Decoding Information for Grasping from the Macaque Dorsomedial Visual Stream.

Area of Science:

• Neuroscience
• Neural Engineering
• Brain-Computer Interfaces

Background:

• Neurodecoders are crucial for neuroprosthetic devices and understanding neural functions.
• Previous research focused on decoding movement information from primate cortex for restoring arm actions.

Purpose of the Study:

• To investigate the decoding of grip configurations from neural signals in the monkey medial posterior parietal cortex.
• To assess the reliability and timing of decoding grasp types for potential neuroprosthetic applications.

Main Methods:

• Trained two Macaca fascicularis monkeys on a reach-to-grasp task with varying visual conditions.
• Acquired population neural activity from area V6A during object vision, delay, and grasp execution.
• Utilized a Bayes classifier to decode object and grip types from neural data.

Main Results:

• Achieved decoding recognition rates significantly above chance level across all analyzed epochs.
• Observed variations in decoding accuracy based on visual task conditions.
• Demonstrated that neural decoding can predict grasp types in advance of movement onset, indicating a changing neural code.

Conclusions:

• Neural signals from the medial posterior parietal cortex reliably represent grip configurations.
• These signals are a promising substrate for developing neuroprostheses to enable dexterous grasping actions.
• The medial posterior parietal cortex is highlighted as a key target for transforming neural activity into control signals for prosthetics.