Measuring instability in chronic human intracortical neural recordings towards stable, long-term brain-computer...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Intracortical brain-computer interfaces (iBCIs) offer intuitive cursor control for individuals with tetraplegia.
• Long-term iBCI performance degrades due to shifts between neural activity and intended movements.
• Identifying performance degradation is crucial for practical, extended iBCI use.

Purpose of the Study:

• To propose and validate a method (MINDFUL) for measuring neural data instability in iBCIs.
• To enable detection of performance degradation without requiring user intention labels.
• To establish a strategy for determining optimal recalibration timing for long-term iBCI functionality.

Main Methods:

• Analysis of longitudinal neural data from two participants with tetraplegia using the BrainGate2 system.
• Development of the MINDFUL method to quantify neural signal instability.
• Correlation analysis between the MINDFUL instability measure and closed-loop cursor performance.

Main Results:

• The MINDFUL method successfully measured neural data instability over extended periods (142 and 28 days).
• Instability measures showed a strong correlation with changes in closed-loop cursor performance (Pearson r = 0.93 and 0.72).
• Performance degradation could be inferred solely from recorded neural activity.

Conclusions:

• The MINDFUL method provides a reliable way to infer online iBCI performance from neural data alone.
• This strategy can guide recalibration timing for sustained and practical iBCI use.
• MINDFUL advances the development of robust, long-term brain-computer interfaces for assistive technology.