Brain and muscle synergies for brain-machine interface decoding

The inspiration for this project stemmed from the need to improve brain-machine interface (BMI) performance, particularly in decoding complex multi-degree-of-freedom (DoF) movements. While the brain handles high-DoF tasks effortlessly, current BMI systems struggle to match this level of control. Our study was motivated by the hypothesis that the brain simplifies high-DoF tasks through synergies, functional groupings of muscles and neural activities. By examining whether these synergies could enhance decoding in implanted BMIs, we aimed to uncover strategies to bridge the gap between natural and artificial control.

We used advanced tools and methods to test this hypothesis. We collected data from two non-human primates performing dexterous finger tasks, using implanted neural and muscle electrodes. I applied three dimensionality reduction techniques to the brain and muscle data: principal component analysis (PCA), non-negative matrix factorization (NMF), and demixed PCA (dPCA). These techniques allowed us to analyze the synergies in neural and muscle activities while testing their effectiveness in denoising, compression and generalization across varied tasks.

Our results showed that dimensionality reduction methods effectively compressed neural and muscle data, achieving substantial reductions in data size while retaining high predictive power. However, synergies did not outperform the full dataset in decoding performance or generalization across task contexts.

This is still work in progress, but will be submitted to a journal in the next few weeks.