MINING OF ECG SIGNAL FOR NEW DIAGNOSTIC INFORMATION

This paper investigates a technique, which extracts new features having potentiality for giving more discriminatory clues from simple ECG (Electrocardiogram) signals. The extracted feature parameters, Average RR Signal Morphology (ARSM) and RR Interval Frequency Histogram (RIFH) of an ECG signal from pre-selected data segment of the MIT-database shows that they may be utilized to detect some clinicallychallenging problems like Paroxysmal disease of Heart. For checking the efficiency of a Multidimensional- feature parameter like thosementioned above, we have employed a novel two dimensional display method. Overall, it was shown that the same technique may be employedfor comparative study of new as well as old feature parameters extracted from biomedical images and signals