A high performing brain-machine interface driven by low-frequency local field potentials alone and together with...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain-machine interfaces (BMIs) enable individuals with motor impairments to control external devices using neural activity.
• High-performance BMIs typically rely on action potentials (spikes), but signal quality can degrade over time.
• Local field potentials (LFPs) present a potential alternative or supplementary signal for BMI control.

Purpose of the Study:

• To evaluate the efficacy of local motor potential (LMP) decoding for BMI control.
• To develop and assess a hybrid BMI combining LMP and spike decoding.
• To improve the robustness and longevity of BMI systems.

Main Methods:

• Recorded neural data (spikes and LFPs) from non-human primates performing a reaching task.
• Implemented and evaluated closed-loop BMI control using biomimetic decoders.
• Compared BMI performance using LMP, spikes, and a hybrid approach.

Main Results:

• LMP decoding facilitated rapid and precise cursor control, outperforming previous LFP-based BMI methods.
• Hybrid decoding using both spikes and LMP significantly enhanced BMI performance, particularly when spike signal quality was compromised.
• LMP decoding demonstrated effectiveness even with limited spike signal availability.

Conclusions:

• Local motor potential (LMP) is a viable and effective control signal for brain-machine interfaces.
• LMP requires minimal power for extraction and can substitute for or augment diminished spike signals.
• This approach may extend the operational lifespan of BMIs, contributing to their clinical applicability.