Denoising electroencephalogram signals using TQWT

Electroencephalographic (EEG) signals provide a valuable window into human brain activity, but their analysis is complicated by the presence of various types of noise and artifacts (electrical interference, ocular movements, muscle activity, etc.). Effective preprocessing is therefore essential to ensure the integrity of the information contained in EEG signals. In this context, we propose the use of the TQWT as an advanced filtering technique. TQWT enables flexible decomposition of the signal into frequency sub-bands by adjusting the Q-factor to optimize frequency and temporal resolution according to the spectral content. This approach facilitates the isolation of characteristic brain rhythms (delta, theta, alpha, beta, gamma), while effectively attenuating unwanted artifacts. Experimental results obtained from EEG recordings demonstrate that TQWT not only improves the signal-to-noise ratio but also preserves the fine morphology of the signals, which is crucial for applications in cognitive neuroscience and brain-computer interfaces. This study highlights the potential of TQWT as a powerful tool for adaptive EEG signal filtering.