Abstract|Motor Neuron Disease 1| Volume 357, SUPPLEMENT 1, e48-e49, October 15, 2015

Reaching and grasping a glass of water by locked-in ALS patients through a BCI-controlled humanoid robot

      Background. Amyotrophic lateral sclerosis (ALS) leads to a complete limb paralysis, dysphagia and anarthria. A brain-computer interface (BCI) technology may aid ALS patients in communication and motor control.
      Objective. To set up a BCI system for wireless control, by ALS patients, of a humanoid robot, with the aim to reach and grasp a glass of water.
      Patients and methods. Four non-demented ALS patients were recruited. Controls were four healthy subjects, matched for demographic variables. A BCI command interpreter was used to control a humanoid robot. The task was to instruct the robot to move towards a glass of water, reach and grasp it (first item) and then bring the glass to the subject (second item). The protocol consisted of a calibration session, an online session, in which the two items are sequentially selected, and a robot session where the two items translate into high level commands. The minimal accuracy of each response and the number of errors each session were evaluated and analysed.
      Results. All ALS patients completed the task (5 trials, 95.5% success). Controls performed comparably with a 100% success over the 5 trials. The minimum accuracy leading to a correct item selection for the robot movement was slightly better for ALS patients (ALS 60% vs controls 53%, p = 0.49).
      Conclusions. ALS patients can successfully control a humanoid robot through a BCI system. This bears the potential to virtually restore the autonomous motion of an ALS patient, enabling him to extend his presence beyond the boundaries of his bed.