Frequency Correlated Filtering for Ambulatory Electrocardiograms

Medical tendencies target nowadays at miniaturization and portability of noninvasive diagnosis devices. Analyzed signals captured outside the clinical environment are sensitive to external interferences and noise. A method based on the frequency dependence of the signal has been developed. The algorithm analyzes electrocardiogram data using wavelets and an exponentially growing thresholding algorithm. The results are promising and display after denoising more accurate electrocardiogram features.