ECG signal denoising using noise invalidation

This paper presents an ECG denoising scheme based on the previously introduced noise invalidation method [1]. The method provides an adaptive data dependent threshold which can be used to denoise ECG signals in an orthogonal basis of choice, by discarding coefficients that are smaller than the proposed threshold. Performance of the noise invalidation method is evaluated using mean square error by denoising a set of ECG signals corrupted by various types of noise, which usually affect ECG signal quality in practice.