Analytical Decision Making from Clinical Data-Diagnosis and Classification of Epilepsy Risk Levels from EEG Signals-A Case Study

Abstract

Making decision under certainty is a perennial part of human predicament. In this research, we review the potential of fuzzy modeling technology as a tool for constructing customized decision making function for medical diagnosis, such as classification of epilepsy risk levels. We investigate the optimization of fuzzy outputs in the classification of epilepsy risk levels from EEG (Electroencephalogram) signals using soft decision as post classifier. The fuzzy techniques are applied as a first level classifier to classify the risk levels of epilepsy based on extracted parameters like energy, variance, peaks, sharp and spike waves, duration, events and covariance from the EEG signals of the patient. Soft Decision Trees are identified as post classifiers on the classified data to obtain the optimized risk level that characterizes the patient's epilepsy risk level. The Performance Index (PI) and Quality Value (QV) are calculated for the above methods. A group of ten patients with known epilepsy findings are used in this study. High PI such as 97.87 % was obtained at QV's of 23.31, SDT optimization when compared to the value of 40% and 6.25 through fuzzy techniques respectively. The performance of SDT is lucidly presented in this work. SDT results are validated with standard Multi Layer Perceptrons (MLP) neural networks.