Volume 64, Issue 6

Due to the excitability of neurons in the brain, a neurological disorder is produced known as epilepsy. The brain activity of patients suffering from epilepsy is monitored through electroencephalography (EEG). The multivariate nature of features from time domain, frequency domain, complexity and wavelet entropy based, and the statistical features were extracted from healthy and epileptic subjects using the Bonn University database and seizure and non-seizure intervals using the CHB MIT database. The robust machine learning regression methods based on regression, support vector regression (SVR), regression tree (RT), ensemble regression, Gaussian process regression (GPR) were employed for detecting and predicting epileptic seizures. Performance was measured in terms of root mean square error (RMSE), squared error, mean square error (MSE) and mean absolute error (MAE). Moreover, detailed optimization was performed using a RT to predict the selected features from each feature category. A deeper analysis was conducted on features and tree regression methods where optimal RMSE and MSE results were obtained. The best optimal performance was obtained using the ensemble boosted regression tree (BRT) and exponential GPR with an RMSE of 0.47, an MSE (0.22), an R Square (RS) (0.25) and an MAE (0.30) using the Bonn University database and support vector machine (SVM) fine Gaussian with RMSE (0.63634), RS (0.03), MSE (0.40493) and MAE (0.31744); squared exponential GPR and rational quadratic GPR with an RMSE of 0.63841, an RS (0.03), an MSE (0.40757) and an MAE (0.3472) was obtained using the CHB MIT database. A further deeper analysis for the prediction of selected features was performed on an RT to compute the optimal feasible point, observed and estimated function values, function evaluation time, objective function evaluation time and overall elapsed time.

The main idea of brain-computer interfaces (BCIs) is to facilitate the lives of patients having difficulties to move their muscles due to a disorder of their motor nervous systems but healthy cognitive functions. BCIs are usually electroencephalography (EEG)-based, and the success of the BCIs relies on the precision of signal preprocessing, detection of distinctive features, usage of suitable classifiers and selection of effective channels. In this study, a two-stage channel selection and local transformation-based feature extraction are proposed for the classification of motor imagery/movement tasks. In the first stage of the channel selection, the channels were combined according to the neurophysiological information about brain functions acquired from the literature, then averaged and a single channel was formed. In the second stage, selective channels were specified with the common spatial pattern-linear discriminant analysis (CSP-LDA)-based sequential channel removal. After the channel selection phase, the feature extraction was carried out with local transformation-based methods (LTBM): local centroid pattern (LCP), one-dimensional-local gradient pattern (1D-LGP), local neighborhood descriptive pattern (LNDP) and one-dimensional-local ternary pattern (1D-LTP). The distinctions and deficiencies of these methods were compared with other methods in the literature and the classification performances of the k-nearest neighbor (k-NN) and the support vector machines (SVM) were evaluated. As a result, the proposed methods yielded the highest average classification accuracies as 99.34%, 95.95%, 98.66% and 99.90% with the LCP, 1D-LGP, LNDP and 1D-LTP when using k-NN, respectively. The two-stage channel selection and 1D-LTP method showed promising results for recognition of motor tasks. The LTBM will contribute to the development of EEG-based BCIs with the advantages of high classification accuracy, easy implementation and low computational complexity.

The aim of this study was to compare a reconfigurable mobile electroencephalography (EEG) system (M-EMOTIV) based on the Emotiv Epoc ® (which has the ability to record up to 14 electrode sites in the 10/20 International System) and a commercial, clinical-grade EEG system (Neuronic MEDICID-05 ® ), and then validate the rationale and accuracy of recordings obtained with the prototype proposed. In this approach, an Emotiv Epoc ® was modified to enable it to record in the parieto-central area. All subjects (15 healthy individuals) performed a visual oddball task while connected to both devices to obtain electrophysiological data and behavioral responses for comparative analysis. A Pearson’s correlation analysis revealed a good between-devices correlation with respect to electrophysiological measures. The present study not only corroborates previous reports on the ability of the Emotiv Epoc ® to suitably record EEG data but presents an alternative device that allows the study of a wide range of psychophysiological experiments with simultaneous behavioral and mobile EEG recordings.

A new technique for electronic fetal monitoring (EFM) using an efficient structure of neural networks based on the Legendre series is presented in this paper. Such a structure is achieved by training a Legendre series-based neural network (LNN) to classify the different fetal states based on recorded cardiotocographic (CTG) data sets given by others. These data sets consist of measurements of fetal heart rate (FHR) and uterine contraction (UC). The applied LNN utilizes a Legendre series expansion for the input vectors and, hence, has the capability to produce explicit equations describing multi-input multi-output systems. Simulations of the proposed technique in EFM demonstrate its high efficiency. Training the LNN requires a few number of iterations (5–10 epochs). The applied technique makes the classification of the fetal state available through equations combining the trained LNN weights and the current measured CTG record. A comparison of performance between the proposed LNN and other popular neural network techniques such as the Volterra neural network (VNN) in EFM is provided. The comparison shows that, the LNN outperforms the VNN in case of less computational requirements and fast convergence with a lower mean square error.

The operating life of endoscopy capsules is limited by the batteries needed for use. The capacity of corresponding batteries, however, is too short to cover the total period of gastrointestinal transit. To overcome this limitation, a novel kind of wireless power transmission was used. It utilizes a permanent magnetic cylinder located outside the patient’s body and is turned by an electric motor, thereby generating a rotating magnetic field. This field in turn causes the motor, by a permanent magnetic sphere, located in a liquid bearing inside the endoscopy capsule, to rotate synchronously. The sphere induces an alternating voltage in coils inside the capsule and thus provides the necessary power. This arrangement was able to transmit power of up to 170 mW. This value is clearly higher than the minimum power of 100 mW needed to operate the electronics of the endoscopy capsules that are actually in use. The volume of both the sphere and the induction coil is smaller than the batteries that are actually integrated in the capsules. By this means, the operating time may be prolonged, in principle up to arbitrary values.

In this study, iris recognition under the influence of diabetes was investigated. A new database containing 1318 pictures from 343 irides – 546 images from 162 healthy irides (62% female users, 38% male users, 21% <20 years old, 61% (20) < 40 years old, 12% (40) <60 years old and 6% more than 60 years old) and 772 iris images from 181 diabetic eyes but with a clearly visible iris pattern (80% female users, 20% male users, 1% <20 years old, 17.5% (20) <40 years old, 46.5% (40) <60 years old and 35% more than 60 years old) – were collected. All of the diabetes-affected eyes had clearly visible iris patterns without any visible impairments and only type II diabetic patients with at least 2 years of being diabetic were considered for the investigation. Three different open source iris recognition codes and one commercial software development kit were used for achieving the iris recognition system’s performance evaluation results under the influence of diabetes. For statistical analysis, the t-test and the Kolmogorov-Simonov test were used.

The key for safe ultrasound (US)-guided punctures is a good visibility of the cannula. When using convex transducers for deep punctures, the incident angle between US beam and cannula varies along the cannula leading to a complex visibility pattern. Here, we present a method to systematically investigate the visibility throughout the US image. For this, different objective criteria were defined and applied to measurement series with varying puncture angles and depths of the cannula. It is shown that the visibility not only depends on the puncture angle but also on the location of the cannula in the US image when using convex transducers. In some image regions, an unexpected good visibility was observed even for steep puncture angles. The systematic evaluation of the cannula visibility is of fundamental interest to sensitise physicians to the handling of convex transducers and to evaluate new techniques for further improvement.

Image denoising stays be a standout amongst the primary issues in the field of image processing. Several image denoising algorithms utilizing wavelet transforms have been presented. This paper deals with the use of wavelet transform for magnetic resonance imaging (MRI) liver image denoising using selected wavelet families and thresholding methods with appropriate decomposition levels. Denoised MRI liver images are compared with the original images to conclude the most suitable parameters (wavelet family, level of decomposition and thresholding type) for the denoising process. The performance of our algorithm is evaluated using the signal-to-noise ratio (SNR), peak signal-to-noise ratio (PSNR) and mean square error (MSE). The results show that the Daubechies wavelet family of the tenth order with first and second of the levels of decomposition are the most optimal parameters for MRI liver image denoising.

The aim of this study was the quantitative evaluation of gastrointestinal cancer cell motility and 5-aminolevulinic acid (5-ALA)-induced fluorescence in vitro using mathematical morphology and structural analysis methods. The results of our study showed that MKN28 cells derived from the lymph node have the highest motility compared with AGS or HCT116 cells derived from primary tumors. Regions of single cells were characterized as most moving, and “tightly packed” cell colonies as nearly immobile. We determined the reduction of cell motility in late passage compared to early passage. Application of 5-ALA caused fluorescence in all investigated cells, and the fluorescence was different with regard to the cell type and application time. We observed higher fluorescence in MKN28 cells. Comprehensive image analysis did not reveal any statistically significant difference in fluorescence intensity between “tightly packed” cell regions, where nearly no motility was registered and loosely distributed cells, where the highest cell motility was registered. In conclusions, our study revealed that MKN28 cells derived from the lymph node have higher motility and 5-ALA-induced fluorescence than AGS or HCT116 derived from primary tumors. Moreover, image analysis based on a large amount of processed data is an important tool to study these tumor cell properties.

The aim of this study was to explore the effects of bone marrow mesenchymal stem cells (BMMSCs) and alendronate sodium (ALN) intervention on osteoporosis (OP). Sixty-eight 6-month-old healthy female Sprague Dawley (SD) rats were used to generate an OP model by removal of the ovaries. After 12 weeks, rats were treated with BMMSCs (BMMSC group) or ALN (ALN group) for 5 weeks. Serum type I collagen C terminal peptide (CTX_1), procollagen type I N-terminal propeptide (PINP), and bone alkaline phosphatase (BALP) were tested along with the femur bone density and other properties, including bone mineral density (BMD), BALP, percent trabecular area (BV/TV), trabecular thickness (Tb.Th), trabecular number (TbN), maximum load, maximum stress, maximum strain, and elastic modulus. BMD, BALP, BV/TV, Tb.Th, TbN, maximum load, maximum stress, maximum strain, and elastic modulus values were higher in the BMMSC group versus the ALN group relative to the control group (p < 0.05); CTX_1, PINP, trabecular separation (Tb.Sp), and osteoclast number (OC.N) were lowest in the BMMSC group versus the ALN group relative to the control group (p < 0.05). Both BMMSCs and ALN could improve the metabolic function and bone quality in osteoporotic mice while restoring the strength and toughness of bones. The intervention effects of BMMSCs are better than ALN in this model.

MEDLINE is a rapidly growing database; to utilize this resource, practitioners and biomedical researchers have dealt with tedious and time-consuming tasks such as discovering, searching, reading and evaluating of biomedical documents. However, making a label for a group of biomedical documents is expensive and needs a complicated operation. Otherwise, compound words, polysemous and synonymous problems can influence the search in MEDLINE. Therefore, designing an efficient way of sharing knowledge and information organization is essential so that information retrieval systems can provide ideal outcomes. For this purpose, different strategies are used in the retrieval of biomedical documents (RBD). However, still a number of unrelated results for the users’ query are obtained in the RBD process. Studies have shown that well-defined clusters in the retrieval system exhibit a more efficient performance in contrast to the document-based retrieval. Accordingly, the present study proposes the Expanding Statistical Language Modeling and Thesaurus (ESLMT) for clustering and retrieving biomedical documents. The results showed that Clustering with ESLM Similarity and Thesaurus (CESLMST) in all those criteria in this study have a higher value than the other compared methods. The results indicated that the mean average precision (MAP) has improved in the Clusters’ Retrieval Derived from ESLM Similarity-Query (CRDESLMS-QET) method in comparison to the previous methods with the Text REtrieval Conference (TREC) data set.