Wavelet-Based Muscle Artefact Noise Reduction for Short Latency rTMS Evoked Potentials

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Repetitive transcranial magnetic stimulation (rTMS) is a therapeutic tool, but EEG analysis is hindered by noise.
• Stimulus artifacts and muscle potentials obscure early cortical evoked potentials (within 20ms post-stimulus).

Purpose of the Study:

• To develop and validate a novel noise reduction method for EEG signals during rTMS.
• To enhance the analysis of cortical evoked potentials by removing artifacts.

Main Methods:

• Utilized wavelet transform with a fourth-order Daubechies mother wavelet.
• Implemented a novel coefficient reduction algorithm based on cortical amplitude thresholds.
• Compared performance against other wavelets (Coiflets 2 & 3) and digital filtering.

Main Results:

• The proposed wavelet transform method effectively reduced noise in EEG signals.
• Coiflet 2 and 3 wavelets also showed effectiveness, with Coiflet 3 performing marginally better than Daubechies 4.
• The method was validated on data from 16 subjects under clinically relevant rTMS conditions.

Conclusions:

• The novel wavelet-based noise reduction technique significantly improves EEG signal quality during rTMS.
• This method facilitates more accurate analysis of cortical evoked potentials, crucial for understanding brain responses and optimizing rTMS therapy.