Subhagata Chattopadhyay

Automated diagnosis of epileptic EEG using entropies

Abstract Epilepsy is a neurological disorder characterized by the presence of recurring seizures. Like many other neurological disorders, epilepsy can be assessed by the electroencephalogram (EEG). The EEG signal is highly non-linear and non-stationary, and hence, it is difficult to characterize and interpret it. However, it is a well-established clinical technique with low associated costs. In this work, we propose a methodology for the automatic detection of normal , pre-ictal , and ictal conditions from recorded EEG signals. Four entropy features namely Approximate Entropy ( ApEn ), Sample Entropy ( SampEn ), Phase Entropy 1 ( S1 ), and Phase Entropy 2 ( S2 ) were extracted from the collected EEG signals. These features were fed to seven different classifiers: Fuzzy Sugeno Classifier (FSC), Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Probabilistic Neural Network (PNN), Decision Tree (DT), Gaussian Mixture Model (GMM), and Naive Bayes Classifier (NBC). Our results show that the Fuzzy classifier was able to differentiate the three classes with a high accuracy of 98.1%. Overall, compared to previous techniques, our proposed strategy is more suitable for diagnosis of epilepsy with higher accuracy.

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.

AUTOMATED DIAGNOSIS OF NORMAL AND ALCOHOLIC EEG SIGNALS

Electroencephalogram (EEG) signals, which record the electrical activity in the brain, are useful for assessing the mental state of a person. Since these signals are nonlinear and non-stationary in nature, it is very difficult to decipher the useful information from them using conventional statistical and frequency domain methods. Hence, the application of nonlinear time series analysis to EEG signals could be useful to study the dynamical nature and variability of the brain signals. In this paper, we propose a Computer Aided Diagnostic (CAD) technique for the automated identification of normal and alcoholic EEG signals using nonlinear features. We first extract nonlinear features such as Approximate Entropy (ApEn), Largest Lyapunov Exponent (LLE), Sample Entropy (SampEn), and four other Higher Order Spectra (HOS) features, and then use them to train Support Vector Machine (SVM) classifier of varying kernel functions: 1st, 2nd, and 3rd order polynomials and a Radial basis function (RBF) kernel. Our results indicate that these nonlinear measures are good discriminators of normal and alcoholic EEG signals. The SVM classifier with a polynomial kernel of order 1 could distinguish the two classes with an accuracy of 91.7%, sensitivity of 90% and specificity of 93.3%. As a pre-analysis step, the EEG signals were tested for nonlinearity using surrogate data analysis and we found that there was a significant difference in the LLE measure of the actual data and the surrogate data.

Comparing Fuzzy-C Means and K-Means Clustering Techniques: A Comprehensive Study

Clustering techniques are unsupervised learning methods of grouping similar from dissimilar data types. Therefore, these are popular for various data mining and pattern recognition purposes. However, their performances are data dependent. Thus, choosing right clustering technique for a given dataset is a research challenge. In this paper, we have tested the performances of a Soft clustering (e.g., Fuzzy C means or FCM) and a Hard clustering technique (e.g., K-means or KM) on Iris (150 x 4); Wine (178 x 13) and Lens (24 x 4) datasets. Distance measure is the heart of any clustering algorithm to compute the similarity between any two data. Two distance measures such as Manhattan (MH) and Euclidean (ED) are used to note how these influence the overall clustering performance. The performance has been compared based on seven parameters: (i) sensitivity, (ii) specificity, (iii) precision, (iv) accuracy, (v) run time, (vi) average intra cluster distance (i.e. compactness of the clusters) and (vii) inter cluster distance (i.e. distinctiveness of the clusters). Based on the experimental results, the paper concludes that both KM and FCM have performed well. However, KM outperforms FCM in terms of speed. FCM-MH combination produces most compact clusters, while KM-ED yields most distinct clusters.

Neural Network Approaches to Grade Adult Depression

Depression is a common but worrying psychological disorder that adversely affects one’s quality of life. It is more ominous to note that its incidence is increasing. On the other hand, screening and grading of depression is still a manual and time consuming process that might be biased. In addition, grades of depression are often determined in continuous ranges, e.g., ‘mild to moderate’ and ‘moderate to severe’ instead of making them more discrete as ‘mild’, ‘moderate’, and ‘severe’. Grading as a continuous range is confusing to the doctors and thus affecting the management plan at large. Given this practical issue, the present paper attempts to differentiate depression grades more accurately using two neural net learning approaches—‘supervised’, i.e., classification with Back propagation neural network (BPNN) and Adaptive Network-based Fuzzy Inference System (ANFIS) classifiers, and ‘unsupervised’, i.e., ‘clustering’ technique with Self-organizing map (SOM), built in MATLAB 7. The reason for using the supervised and unsupervised learning approaches is that, supervised learning depends exclusively on domain knowledge. Supervision may induce biasness and subjectivities related to the decision-making. Finally, the performance of BPNN and ANFIS are compared and discussed. It was observed that ANFIS, being a hybrid system, performed much better compared to the BPNN classifier. On the other hand, SOM-based clustering technique is also found useful in constructing three distinct clusters. It also assists visualizing the multidimensional data clusters into 2-D.

A survey and comparative study on the instruments for glaucoma detection

Glaucoma is one of the leading causes of irreversible blindness worldwide. It has been proposed that the intraocular pressure is a causative factor in the development of glaucoma, which is an optic neuropathy. This paper surveys the use of tonometers, gonioscopes, optical coherence tomographs, scanning laser polarimeters, scanning laser ophthalmoscopes (also known as scanning laser tomographs) and corneal pachymeters for the diagnosis and management of glaucoma. The working mechanisms as well as the comparative advantages and disadvantages of each of these instruments are presented.

Some studies on fuzzy clustering of psychosis data

Clustering is a well-known method of data mining, which aims at extracting useful information from a data set. Clusters could be either crisp (having well-defined boundaries) or fuzzy (with vague boundaries) in nature. The present paper deals with fuzzy clustering of psychosis data. A set of statistically generated psychosis data are clustered using Fuzzy C-Means (FCM) algorithm and entropy-based method and its proposed extensions. From the clusters, we finally decide on patient distributions response-wise. Comparisons are made of the above algorithms, in terms of quality of clusters made and their computational complexity. Finally, the multidimensional best set of clusters are mapped into 2-D for visualisation, using a Self-Organising Map (SOM).

A framework for assessing ICT preparedness for e-health implementations

Electronic health (e-health) is probably one of the most significant contributions of Information Communication Technology (ICT) in present daypsilas healthcare. ICT efficiently bridges healthcare sector and technology for e-health implementation, which is a costly affair due to involvement of considerable amount of planning and investment. Preparedness, on the other hand can be defined as a state of readiness prior taking any action and applicable for implementing any e-health project. Study of preparedness essentially i) renders insight to the existing resources, ii) specifies the requirements for successful implementation of a project, and iii) helps set up strategies for the said implementation. Preparedness may be assessed at various levels of e-health implementations, such as ICT, Application, Service, Process and Government or Organizational levels. The present work focuses at the very initial level i.e. ICT and proposes an ICT-preparedness-framework for e-health implementation. The proposed ICT-preparedness framework is a conceptual one and is based on two different applications - A) connected graph-based approach to capture and in turn quantify some of the ICT constructs (Hardware, Connectivity, Software and Skills) and their respective indicators and B) a fuzzy set-based technique to assess the preparedness levels of these constructs. Finally the framework is discussed with an e-health scenario on Tele-cardiology.

A Novel Mathematical Approach to Diagnose Premenstrual Syndrome

Diagnosis of Premenstrual syndrome (PMS) is a research challenge due to its subjective presentation. An undiagnosed PMS case is often termed as ‘borderline’ (‘B’) that further add to the diagnostic fuzziness. This study proposes a methodology to diagnose PMS cases using a combined knowledge engineering and soft computing techniques. According to the guidelines of American College of Gynecology (ACOG), ten symptoms have been selected and technically processed for 50 cases each having class labels—‘B’ or ‘NB’ (not borderline) using domain expertise. Any Attribute that fails normality test has been excluded from the study. Decision tree (DT) has then been induced in obtaining the initial class boundaries and mining the important Attributes to classify PMS cases. Prior doing so, the best split criterion has been set using the maximum information gain measure. Initial information about classification boundaries are finally used to measure fuzzy membership values and the corresponding firing strengths have been measured for final classification of PMS ‘B’ cases.

Application of Bayesian Classifier for the Diagnosis of Dental Pain

Toothache is the most common symptom encountered in dental practice. It is subjective and hence, there is a possibility of under or over diagnosis of oral pathologies where patients present with only toothache. Addressing the issue, the paper proposes a methodology to develop a Bayesian classifier for diagnosing some common dental diseases (D = 10) using a set of 14 pain parameters (P = 14). A questionnaire is developed using these variables and filled up by ten dentists (n = 10) with various levels of expertise. Each questionnaire is consisted of 40 real-world cases. Total 14*10*10 combinations of data are hence collected. The reliability of the data (P and D sets) has been tested by measuring (Cronbach’s alpha). One-way ANOVA has been used to note the intra and intergroup mean differences. Multiple linear regressions are used for extracting the significant predictors among P and D sets as well as finding the goodness of the model fit. A naïve Bayesian classifier (NBC) is then designed initially that predicts either presence/absence of diseases given a set of pain parameters. The most informative and highest quality datasheet is used for training of NBC and the remaining sheets are used for testing the performance of the classifier. Hill climbing algorithm is used to design a Learned Bayes’ classifier (LBC), which learns the conditional probability table (CPT) entries optimally. The developed LBC showed an average accuracy of 72%, which is clinically encouraging to the dentists.