Decoding of intended saccade direction in an oculomotor brain-computer interface.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Current invasive brain-computer interface (BCI) research primarily focuses on motor cortex signals for limb function restoration.
• The oculomotor system offers potential advantages for BCI applications requiring rapid spatial target selection, such as in augmentative and alternative communication (AAC).

Purpose of the Study:

• To demonstrate the feasibility of an intracortical BCI utilizing the oculomotor system for high-accuracy control.
• To investigate the decoding of intended saccadic eye movement direction from frontal cortical activity.

Main Methods:

• Development of a chronic intracortical BCI in non-human primates.
• Decoding of intended saccade direction in real time using neural activity from multiple frontal cortical areas.
• Analysis of local field potential (LFP) magnitude and spiking activity for decoding performance.

Main Results:

• High-accuracy real-time decoding of intended saccade direction was achieved, particularly for contralateral targets.
• Effective decoding was observed from the beginning of the BCI session, indicating minimal training requirement.
• High-frequency LFP magnitude (80-500 Hz) outperformed spiking activity for decoding, simplifying BCI implementation.
• Frontal eye fields and supplementary eye fields were primary contributors to decoding, with less contribution from dorsolateral prefrontal cortex.

Conclusions:

• Intracortical oculomotor BCIs are feasible and capable of high-accuracy control.
• These BCIs require little to no practice, making them user-friendly.
• Oculomotor BCIs are well-suited for 'point and click' computer operations, significantly benefiting AAC users.