Spatial co-adaptation of cortical control columns in a micro-ECoG brain-computer interface

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electrocorticography (ECoG) is a key technology for brain-computer interfaces (BCIs).
• Limited research exists on chronic ECoG electrode performance specifically for BCI applications.
• Understanding electrode spacing and decoding strategies is crucial for BCI advancement.

Purpose of the Study:

• To investigate the impact of inter-electrode distance on chronic ECoG BCI performance.
• To compare BCI performance using fixed versus adaptive decoding weights.
• To explore neural adaptation in the context of closed-loop ECoG BCIs.

Main Methods:

• Utilized epidural micro-ECoG signals (75-105 Hz) with 300 μm electrodes for 1D and 2D BCI tasks.
• Tested inter-electrode distances ranging from 3 mm to 15 mm.
• Compared fixed decoding weights on a subset of channels against adaptive weights using the entire electrode array.

Main Results:

• Successful BCI control was achieved with electrodes spaced 9 mm and 15 mm apart.
• Electrode spacing of 3 mm led to decreased performance and increased signal correlation.
• Adaptive decoding weights significantly improved 2D BCI performance (80%-90%) compared to fixed weights (50%-60%).
• Adaptive decoding increased modulation in previously unused channels.

Conclusions:

• Neural activity adapts to generate control signals, with a practical minimal spatial resolution of approximately 3 mm for micro-ECoG BCI.
• The interplay between neural adaptation and machine learning is vital for optimizing ECoG BCI performance.
• Findings inform the design and implementation of more effective chronic ECoG BCIs.