Motor Hotspot Identification Based on Electroencephalography

Transcranial direct current stimulation (tDCS) is an emerging technique that applies DC current to the scalp to improve the motor function of patients with neurological disorders. The target location of tDCS is generally defined by the motor hotspot identified by using motor-evoked potential (MEP). However, the traditional MEP-based method needs use of transcranial magnetic stimulation (TMS), and it requires the empirical judgement. In this study, we propose a new motor hotspot detection method based on electroencephalography (EEG) acquired during a motor task. EEG data were measured from twelve subjects while they repeatedly performed a simple finger tapping task. Power spectral densities (PSDs) were extracted frin the EEG data as features, and a linear regression and artificial neural network (ANN) were used to estimate an individual motor hotspot location. The 3D coordinate information of individual motor hotspots detected by TMS were compared with those estimated by our proposed EEG-based approach. ANN outperformed a linear regression method in terms of the error distance, and a minimum distance between the motor hotspots identified by TMS and ANN was only 0.18 cm, demonstrating the feasibility of our proposed novel EEG-based motor hotspot detection method.