Fig. 1.

System introduction. A The flow of proposed system. While the system is operated, the patients are instructed to imagine themselves performing lower-limb movements. The EEG cap receives the raw EEG signals of the patient, which are then transmitted to a computer for analysis via Bluetooth. Then, the computer sends the control command to the lower-limb robot, and the robot conducts lower-limb training for the SCI patients. The robot has an angle sensor and sends the crank angle information to the display when running. The crank angle on the screen matches the actual crank angle. B The system includes an EEG cap with 8 electrodes, an EEG collector, a computer, a monitor, and a lower-limb driving robot (maximum torque: 12 N m; height: 0.9–1.2 m; crank length: 0.1 m; rotation speed: 0–100r/min). C Robot usage and game display interface. The patient is asked to lie down on a bed and their lower limbs are fixed to the robot, and the moving parts of the system drive the patient’s lower limb in a pedaling motion, giving muscle and sensory feedback to the patient. The monitor displays a riding game to simulate actual pedaling. The speed of the pedaling movement in the riding game is consistent with that of the patient. The game shows the remaining training time (top center panel), virtual riding distance (top right panel), speed (bottom center panel), muscle strength value (bottom panel, to the left of “speed”), and motor intentions (bottom panel, to the right of “speed”). The inset windows on the left and right display the cyclist’s legs, such that the patient can directly observe the changes in the movement speed of both the legs. These views provide real-time feedback to patients