delivery by biohybrids. Electrospinning of 3D tubes is a formidable task since the deposited nanofibers on the rotating mandrel adhere so strongly that newly formed tubes cannot be removed. A specific electrospinning instrument setup with a "removable mandrel sleeve" technique facilitated the reproducible synthesis of tubes for guided bone and future vascular applications. It is shown, that this technique not only allows the preparation of tubes but also of self-supported 3D rings and sandwich panels which can be functionally doped by growth factors such as rhBMP-2 to biohybrids.

Introduction: In this work, we use simulated data to quantify the different failure mechanisms of a previously presented low-cost jump height measurement system, based on widely available consumer smartphone technology. Methods: In order to assess the importance of the different preconditions of the jump height measurement algorithm, we generate a synthetic dataset of 2000 random jump parabolas for 2000 randomly generated persons without real-world artifacts. We then selectively add different perturbations to the parabolas and reconstruct the jump height using the evaluated algorithm. The degree to which the manipulations influence the reconstructed jump height gives us insights into how critical each precondition is for the method’s accuracy. Results: For a subject-to-camera distance of 2.5 meters, we found the most important influences to be tracking inaccuracies and distance changes (non-vertical jumps). These are also the most difficult factors to control. Camera angle and lens distortion are easier to handle in practice and have a very low impact on the reconstructed jump height. The intraclass correlation value ICC(3,1) between true jump height and the reconstruction from distorted data ranges between 0.999 for mild and 0.988 for more severe distortions. Conclusion: Our results support the design of future studies and tools for accurate and affordable jump height measurement, which can be used in individual fitness, sports medicine, and rehabilitation applications.

Involuntary female infertility represents a sensitive issue, which is frequently caused by an impaired patency of the Fallopian tube. Current treatment options, such as in vitro fertilisation (IVF) or tubal reconstruction, are related to high costs or surgical risks. Therefore, a need for an alternative highly safe and effective minimally invasive therapy option is assumed. The current work presents a proof-of-concept for a novel microstent in combination with a delivery system for treatment of proximal Fallopian tube occlusions. For this purpose, a microstent prototype was implanted into a porcine Fallopian tube ex vivo using the presented delivery system. The results of the procedure were analyzed using a micro-computed tomography (μ-CT). Based on the μ-CT imaging, critical parameters of the microstent, such as radial patency, were assessed. The microstent implantation leads to an increased opening area of the Fallopian tube. Additionally, the microstent does not impair the anatomical shape of the tube epithelium and adapts well to the anatomical path. In further studies, the functionality of the Fallopian tube will be examined after microstent implantation in vivo.

Sensor-integrated needles with impedance or other electrochemical sensors have gained increased attention in research. Their potential use is widespread and covers applications such as early disease diagnostics or needle guidance for needle-based interventions. However, fabrication and sensor integration are crucial parts of realizing functionalized needles. In this paper, a review of existing research on microscale sensor arrays on curved needle surfaces is presented. It focusses on medical needles excluding needle-type sensors. Especially, means of fabrications including flexible thin-film-based sensors or photolithography are discussed. Although they provide satisfactory results, no manufacturing method is yet ready for bulk production.

Full-field electroretinography (full-field ERG) forms the diagnostic basis for numerous pathologies of the eye. For this reason, fast and accurate diagnostics in the field of ophthalmology are essential. Two examination techniques, full-field ERG and pupillometry were combined in a diagnostic device developed by ICM e.V. to reduce the examination process for both examiners and patients. In this paper, the device is examined for the quality of the full-field ERG measurements. A feasibility study with 12 healthy subjects (3 f, 9 m, 36.33 ± 11.94 years) was conducted to evaluate the device. The results showed that the peak times for both light- and dark-adapted measurements were within the range of the researched literature values. However, the amplitudes were markedly lower in both measurements compared to the averaged literature values (dark-adapted about 8.5-fold and light-adapted about 5.5-fold) and are clearly outside the range of values researched. The main reason for this is the use of cup electrodes, which were placed on the skin of the lower eyelid. Nevertheless, plausible and comparable analysis values could be obtained with the eye diagnostic device PEP-2000. Further studies with wire electrodes will be performed.

Retractors are used in surgery for the displacement of tissue and organs and to keep the surgical area open. The retraction can lead to a tissue compression that can cause irritation on nerves, perfusion deficits, hematoma or late sequelae like wound healing disorders, pain and numbness. The retraction force and thus the load on the tissue can not be measured with retraction systems today. We present a new concept for a retractor with an integrated spring element and an indicator to visualize retraction force and allow an adjustment of the organ compression. The new retractor was designed considering organ compression limits, requirements on sterility and easy manufacturing. Two retractors were built up as a prototype by 3D printing and tested in a distinguished test setup. The first test of the retractors was performed on a force measurement test bench using a fixed retractor, and in a second test, a liver specimen in a bowl. Tests were repeated for both retractor variants. The tests showed how different spring characteristics could be realized with the new retractor design. Depending on the spring characteristics, the compression could be limited to a certain level and flexibility that can compensate for unwanted patient motion. An indicator is integrated into the design showing the applied load on the retraction system. The presented concept can help to increase patient safety and reduce tissue trauma and late sequelae. The indicator is an easy way to visualize applied forces and allow adaption of the setup to the surgeon's needs.

Automated surgical tool classification will improve the workflow of surgery. Previous research tackled this task mainly in cholecystectomy procedures due to availability of a relatively large and labelled set (Cholec80 dataset). However, the complexity of the procedure type has an impact on the robustness of the deep learning approaches. Therefore, the classification capability of CNNs on data of more complex procedures with many surgical tools was investigated. In this work, laparoscopic videos of 14 gynaecological procedures were recorded and labelled for surgical tool presence. Then, the DenseNet-121 model was trained to identify surgical tools according to functionality. Experimental results imply high classification performance for some surgical tools. The mean average precision over all the tools was 67%. This study is an initial benchmark for detecting surgical tools in realistic settings.

Pupillometry forms the diagnostic basis for numerous pathologies of the eye. For this reason, fast and accurate diagnostics in the field of ophthalmology are essential. Two examination techniques, full-field ERG and pupillometry were combined in a diagnostic device developed by ICM e.V. to reduce the examination process for both examiners and patients. In this paper, the device is examined for the quality of the pupillometry measurements. A study with 12 healthy subjects (3 f, 9 m, 36.33 ± 11.94 years) was conducted to evaluate the device. The results showed that the minimal pupil diameter is 40 % higher than the literature values. The main reason for the differences is the low light intensity of 15 cd/m2. However, the maximum pupil diameter is within the range of the researched values. The results of the pupillary reaction measurements show that the values obtained (amplitude, contraction time and peak time) are within the range of literature values. The latency time of 690 ms is 40 % too high. The reason for this could be the moderate pupil detection rate of 50-70 %. Nevertheless, plausible and comparable analysis values could be obtained with the eye diagnostic device PEP-2000. Further work will look at improving pupil detection rates.

This paper proposes a cost-effective VR surgical training system which computes haptic feedback forces when a VR surgical tool interacts with virtual tissues. A 3 degrees of freedom (DoF) reverse linear Delta mechanism is used to render computed force feedback data which are then received by the fingertip of the operator. Additionally, the moving plate allows rendering of surface properties and lateral forces occurring due to a tumor with different stiffness parameters below the skin surface. Controllers are designed and implemented to regulate the haptic feedback device’s end-effector position and applied force. The virtual surgical instruments are controlled by a 7DoF serial link manipulator which captures the operator’s movement by the utilization of various sensors. The controllers to regulate forces as well as the positions are evaluated with the proposed haptic feedback device. The mean RMSE of the force and mean error of the angular displacement are 0.0707 N and 1.95°, respectively. The presented system can provide multisensory feedback including visual, auditory and haptic feedback interactively depending on the operator’s input in the presented VR surgical training system.

Cardiovascular diseases are among the most common diseases with high mortality, including aortic valve stenosis and insufficiency. Minimally invasive implantation of transcatheter aortic valve prosthesis (TAVI) has become the standard procedure for patients with increased risk for open surgery. It is commonly accepted that the long-term outcome of aortic valve replacement depends on hemodynamic performance. This motivates the analysis of the velocity field in the vicinity of the TAVI. Computational fluid dynamics (CFD) methods have been established in the past, but show limitations in terms of computational effort when rapid design optimization or patient-specific decision making in real time is required. In this study we show the usage of PINNs for predicting fluid flow through a TAVI device. We also show a method of enforcing boundary conditions for this specific problem. Due to the physics involved in the training process, this principle does in theory not require additional training data. To validate the method, we performed CFD simulations that solved the Navier-Stokes equations be means of finite volume methods. Besides the good estimation of the main flow components, discrepancies between CFD and PINN results are present. Nevertheless, the flow structures have certain similarities in the coarse spatial localization of the vortex patterns occurring in flow around the TAVI device.

Spinal cord injury (SCI) has a drastic effect on the quality of life of those affected. There is thus a great need for new knowledge that may increase their quality of life. The goal of this research was to develop and set up a measurement system for exact measurements of times between events in a patellar reflex test and a transcutaneous spinal cord stimulation (tSCS) and compare the two in healthy and spinal cord-injured individuals. 21 individuals with a healthy nervous system and 2 spinal cord-injured individuals were subjected to the patellar reflex test and tSCS. The patellar reflex of one individual with a complete SCI at vertebras C6-C7 after an accident was delayed by 24% compared with the healthy subjects, while the reflex time of the other SCI individual, who had cancer in vertebra T3, was shortened by 23%. There was also a difference in the reflex between the patellar reflex test and the tSCS due to action potential conduction distance. In this study, a processing method was created for comparing muscle signals generated by patellar strikes or electrical stimulation to the posterior nerve roots. Data from the subjects with central nervous system lesions showed characteristic differences from reference data.

Paroxysmal brain state changes, such as microsleep events in drivers, are presumably subcortically induced and accompanied by cortical processes. This raises questions of how stable and persistent are the cortical processes that can be observed with EEG in real time. For this purpose, recordings of four night-driving simulation studies including 79 subjects are used to analyze how large the time window and how the temporal offset of the EEG segment must be to achieve maximum classification rate. From each EEG segment, power spectral densities were estimated using modified periodogram method and averaged in narrow spectral bands. They were then processed using gradient boosting machines in order to map them to one of two brain state types: microsleep or sustained attention. Segment length and offset were found to have moderate and dramatic effects on recognition rate, respectively.

Correctly torquing bone screws is important for implant fixation longevity and strength. A model-based smart screwdriver has been proposed to regulate torque, however current models ignore axial force and the initial engagement of the screw. This was addressed here by deriving a model based on the concept of a net axial force generated by a difference in contact areas on the inward and outward sides of the screw threads. This force is opposed by the shear strength of the material around the threads. The results of the derivation were able to predict the effects of different axial forces during insertion in relation to the hole material strength. The results may be used to compensate for initial thread breakage in future model-based smart screwdrivers to improve their accuracy.

Soft, stretchable, and flexible electrodes offer a unique and nonintrusive option to monitor essential biopotential signals from the skin, which can be related to the functioning and status of different organs including the heart. Such stretchable electrodes, by forming an intimate contact to the human skin, not only increase the signal-to-noise ratio of the recorded biopotentials and reduce eventual signal artifacts, but are also more pleasant to wear. We developed a stretchable electrode patch for electrocardiogram (ECG) monitoring utilizing the medical-grade, transparent, and stretchable material Styrene-ethylene-butylene-styrene (SEBS). Our fabrication process is compatible with standard micro- and nano clean room processes and could be up-scaled accordingly.

The development of novel and innovative biomaterials and biomedical devices like medical implants is a sophisticated task that requires expertise ranging from physicians over engineers and natural scientists to regulatory officers for clinical translation to be most efficient. This short communication should highlight the specific requirements and perspectives of the different disciplines and conclude on the importance of interdisciplinary training for early-stage researchers. For this purpose, experts for clinical practice, material development, biological testing and medical device approval with profound experience in interdisciplinary research summarized and discussed current literature and knowledge. Multiple interactions right from the start of the design phase until the medical device approval could clearly be identified. As not all relevant study programs already include interdisciplinary training, implementing structured interdisciplinary doctoral programs is advised to accelerate innovative biomedical implant development.

Auscultation methods enable non-invasive diagnosis of diseases, e.g. of the heart, based on heartbeat sounds. Regular, early examinations using machine learning techniques could help to detect diseases at an early stage to prevent serious health conditions and then provide optimal therapy through continuous monitoring. There is already a lot of work on human data using AI algorithms to detect patterns in signals or images. However, there is hardly no work on detecting heart murmurs with digital such as Myxomatous Mitral Valve Disease. In this paper, we present a canine auscultation project that aims to provide a tool to establish a baseline of classification parameters from audio signals that could be used to monitor canine health status by analyzing deviations from this baseline. In the future, data analysis could also lead to prediction and early detection of other diseases.

If the alignment of a leg prosthesis is inadequate there will be the danger of arthritic defects occurring in the preserved knee joint. Acoustic-kinetic joint analysis in orthopaedic technology allows monitoring of the knee joint of an amputee after the alignment of a leg prosthesis and to recognize early joint defects resulting in adaptation of the alignment of the prothesis. A test rig was developed which includes acoustic emission analysis of the knee, video-based gait analysis, ground reaction force measurement and the measurement of the holding forces of the hands. During the examination a subject must do 3 one-legged knee bends during 10 sec with the preserved leg. To support the subject two grab handles were fixed with force sensors to a frame. A series of 6 measurements were recorded with breaks of 2 min. The recorded forces of the leg show a double hump pattern during the knee bend. The rising edge of the flexion phase was not uniform. The forces show no conclusive uniform pattern. The fitness of the probands acts as an indicator for the pattern of the force signals. The signals of the measurements show the necessity of the support for both hands to execute a one-legged knee bend. The understanding of the 3 measured forces is important for the arthritic diagnosis of the knee joint.

Various types of magnetic nanoparticles have been developed for a range of medical applications and some have the potential to treat diseases. In this paper, a new method for controlling the position of these magnetic particles by changing the magnetic field surrounding them using three Halbach ringsis investigated with the aid of a simulation. The inner ring controls the direction, and the position of the two outer rings can keep the group of particles together. In the simulation, the magnetic field is calculated analytically for each time step and the movement of the particles is simulated by applying the corresponding forces. These result from the manipulable external magnetic field, the magnetic field between the particles, the Lennard-Jones potential to model attractive interactions between the particles and the drag. In the simulation, we show how the particles are manipulated using a Levenberg-Marquard algorithm and develop a closedloop control for the position of the grouped particles by manipulating the magnetic field inside the Halbach rings by changing the angular position of these rings.

Capacitive extracellular stimulation is a common method in implanted stimulation electrodes. The basis for investigating the transmission of stimuli from an electrode to adhered cells are in vitro experiments using calcium imaging or patch clamp measurements. Computational spatial models are used to simulate the mechanism of signal transmission at the cell-electrode interface but require high computing power. In this work, the Stern model to characterize the electrochemical double-layer (EDL) formation and a modified two-domain model are combined into a holistic equivalent circuit modelling capacitive cell stimulation. The described parameters can be directly associated with physicochemical effects. A simulation of the involved control and measurement systems allows the validation of the model with in vitro patch clamp recordings. The relationship of the cell’s distance to the electrode and efficacy of signal transmission could be observed. With this concept we aim to convert different complex approaches into a simple model and thus give an overview of the mechanisms of stimulation. We want to facilitate the interpretation of measured signals especially in voltage clamp measurements.

Recent in-vivo work showed the suitability of Pickering-stabilized antibubbles in harmonic imaging and ultrasound-guided drug delivery. To date, however, theoretical considerations of antibubble core properties and their effects on antibubble dynamics have been rather sparse. The purpose of this study was to investigate the influence of skeletal friction on the damping of a pulsating antibubble and the pulsation phase of an antibubble relative to the incident sound wave. Numerical simulations were performed to compute damping terms and pulsation phases of micron-sized antibubbles with thin elastic shells and 30% endoskeleton volume fraction. The simulations showed that the damping owing to skeleton presence dominates the damping mechanism for antibubbles of radii less than 2.5 μm, whilst it is negligible for greater radii. The pulsation phase of such small antibubbles was simulated to have a phase delay of up to 1/6 π with respect to pulsating free gas bubbles. Our findings demonstrate that the presence of an endoskeleton inside a bubble influences pulsation phase and damping of small antibubbles. Antibubbles of radii less than 3 μm are of interest for the use as ultrasound contrast agents.

Electrical impedance tomography (EIT) has the potential for monitoring perfusion in addition to respiration on the bedside. Several separation methods were reported, e.g. Filtering, ECG-gating and PCA. However, the separation is not trivial, which is why harmonic analysis was introduced in EIT data analysis. It is based on several assumptions and therefore prove of plausibility is necessary. In this contribution a two-dimensional thorax simulation is introduced, which includes simplified lungs and heart with changing shapes and conductivities due to volume variations and lung perfusion. Harmonic analysis was applied on the simulated data. The separation results are in good agreement with the simulation settings. Further investigations will be required due to limitations of the simulation. Nevertheless, harmonic analysis delivered precise results in ideal and strictly defined conditions.

Several contributions are presented to aid the analysis of hyperspectral images by utilizing appropriate optical models. The algorithms are described by means of the perfusion of human skin and a corresponding two-layer tissue model. The particular role of the mean photon path length within the medium is discussed and two approaches to solve the inverse problem, i.e., the derivision of perfusion parameters from given spectra, are outlined.

Smart underwear was designed to meet the demands of elderly people be monitored in outpatient environments. Especially the demands of women have been addressed, who have heart failure but who still want to live in their home. The smart bra comprises textile electrodes and an ECG event recorder. The recorder transmits the data to the doctor only if artificial intelligence (AI) has discovered critical events in ECG signals. An implemented convolutional neural network, which uses the wavelet transformation of the ECG as input and to determine an interference index for this ECG signal is implemented. The smart bra can be part of new remote monitoring services such as e-consultation or as a new home-based follow-up to ensure seamless patient care after hospital or rehabilitation stays. Healthcare personnel has better access to the patient´s digital ECG data without direct consultation, independently of where the patient lives and the patient can feel safer.

For the development of new types of hip implants for acetabulum revision, it is beneficial to analyse the acetabular defects of the indication group in advance. In order to be able to specially compare the bone defects with each other, a normalisation and accompanying scaling of the pelvis is necessary. Uniform scaling is required so that the bone structures are not distorted. In the following study, three scaling methods based on the minimal bounding box and sphere principle are compared with a method using 14 landmarks on the pelvis.The landmark method is applied to determine the true scaling factor. For the comparison of the different methods, 40 female pelvic models with an acetabular defect are analysed. In the comparison of the scaling methods, the method using minimal bounding spheres shows the least deviation from the landmark method (mean difference 3.30 ± 2.17 %). Due to the fact that no preprocessing (definition of the landmarks) is required and the fast implementation of the algorithm, the minimal bounding sphere is to be preferred to the landmark method for a fast size estimation.

Sleep quality has a significant impact on human physical and mental health. The detection of sleep-wake states is thus of significant importance in the study of sleep. The performance of classical machine learning models for automated sleep detection depends on the signals considered and feature extraction methods. Moreover, hand-crafted features are highly dependent on the experts’ experience and their prior knowledge about different physiological signals and conditions of the subjects. To overcome this limitation, this paper develops an end-to-end deep learning approach for automated feature extraction and detection of sleep-wake states using single channel raw EEG signals. Moreover, we leverage transfer learning to train and fine tune the proposed model to avoid the complexities associated with building a deep learning model from scratch. Using polysomnography (PSG) data of 20 patients, our results demonstrate the effectiveness of the proposed deep learning pipeline, achieving an excellent test performance in detecting sleep events with an overall sensitivity and precision of 92.7% and 92.1% respectively. The results demonstrate that the proposed approach can achieve superior performance compared to state-of-the-art studies on Sleep- Wake classification. Furthermore, it can attain reliable results as an alternative to classical methods that heavily rely on expert defined features.