Microprocessor based detection of epileptic discharges

Abstract

Parametric and Non-parametric methods have been developed for the detection and interpretation of EEG data for normal and abnormal patients. These methods have been implemented on mainframe computers or dedicated microcomputers. The heuristic methods are suitable for implement on dedicated microprocessor based system as they involve less degree of computation in comparision to the parametric methods. In this work a microprocessor based system has been developed and heuristic pattern recognition technique has been applied, which is based on the measurement of amplitude, duration slopes etc. for the detection of spike and sharp waves as well as the different frequency band of background activity. The computed values of amplitude and durations are shown on the graphs from which the different symptoms based on EEG are determined. © 1991 Kluwer Academic Publishers.