NEURAL-NETWORK-BASED ADAPTIVE MATCHED FILTERING FOR QRS DETECTION

We have developed an adaptive matched filtering algorithm based upon a n artificial neural network (ANN) for QRS detection. We use an ANN ada ptive whitening filter to model the lower frequencies of the ECG which are inherently nonlinear and nonstationary. The residual signal which contains mostly higher frequency QRS complex energy is then passed th rough a linear matched filter to detect the location of the QRS comple x. We developed an algorithm to adaptively update the matched filter t emplate from the detected QRS complex in the ECG signal itself so that the template can be customized to an individual subject. This ANN whi tening filter is very effective at removing the time-varying, nonlinea r noise characteristic of ECG signals. Using this novel approach, the detection rate for a very noisy patient record in the MIT/BIH arrhythm ia database is 99.5%, which compares favorably to the 97.5% obtained u sing a linear adaptive whitening filter and the 96.5% achieved with a bandpass filtering method.