12 Lead QRS Window Detection: Using Feature Extraction and Statistical Parameters

A 12-lead Electrocardiogram (ECG) is the fundamental clinical test that a doctor uses to check for any abnormalities in the cardiac muscle. The QRS complex has relatively higher energy and is the most prominent component of an ECG signal. This paper proposes a novel robust technique to extract a window signal that contains this QRS complex of a noisy 12 lead Electrocardiogram (ECG). Algorithms for further automated analysis depend on the correct detection of the QRS complex. An ECG signal is processed to create a feature signal with relatively high amplitudes in the QRS complex region. Statistical parameters and multiple stages of different thresholds are used in combination with conventional feature enhancement techniques to identify correct QRS complexes from any false positives.