Matching a wavelet to ECG signal.

In this paper we develop an approach to synthesize a wavelet that matches the ECG signal. Matching a wavelet to a signal of interest has potential advantages in extracting signal features with greater accuracy, particularly when the signal is contaminated with noise. The approach that we have taken is based on the theoretical work done by Chapa and Rao. We have applied their technique to a noise-free ECG signal representing one cardiac cycle. Results indicate that a matched wavelet, that was able to capture the broad ECG features, could be obtained. Such a wavelet could be used to extract ECG features such as QRS complexes and P&T waves with greater accuracy.