Adaptive cancellation techniques for noise reduction in electrocardiography

Mukalaf, A

Abstract

This thesis describes an investigation of the application of adaptive filtering techniques in electrocardiography (ECG), with particular reference to exercise testing. The main objective in this study was to observe the effectiveness of adaptive methods for noise reduction. A number of techniques, including averaging, were tested for comparison and the limitation and inadequacy of these techniques in noise reduction were used as the basis for extending the application of adaptive filtering techniques.
The theory and application of adaptive filtering relating to noise reduction in ECG has been developed, based upon the adaptive transversal filter using the Widrow-Hoff algorithm. The application of a new version of the adaptive filter, referred to the minimal time-sequence adaptive filter to enhance the ECG, was tested in order to improve cancellation with less distortion.
The coherence function was studied as the basis for selecting electrode placement for particular adaptive filtering applications. The adaptive filtering technique was found to be an appropriate method for noise reduction and can improve the signal-to noise ratio by up to 15dB. It was also found, through the assessment of the results presented from 22 patients, that the use of adaptive filters in the exercise ECG is appropriate, after the removal of a d.c. variation.

Citation

Mukalaf, A. (1987). Adaptive cancellation techniques for noise reduction in electrocardiography