Noise Reduction of 15-lead Electrocardiogram Signals Using Signal Processing Algorithms

Abstract

In the research, a signal processing algorithm is proposed to reduce the noise and baseline wander in ECG spectra. The ECG spectra are first processed by symmetrical triangle smooth (can also be considered as a low pass filter) to reduce the noise, then a FIR high pass filter is applied to decrease the baseline wander. The ROC curves of the same classification test for myocardial infarction using ECG spectra processed by the standard and proposed method are plotted. The area under the curve can be improved up to 29.3%, which indicates the proposed ECG spectra can provide a more accuracy diagnostic result for myocardial infarction.