Volume 66, Issue 12 - Special Issue: Automation in medicine: from homecare to clinical applications / Walter Karlen, Thomas Schauer, Knut Möller, Olaf Simanski

Zusammenfassung Durch die dynamische Entwicklung von mobilen Sensoren bieten sich Medizinern ständig neue, potentiell kosteneffektive Wege für Diagnostik, Monitoring und Screening. Angetrieben durch den wachsenden Erfolg auf dem Konsumentenmarkt streben immer mehr Wearables & Co in den Gesundheitsmarkt. Weltweit beschäftigen sich Forschungsgruppen seit Jahren mit den Auswirkungen dieser mobilen Technologien auf unser Gesundheitswesen. In der Menge der zahlreichen Möglichkeiten und in Anbetracht der rapiden Entwicklung sollte jedoch bedacht werden, dass für einen erfolgreichen Einsatz am Patienten einige Herausforderungen und Risiken berücksichtigt werden müssen. Der Artikel bietet einen Einstieg in das breite Themenfeld rund um die Einführung von Wearables in die klinische Routine und berichtet über eine Auswahl an realisierten und potentiellen Einsatzmöglichkeiten.

Recently, telemonitoring of vital signs has gained a lot of research interest. Especially for electrocardiogram (ECG), which is among the most frequently measured vital sign. However, the integration of classical ECG Holter in wearables is problematic since shielded cables and gel electrodes are required to get ECG signals of highest quality. We have recently introduced a novel sensing architecture based on active electrodes (so-called cooperative sensors) that significantly reduces the cabling complexity of the monitoring device. After briefly recalling the principle of cooperative sensors this paper details how they address rejection of common-mode voltage induced by electromagnetic disturbances. The proposed approach uses an auto-identification technique based on a continuous-time calibration of the sensor system and a digital control loop. To demonstrate the reliability of the proposed approach, a 12-lead ECG monitoring system was implemented with the new common-mode rejection method. Measurements on four healthy volunteers showed that the signal quality obtained with the cooperative-sensor system (using dry electrodes) is equivalent to the one measured with a gold standard medical device (using gel electrodes) in exercise stress tests.

Undesired forces during human-robot interaction limit training effectiveness with rehabilitation robots. Thus, avoiding such undesired forces by improved mechanics, sensorics, kinematics, and controllers are the way to increase exoskeleton transparency. In this paper, the arm therapy exoskeleton ARMin IV+ was used to compare the differences in transparency offered by using the previous feed-forward model-based controller, with a disturbance observer in a study. Systematic analysis of velocity-dependent effects of controller transparency in single- and multi-joint scenarios performed in this study highlight the advantage of using disturbance observers for obtaining consistent transparency behavior at different velocities in single-joint and multi-joint movements. As the main result, the concept of the disturbance observer sets a new benchmark for ARMin transparency.

The generation of precise hand movements with functional electrical stimulation (FES) via surface electrodes on the forearm faces several challenges. Besides the biomechanical complexity and the required selectivity, the rotation of the forearm during reach-and-grasp tasks leads to a relative change between the skin and underlying tissue, resulting in a varying FES response. We present a new method for automatic adaptation of virtual electrodes (size, position) and stimulation intensity in an electrode array to guarantee a secure grasp during forearm movements. The method involves motion tracking of arm and hand with inertial sensors. This enables the estimation of grasping strength when using elastic objects. Experiments in healthy volunteers revealed that our method allows generating a strong, stable grasp force regardless of the rotational state of the forearm.

A hysteresis loop analysis (HLA) method for breast cancer diagnosis based on a non-invasive digital imaging elasto-tomography (DIET) screening system is evaluated using data from 3 clinical trial patients, comprising 2 healthy breasts and 4 breasts with cancer. The identified mechanical nominal stiffness with ∼2x higher values compared to healthy tissue stiffness localized the correct cancerous area (CA) or tumor location, matching the mammography detection for all 4 breasts with cancer. The difference in identified stiffness varies across different frequencies and individuals. However, the identified stiffness for all healthy breasts and/or health tissue regions are consistent across frequencies, avoiding false positive diagnosis. The overall approach can be implemented automatically without requiring a skilled operator, thus reducing the screening cost.

Recent studies demonstrate therapeutic benefits in retinal laser therapy even for non-visible effects of the irradiation. However, in practice, ophthalmologists often rely on the visual inspection of irradiation sites to manually set the laser power for subsequent ones. Since absorption properties vary strongly between sites, this procedure can lead to under- or over-treatment. To achieve safe automatic retinal laser therapy, this article proposes a robust control scheme based on photoacoustic feedback of the retinal temperature increase. The control scheme is further extended to adapt to real-time parameter estimates and associated bounds on the uncertainty of each irradiation site. Both approaches are successfully validated in ex vivo experiments on pigs’ eyes, achieving consistent irradiation durations of 55 ms despite the uncertainty in absorption properties.

The CPAP (Continuous Positive Airway Pressure) therapy is the gold standard to treat the obstructive sleep apnoea syndrome (OSAS). The CPAP-device generates a positive pressure to splint the upper respiratory tracts preventing an collapse of the tracts and the pharynx. It has been shown that many CPAP-devices cannot maintain the pressure set points adjusted by the medical staff, because the mask pressure is seriously influenced by the patient’s breathing. In this work a method is provided to estimate the breathing effort of the patient. Therefore a model of the breathing therapy system is introduced and the application of the Kalman filter is described. The estimated breathing effort could then be used in further control strategies to improve the control quality.

Conventional bone surgery leads to unwanted damage to the surrounding tissues and a slow healing process for the patients. Additionally, physicians are not able to perform free cutting shapes due to the limitations of available systems. These issues can be overcome by robot-assisted contactless laser surgery since it provides less mechanical stress, allows precise functional cuts, and leads to faster healing. The remaining drawback of laser surgery is the low ablation rate that is not yet competitive with conventional mechanical piezo-osteotomes. Therefore, we aim at maximizing the efficiency in hard tissue laser ablation by optimizing the lateral movement speed for different irrigation conditions. The results of this study show a non-linear relationship between cutting rates, speeds, and depths that should be critically considered for integration in robotic laser surgery.