A vision model automates the sorting of noisy EEG signals
Cleaning up an EEG normally takes hours of manual inspection by an expert to figure out which components are real brain signal and which are muscle or eye-movement noise. A computer vision system now does this sorting instead of the specialist, opening the door to large-scale EEG studies and near real-time clinical use.
practitioners › 89.45% accuracy, processing 7,200x faster than manual inspection, compatible with ICLabel and EEGLab.
arXiv:2607.21654 (q-bio.NC), ICMLA 2024 →