The goals of this study are (1) to evaluate the rate of stereoEEG brain-computer interface (BCI) classification accuracy and (2) to collect the dataset of neuronal signals recorded from stereoEEG electrodes during motor performance, motor imagery or brain-computer interface control. The study enrolls hospitalised patients suffering from resistant epilepsy with already implanted intracranial stereoEEG electrodes for medical reasons (i.e. for preoperative localization of the epileptogenic foci). The number and location of electrodes are determined solely for the clinical purposes of stereoEEG monitoring and are not related to the protocol of the current study. After obtaining informed consent to participate in the study, each patient will participate in one experimental session lasting no more than 60 minutes, recording brain signals associated with hand movement, motor imagery, and BCI control. All tasks and instructions presented during the study session are not pro-epileptogenic and cannot provoke an epileptic attack. The experiments will take place in the patient's room, without interruption of observation by the department's medical staff. The data recorded in this study will be used to improve or develop new algorithms for decoding motor signals from deep brain structures for their potential use in invasive BCIs.