Analysis of a Differential Noise Detection Filter in ECG Signals

Noise detection presents a big challenge in wearable heart monitor technology and beat detection. The distorted signal is hard to interpret and as a consequence, valuable information may be lost. In this paper, we present our research for developing a noise detection filter based on a differential filter. Our analysis offers in-depth evaluation of the optimal values of two critical parameters - window size and Signal-to-Noise threshold. Our final goal is to minimize errors in QRS detection and beat classification caused by noise-distorted data.