Wenwei Yu

Application of recurrence quantification analysis for the automated identification of epileptic EEG signals.

Epilepsy is a common neurological disorder that is characterized by the recurrence of seizures. Electroencephalogram (EEG) signals are widely used to diagnose seizures. Because of the non-linear and dynamic nature of the EEG signals, it is difficult to effectively decipher the subtle changes in these signals by visual inspection and by using linear techniques. Therefore, non-linear methods are being researched to analyze the EEG signals. In this work, we use the recorded EEG signals in Recurrence Plots (RP), and extract Recurrence Quantification Analysis (RQA) parameters from the RP in order to classify the EEG signals into normal, ictal, and interictal classes. Recurrence Plot (RP) is a graph that shows all the times at which a state of the dynamical system recurs. Studies have reported significantly different RQA parameters for the three classes. However, more studies are needed to develop classifiers that use these promising features and present good classification accuracy in differentiating the three types of EEG segments. Therefore, in this work, we have used ten RQA parameters to quantify the important features in the EEG signals.These features were fed to seven different classifiers: Support vector machine (SVM), Gaussian Mixture Model (GMM), Fuzzy Sugeno Classifier, K-Nearest Neighbor (KNN), Naive Bayes Classifier (NBC), Decision Tree (DT), and Radial Basis Probabilistic Neural Network (RBPNN). Our results show that the SVM classifier was able to identify the EEG class with an average efficiency of 95.6%, sensitivity and specificity of 98.9% and 97.8%, respectively.

Application of Higher Order Spectra for the Identification of Diabetes Retinopathy Stages

Diabetic retinopathy (DR) is a condition where the retina is damaged due to fluid leaking from the blood vessels into the retina. In extreme cases, the patient will become blind. Therefore, early detection of diabetic retinopathy is crucial to prevent blindness. Various image processing techniques have been used to identify the different stages of diabetes retinopathy. The application of non-linear features of the higher-order spectra (HOS) was found to be efficient as it is more suitable for the detection of shapes. The aim of this work is to automatically identify the normal, mild DR, moderate DR, severe DR and prolific DR. The parameters are extracted from the raw images using the HOS techniques and fed to the support vector machine (SVM) classifier. This paper presents classification of five kinds of eye classes using SVM classifier. Our protocol uses, 300 subjects consisting of five different kinds of eye disease conditions. We demonstrate a sensitivity of 82% for the classifier with the specificity of 88%.

Automated identification of normal and diabetes heart rate signals using nonlinear measures

Diabetes mellitus (DM) affects considerable number of people in the world and the number of cases is increasing every year. Due to a strong link to the genetic basis of the disease, it is extremely difficult to cure. However, it can be controlled to prevent severe consequences, such as organ damage. Therefore, diabetes diagnosis and monitoring of its treatment is very important. In this paper, we have proposed a non-invasive diagnosis support system for DM. The system determines whether or not diabetes is present by determining the cardiac health of a patient using heart rate variability (HRV) analysis. This analysis was based on nine nonlinear features namely: Approximate Entropy (ApEn), largest Lyapunov exponet (LLE), detrended fluctuation analysis (DFA) and recurrence quantification analysis (RQA). Clinically significant measures were used as input to classification algorithms, namely AdaBoost, decision tree (DT), fuzzy Sugeno classifier (FSC), k-nearest neighbor algorithm (k-NN), probabilistic neural network (PNN) and support vector machine (SVM). Ten-fold stratified cross-validation was used to select the best classifier. AdaBoost, with least squares (LS) as weak learner, performed better than the other classifiers, yielding an average accuracy of 90%, sensitivity of 92.5% and specificity of 88.7%.

EMG prosthetic hand controller discriminating ten motions using real-time learning method

We discuss the necessity of a learning mechanism for an EMG prosthetic hand controller, and the real-time learning method is proposed and designed. This method divides the controller into three units. The analysis unit extracts useful informations for discriminating motions from the EMG. The adaptation unit learns the relation between EMG and control command and adapts operators characteristics. The trainer unit makes the adaptation unit learn in real-time. Experiments show that the proposed controller discriminates ten forearm motions, which contain four wrist motions and six hand motions, and learns within 4/spl sim/25 minutes. The average of the discriminating rate is 91.5%.

EMG automatic switch for FES control for hemiplegics using artificial neural network

Abstract Functional Electrical Stimulation (FES) is an effective and developing method to restore functions for paraplegic patients. In this research, we focused on the switching problem of FES, which is one of the obstacles that prevent FES from further practical uses. An adaptive switching for FES control for the lower limbs’ activities of hemiplegic patients was developed, based on the consideration that, lower limbs’ activities need the synchronization of limbs of both sides. Electromyogram (EMG) signals detected from normal side of hemiplegic patients were used to recognize the activities that the patients intend to do. The recognition results were utilized as the switching signals. However, motion patterns represented and analyzed by EMG are distinctive of individual variations and characteristic alternation, which inevitably result in classification errors in EMG analyzing. To overcome these problems, a feed-forward artificial neural network (ANN) was embedded in an on-line process to form an analyzing system that can adapt to individual characteristics and trace the nonstationary factor. Furthermore, in order to enable the analyzing system to recognize right timings from the EMG-described dynamical processes of activities, such as standing-up and walking, a practical training-set construction method that utilizes additional reference data was proposed. The proposed switching system was applied to a FES system that supports standing-up and walking for a hemiplegics subject, to verify the effectiveness.

Mutual Adaptation in a Prosthetics Application

Prosthetic care for handicapped persons requires new and reliable robotics technology. In this paper, developmental approaches for prosthetic applications are described. In addition, the challenges associated with the adaptation and control of materials for human hand prosthetics are presented. The new technology of robotics for prosthetics provides many possibilities for the detection of human intention. This is particularly true with the use of electromyogram (EMG) and mechanical actuation with multiple degrees of freedom. The EMG signal is a nonlinear wave, and has time dependency and big individual differences. The EMG signal is a nonlinear wave that has time dependency and significant differences from one individual to another. A method for how an individual adapts to the processing of EMG signals is being studied to determine and classify a human’s intention to move. A prosthetic hand with 11 degrees of freedom (DOF) was developed for this study. In order to make it light-weight, an adaptive joint mechanism was applied. The application results demonstrate the challenges for human adaptation. The f-MRI data show a process of replacement from a phantom limb image to a prosthetic hand image.

Psycho-physiological assessment of a prosthetic hand sensory feedback system based on an auditory display: a preliminary study

BackgroundProsthetic hand users have to rely extensively on visual feedback, which seems to lead to a high conscious burden for the users, in order to manipulate their prosthetic devices. Indirect methods (electro-cutaneous, vibrotactile, auditory cues) have been used to convey information from the artificial limb to the amputee, but the usability and advantages of these feedback methods were explored mainly by looking at the performance results, not taking into account measurements of the user’s mental effort, attention, and emotions. The main objective of this study was to explore the feasibility of using psycho-physiological measurements to assess cognitive effort when manipulating a robot hand with and without the usage of a sensory substitution system based on auditory feedback, and how these psycho-physiological recordings relate to temporal and grasping performance in a static setting.Methods10 male subjects (26+/-years old), participated in this study and were asked to come for 2 consecutive days. On the first day the experiment objective, tasks, and experiment setting was explained. Then, they completed a 30 minutes guided training. On the second day each subject was tested in 3 different modalities: Auditory Feedback only control (AF), Visual Feedback only control (VF), and Audiovisual Feedback control (AVF). For each modality they were asked to perform 10 trials. At the end of each test, the subject had to answer the NASA TLX questionnaire. Also, during the test the subject’s EEG, ECG, electro-dermal activity (EDA), and respiration rate were measured.ResultsThe results show that a higher mental effort is needed when the subjects rely only on their vision, and that this effort seems to be reduced when auditory feedback is added to the human-machine interaction (multimodal feedback). Furthermore, better temporal performance and better grasping performance was obtained in the audiovisual modality.ConclusionsThe performance improvements when using auditory cues, along with vision (multimodal feedback), can be attributed to a reduced attentional demand during the task, which can be attributed to a visual “pop-out” or enhance effect. Also, the NASA TLX, the EEG’s Alpha and Beta band, and the Heart Rate could be used to further evaluate sensory feedback systems in prosthetic applications.

Application of infrared thermography in computer aided diagnosis

Abstract The invention of thermography, in the 1950s, posed a formidable problem to the research community: What is the relationship between disease and heat radiation captured with Infrared (IR) cameras? The research community responded with a continuous effort to find this crucial relationship. This effort was aided by advances in processing techniques, improved sensitivity and spatial resolution of thermal sensors. However, despite this progress fundamental issues with this imaging modality still remain. The main problem is that the link between disease and heat radiation is complex and in many cases even non-linear. Furthermore, the change in heat radiation as well as the change in radiation pattern, which indicate disease, is minute. On a technical level, this poses high requirements on image capturing and processing. On a more abstract level, these problems lead to inter-observer variability and on an even more abstract level they lead to a lack of trust in this imaging modality. In this review, we adopt the position that these problems can only be solved through a strict application of scientific principles and objective performance assessment. Computing machinery is inherently objective; this helps us to apply scientific principles in a transparent way and to assess the performance results. As a consequence, we aim to promote thermography based Computer-Aided Diagnosis (CAD) systems. Another benefit of CAD systems comes from the fact that the diagnostic accuracy is linked to the capability of the computing machinery and, in general, computers become ever more potent. We predict that a pervasive application of computers and networking technology in medicine will help us to overcome the shortcomings of any single imaging modality and this will pave the way for integrated health care systems which maximize the quality of patient care.

Study on the Effects of Electrical Stimulation on the Pattern Recognition for an EMG Prosthetic Application

The need of biofeedback in man-machine interfaces is of vital importance for the development of subconscious control with external devices. In order to obtain extended proprioception, in other words, to include the external devices into the body schema, we need to provide with more feedback channels to the human body. In this study we look into the use of electrical stimulation as biofeedback and its effects over the pattern recognition process from the EMG signals that controls the hand movements

On-line supervising mechanism for learning data in surface electromyogram motion classifiers

This paper proposes a mechanism that supervises the learning data set for the classification from electromyogram to forearm motion. The supervising mechanism contains automatic data addition and automatic data elimination processes. It also contains manual data addition that is the same as our former algorithm. Both the automatic addition and elimination processes evaluate success or failure of classification from the continuity of the classifiers outputs. These processes assume that a person cannot change his motion within a certain interval. In experiments, a system with the supervising mechanism embedded classifies ten forearm motions from two channels of the electromyogram. First, this paper makes it clear that the proposed mechanism is effective by comparison with six settings, including the absence of a supervising pattern. Next, it is verified that the system can adapt to alteration of the operators characteristics by a sensor-shifting test in which we move one sensor after the operators training. From the experimental results, it is concluded that the proposed mechanism can adjust the generated decision boundaries for improvement of classification ability, and in addition is capable of tracking the alteration of the operators characteristics through time.