Suppression of impulse noise using adaptive filters

The electrical activity of muscles is usually modeled using Gaussian probability distribution function. Such assumption is not always true, because real-life muscle noise has impulse character as well. Adaptive threshold technique is one important step in the detection of QRS Complexes. However, presence of impulse noise may trigger false positives. The proposed system uses adaptive filters to suppress impulse noise in ECG signal and avoid the false positives. It was observed that despite the noise, the algorithm identifies the system and adjusts coefficients accordingly. Significant improvement in Signal-to-Noise ratio is achieved after suppressing impulse noise.