Towards a Naturalistic Brain-Machine Interface: Hybrid Torque and Position Control Allows Generalization to Novel Dynamics

摘要: Realization of reaching and grasping movements by a paralytic person or an amputee would greatly facilitate her/his activities daily living. Towards this goal, control computer cursor robotic arm using neural signals has been demonstrated in rodents, non-human primates humans. This technology is commonly referred to as Brain-Machine Interface (BMI) achieved predictions kinematic parameters, e.g. position velocity. However, execution natural movements, such swinging baseball bats different weights at the same speed, requires advanced planning for necessary context-specific forces addition control. Here we show, first time, virtual with representative inertial parameters real-time torques (M. radiata). We found that leads ballistic, possibly more naturalistic than alone, adding influence hybrid torque-position changes feedforward behavior these BMI movements. In addition, level was achievable utilizing recordings from either contralateral ipsilateral M1. also observed changed under novel external dynamic environments comparable Our results demonstrate inclusion torque drive neuroprosthetic device gives user direct handle on movement execution, especially when dealing changing environments. anticipate our be starting point sophisticated algorithms sensorimotor neuroprostheses, eliminating need fully automatic kinematic-to-dynamic transformations currently used traditional kinematic-based decoders. Thus, propose torques, other force related variables, should allow user.