Decoding grasp types from the monkey motor cortex and on-line control of a dexterous artificial hand

Abstract

Online decoding of grasp gestures and real-time control of a prosthetic hand using signals from dorsal premotor (PMd) and primary motor cortex (M1) is of fundamental importance in order to restore grasping capabilities in paralysed subjects. To this ultimate aim, authors of this study are addressing the problem of decoding signals from M1 to control an external device. In particular, this paper presents preliminary results of a study in which four different grasp types and the rest state have been successfully decoded from the brain signals of two trained monkeys while they were grasping different shaped objects. An artificial hand replicating the monkeys movements has been online controlled through the neural activity measured. Our results enabled asynchronous neural control of prosthetic hand gestures, which underline a feasible hand neural prosthesis in BMIs.