Noise cancellation in brain waves using a new diffusion normalized least power based algorithm for brain computer interface applications

Abstract

In remote health care monitoring applications, continuous brain signal monitoring is a key task. While recording electroencephalogram (EEG) signals in clinical environments various artifacts like respiration artifact (RA), electromyogram (EMG) or electro muscle artifact and electrode motion artifact (EMA) effects the original brain activity plot. In this paper new adaptive noise canceler algorithm is proposed to improve EEG signal quality in clinical scenario. Diffusion normalized least power algorithm (DNLMP) is one of the better versions of adaptive normalization techniques. Due to advantages of normalization a new robust diffusion algorithm is implemented for noise cancellation experiments. Here error signal is considered with artifacts signal and it is diminished by using normalization factor of the weight recursion. The performance of the developed noise cancellers is tested on real EEG signals with artifacts. As computational complexity is an important parameter in practical applications, we combined robust DNLMP with signum based techniques to achieve lesser computational complexity. Experimental results confirmed that proposed realizations perform better than the counter parts s in terms of excess mean square error (EMSE), signal to noise ratio improvement (SNRI) and computational complexity.

Keywords

• Adaptive noise cancellers
• Artifacts
• Brain waves
• Health care monitoring
• NLMP algorithm