Summary: New single-neuron recordings show the hand knob region of the premotor cortex contains information about movements across the whole body and uses a compositional code linking similar actions from different limbs.
Source: Cell Press
Mapping how discrete brain regions correspond to thoughts, actions, and specific neural functions is a central goal in neuroscience. Noninvasive tools like fMRI and EEG have outlined broad functional areas, but they cannot resolve activity at the level of individual neurons. This limitation leaves open important questions about how single neurons represent different body parts and movements.
In a paper published March 26 in the journal Cell, researchers report microelectrode array recordings from two human participants that map motor representations down to the single-neuron level. The results reveal that the so-called hand knob area of the premotor cortex—previously associated mainly with hand and arm control—encodes movements across the entire body. The study also uncovers how neurons coordinate to represent limb identity separately from the specific movement, revealing a compositional neural code.
“For the first time, we show that a brain region long thought to represent only the arm and hand actually contains information about movements across the whole body,” says Frank Willett, the study’s first author and a postdoctoral fellow in the Neural Prosthetics Translational Laboratory at Stanford University and the Howard Hughes Medical Institute. “We also found a shared neural code that links body parts together, which has implications for how the brain generalizes motor skills.”
This work is part of BrainGate2, a multisite pilot clinical trial developing and testing implantable neurotechnology to restore communication and independence to people with paralysis and related conditions. A core goal of the Stanford team has been to translate intracortical signals into reliable control of assistive devices and communication tools.
The study involved two individuals with chronic tetraplegia—one with a high-level spinal cord injury and the other with amyotrophic lateral sclerosis. Both participants had microelectrode arrays implanted in the hand knob region of the precentral gyrus. While attempting specific movements such as lifting a finger or rotating an ankle, the participants’ single-neuron action potentials were recorded and analyzed.
The researchers were surprised to find that neurons in the hand knob area responded not only to attempted hand and arm movements but also to movements of the legs, face, and other body regions. Further analysis showed that neural patterns for matching movements across different limbs—such as raising a wrist and raising an ankle—were more similar than expected given the different muscles involved.
“When we compared analogous movements across limbs, the resulting neural activity patterns were remarkably alike,” Willett explains. “That suggests a compositional organization in motor cortex in which one component represents which limb is involved and another represents the type of movement. This arrangement could make it easier for the brain to transfer learned skills from one limb to another.”
These findings have direct relevance for brain-computer interface (BCI) design. Previously, researchers assumed that separate implants across widely distributed motor areas would be required to decode commands for many different body parts. The new results suggest that a single implant in the hand knob area may provide signals sufficient to control movements or interfaces associated with the whole body.
One promising application is communication for people with paralysis or locked-in syndrome. Diverse neural signals tied to different intended body movements could be mapped to distinct computer commands—for example, different types of cursor clicks or selection gestures—potentially improving the accuracy and flexibility of BCI-driven communication compared with relying only on hand or arm signals.
Funding: This research was supported by the Office of Research and Development, Rehabilitation R and D Service, Department of Veterans Affairs; the Executive Committee on Research of Massachusetts General Hospital; NIDCD; NINDS; Larry and Pamela Garlick; Samuel and Betsy Reeves; the Wu Tsai Neuroscience Institute at Stanford; the Simons Foundation Collaboration on the Global Brain; the Office of Naval Research; and the Howard Hughes Medical Institute.
Source:
Cell Press
Media Contacts:
Carly Britton – Cell Press
Image Source:
Public domain image
Original Research: Open access
“Hand Knob Area of Premotor Cortex Represents the Whole Body in a Compositional Way” — Frank Willett et al., Cell, doi: 10.1016/j.cell.2020.02.043.
Abstract
Hand Knob Area of Premotor Cortex Represents the Whole Body in a Compositional Way
Highlights
• The hand knob region of premotor cortex is tuned to movements across the entire body.
• A compositional neural code links analogous movements across all four limbs.
• Movement type and limb identity are represented separately, which may facilitate skill transfer.
• A discrete intracortical BCI can accurately decode movements for all four limbs from signals in the hand knob.
Summary
Despite decades of research since the motor homunculus was proposed, the fine-grained intermixing of body-part representations in human motor cortex at single-neuron resolution has remained unclear. Using multi-unit intracortical recordings in people with tetraplegia, the study demonstrates robust representation of face, head, arm, and leg movements within the hand knob area of premotor cortex. The neural code has two separable components: a limb-coding component indicating which limb is to be moved and a movement-coding component in which analogous movements from different limbs are represented similarly (for example, a hand grasp and a toe curl). This compositional organization could help the brain generalize motor skills between limbs and provides a framework for designing whole-body BCIs that leverage signals from a single cortical region.