Adaptive Gaussian filtering in routine ECG/VCG analysis

Different types of filters are used in systems for the analysis of electrocardiograms and/or vectorcardiograms. These filters are mostly locally used to enhance the relevant signal components in order to improve the performance of one or more subtasks or modules rather than to provide filtered signals to be used in the feature extraction and classification part of the system. Filters are also used in the module dealing with the recognition of onsets and endpoints of the major components in the ECG. This paper reviews the theoretical aspects of an adaptive Gaussian filter 1, and describes the modifications that were necessary to make this filter suitable for application in the waveform-recognition module. The efficiency of the filter is demonstrated with the aid of an internationally validated database of ECG's and VCG's.