Decoding Saccadic Directions Using Epidural ECoG in Non-Human Primates.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Brain-Computer Interfaces

Background:

• Existing brain-computer interfaces (BCI) primarily rely on SSVEP, P300, or motor imagery, necessitating diverse protocols for different patient needs.
• Locked-in syndrome patients require alternative communication methods, highlighting the need for innovative BCI systems.

Purpose of the Study:

• To develop and validate a novel BCI paradigm based on predicting eye movement direction using epidural electrocorticography (eECoG) signals.
• To assess the feasibility of using a choice saccade (CS) task for BCI applications in non-human primates.

Main Methods:

• Trained non-human primates on a choice saccade (CS) task.
• Recorded brain activity using epidural electrocorticography (eECoG).
• Utilized a support vector machine (SVM) with recursive feature elimination (RFE) to decode eECoG signals and predict eye movement direction.

Main Results:

• Successfully predicted upcoming eye movement direction with mean accuracies of 80% for 2 directions and 43% for 4 directions.
• Identified key brain regions, including the frontal eye field (FEF), supplementary eye field (SEF), and superior parietal lobule (SPL), as crucial for classification.
• Determined that the alpha (α) frequency band and specific time windows post-cue and pre-saccade were most informative for prediction.

Conclusions:

• A saccade-based BCI paradigm is feasible and effective for decoding intended movement direction from eECoG signals.
• This approach offers a potential pathway for creating intuitive and convenient 2D communication platforms for individuals with severe communication disabilities.