Sub-mm functional decoupling of electrocortical signals through closed-loop BMI learning
Summary
Researchers explored brain-machine interfaces (BMI) by decoupling neural signals from closely spaced electrodes. Closed-loop learning enabled rats to control a cursor using high gamma power, demonstrating precise volitional control.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Neural Engineering
Background:
- Volitional control of neural activity is fundamental to Brain-Machine Interface (BMI) systems.
- Investigating signal decoupling from closely spaced electrodes is crucial for high-resolution neural recording.
Purpose of the Study:
- To determine if subdural field potentials from microelectrodes (<1mm apart) can be decoupled using closed-loop BMI learning.
- To assess the stability and performance of novel microelectrode arrays for neural recording.
Main Methods:
- Fabrication of custom, flexible microelectrode arrays with 200 µm pitch and platinum black deposition.
- Chronic subdural implantation over the primary motor cortex (M1) in rats.
- In vivo monitoring of electrode impedance and closed-loop training for a center-out task using high gamma power (70-110 Hz).
Main Results:
- Successfully decoupled neural signals from microelectrodes separated by less than 1 mm.
- Demonstrated stable neural interfaces with consistent electrode impedance.
- Rats learned to perform a center-out task by volitionally modulating high gamma power for cursor control.
Conclusions:
- Closed-loop BMI learning can effectively decouple neural signals from adjacent microelectrodes.
- Custom microelectrode arrays show promise for stable, high-resolution neural recording and BMI applications.
- This approach facilitates precise volitional control of neural activity for advanced BMI systems.
