Time robust movement direction decoding in Local Field Potentials using channel ranking.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Brain Machine Interfaces (BMI) decode movement using neural signals like Local Field Potentials (LFP) and Single Unit Activity (SUA).
• A key challenge is developing time-robust decoders for intra-cortical recordings.
• Existing methods struggle with the dynamic range and variability of absolute signal power.

Purpose of the Study:

• To develop a novel, time-robust decoding technique for Brain Machine Interfaces (BMI).
• To improve the stability and accuracy of movement direction decoding from intra-cortical recordings.
• To overcome limitations associated with absolute signal power in Local Field Potentials (LFP).

Main Methods:

• A new technique using qualitative information from multiple LFP channels, rather than absolute power.
• Implementation of a power-based inter-channel ranking system to define LFP channel quality.
• Introduction of a parameter-based ranking system for channels with comparable power.

Main Results:

• Stable decoding of eight movement directions with over 56% average efficiency for two weeks.
• Achieved 46% decoding power at two weeks, significantly outperforming traditional methods (13%).
• Demonstrated superior decoding power in force field conditions compared to existing approaches.

Conclusions:

• The novel LFP-based qualitative channel analysis provides robust and stable BMI decoding.
• This method significantly enhances decoding performance and longevity over traditional techniques.
• The approach shows promise for advanced neuroprosthetics and motor control research.