Applying Independent Component Analysis on ECG Cancellation Technique for the Surface Recording of Trunk Electromyography

Surface electromyography (sEMG) recorded from the trunk area may reflect underlying muscular function, and is the current standard for in vivo functional examination. However, sEMG of this area, including the low back musculature, usually encounters substantial interference from strong cardiac signals. It is therefore imperative to remove electrocardiogram (ECG) interference from sEMG data. This paper discusses a denoise method using independent component analysis (ICA) and a high-pass filter to effectively suppress the interference of ECG in sEMG recorded from trunk muscles. The performance of this technique was evaluated with simulation experiments. To compare the outcome of the ICA and filtering technique to the original sEMG signal, correlation coefficients in both time-domain waveform and frequency spectrum were computed. In addition, different filter bands were evaluated. The ICA ECG cancellation with a 30 Hz high-pass filter showed higher mean correlation coefficients in the time domain (0.97plusmn0.08) and in the frequency spectrum (0.99plusmn0.06) than any other techniques. This suggests that the ICA ECG cancellation technique with a 30 Hz high-pass filter would be the most appropriate method to extract useful sEMG signals from trunk muscles