Dear colleague,

When we start to think about it. Is removing noise from ECGs reasonable in the first place? For sure there is a signal from chest muscles associated with breathing and signals from muscles due to movements but simultaneously there are but-s.

How do we know that signal that we call noise is not an integral part of ECG recordings? We do not. There is no effective method to discern different sources of noise except parallel recording of breathing activity and muscle movements. Even this does not discern signals perfectly.

Instead of removing noise, it might be very helpful to use methods that are robust against noise in the signal. All depends on data that you would like to extract from ECG recordings. This is, of course, a topic for a completely different question.

All the best,

Jiri