Decoding Saccade Intention From Primate Prefrontal Cortical Local Field Potentials Using Spectral, Spatial, and...

Area of Science:

• Neuroscience
• Brain-Computer Interfaces
• Cognitive Science

Background:

• Invasive Brain-Computer Interfaces (iBCIs) typically rely on motor cortex signals.
• The potential of using signals encoding global cognitive variables for iBCI improvement is under investigation.
• Lateral prefrontal cortex (LPFC) involvement in attention and eye movement planning is known.

Purpose of the Study:

• To investigate the encoding of eye-movement intention using Local Field Potential (LFP) signals from the LPFC.
• To assess the utility of LPFC LFP signals for decoding saccade goals in a delayed saccade task.
• To develop a robust workflow for iBCIs utilizing LPFC signals.

Main Methods:

• Recorded LFP signals from the LPFC of macaque monkeys during a delayed saccade task.
• Applied spectral dimensionality reduction to analyze broadband LFP activity.
• Utilized support vector machine classification to decode saccade target locations.

Main Results:

• Broadband LFP signal dynamics in low dimensions revealed saccade target separation during the pre-movement fixation period.
• LPFC LFP signals demonstrated discriminative potential for decoding intended saccade goals.
• A general workflow for iBCIs using LPFC LFP was proposed, showing robustness to electrode signal loss.

Conclusions:

• Broadband LFP signals from the LPFC can effectively decode intended saccade target locations.
• LPFC signals offer a valuable data source for enhancing iBCI performance by incorporating cognitive variables.
• The proposed decoding workflow provides a promising direction for future iBCI development.