Decoding continuous goal-directed movement from human brain-wide intracranial recordings

Abstract

Reaching out your hand is an effortless yet complex behavior that is indispensable in daily life. Neural correlates of reaching behavior have been observed and decoded beyond the motor cortex, but the degree and granularity of movement representation are not fully understood. Here, we decode 12 kinematics of goal-directed reaching behavior from 18 participants implanted with stereotactic-electroencephalography electrodes performing a 3D reaching task. The decoder is able to decode continuous movement kinematics using low-, mid-, and high-frequency information in all participants using preferential subspace identification. Neural correlates of movements are observed throughout the brain, including deeper structures. Switching to a goal-centric reference frame enables the decoder to decode hand position, indicating that low-frequency activity is involved in higher-order processing of movements. Our results strengthen the evidence that brain-wide motor-related dynamics can be decoded and may provide opportunities for brain-computer interfaces for individuals with a compromised motor cortex.

Keywords: BCI; CP: Neuroscience; PSID; brain-computer interface; goal-directed; hand movement; kinematics; motor decoding; reference frame; sEEG.