eLife | Neural Geometry from Mixed Sensorimotor Selectivity for Predictive Sensorimotor Control | He Cui Lab
The He Cui Laboratory at the Chinese Institute for Brain Research (CIBR) published a research article title “Neural Geometry from Mixed Sensorimotor Selectivity for Predictive Sensorimotor Control” in eLife on August 7, 2024. The study reveals that the activity of most neurons is jointly tuned to both reach direction and target motion. Principle component analysis (PCA) on the neural population shows a clear orbital neural geometry in low-dimensional space at movement onset: The neural states were distributed in reach-direction order, and formed ringlike structures whose slopes were determined by target-motion conditions. The single-unit mixed selectivity and the ring structure also emerge in RNNs without specified instructions. The results of this study expand the functional understanding of the motor cortex, suggest that sensory modulation may be an important part of the process of predictive motor control, help exploring the neural mechanism of motor control of moving targets, and lay a foundation for the application of brain-computer interface in dynamic environments.
