On the use of ECG waveforms for health status estimation

Abstract

We present a system that uses ECG time series for reaching decisions regarding the user’s health status. ECG waveforms can provide us with certain features that have diagnostic value for several cases. The system consists of two modules. The first performs the analysis of the ECG data so as to extract the aforementioned features, and the second is a hybrid intelligence system combining neural and neurofuzzy networks that uses these features to reach decisions according to certain rules provided by medical experts. The output of the system is a classification of the user’s health state to certain alert levels. The system presented here can be integrated in an artificial agent monitoring a user and deciding for his health state.