Accurate performance metrics for removing noise from the electrocardiogram (ECG) are difficult to define due to the inherently complicated nature of the noise and the absence of knowledge about the underlying dynamical processes. By using a previously published model for generating realistic artificial ECG signals and adding both stochastic and deterministic noise, a method for assessing the performance of noise reduction techniques is presented. Independent component analysis (ICA) and nonlinear noise reduction (NNR) are employed to remove noise from an ECG with known characteristics. Performance as a function of the signal to noise ratio is measured by both a noise reduction factor and the correlation between the cleaned signal and the original noise-free signal.
Patrick E. McSharry,
Gari D. Clifford,
"A comparison of nonlinear noise reduction and independent component analysis using a realistic dynamical model of the electrocardiogram", Proc. SPIE 5467, Fluctuations and Noise in Biological, Biophysical, and Biomedical Systems II, (25 May 2004); doi: 10.1117/12.548726; https://doi.org/10.1117/12.548726
Patrick E. McSharry, Gari D. Clifford, "A comparison of nonlinear noise reduction and independent component analysis using a realistic dynamical model of the electrocardiogram," Proc. SPIE 5467, Fluctuations and Noise in Biological, Biophysical, and Biomedical Systems II, (25 May 2004);