Comparing a BCI communication system in a patient with Multiple System Atrophy, with an animal model

Area of Science:

• Neuroscience
• Biomedical Engineering
• Machine Learning

Background:

• Paralysis significantly impairs communication abilities in affected individuals worldwide.
• Multiple System Atrophy (MSA) is a neurodegenerative disease causing paralysis and other neural deficits.
• Brain-Computer Interfaces (BCIs) offer a potential solution for restoring communication in paralyzed patients.

Purpose of the Study:

• To develop and evaluate a microelectrode-based BCI system for communication in a patient with MSA.
• To assess the BCI system's effectiveness by testing it in a non-human primate (NHP).
• To compare the performance of Linear Discriminant Analysis (LDA) and Long Short-Term Memory (LSTM) models in decoding neural signals.

Main Methods:

• A microelectrode-based BCI was implanted in a human patient with MSA and an NHP.
• Neural data from both subjects were used to train LDA and LSTM-based Artificial Neural Network (ANN) models.
• The models were trained to perform binary classification for decoding movement versus non-movement.

Main Results:

• The LDA model achieved up to 72.7% accuracy in binary decoding for the human patient, with highly variable performance.
• The BCI system demonstrated high accuracy in the NHP, with LDA achieving 82.7% and LSTM achieving 83.7%.
• The BCI system showed significantly lower accuracy in the human patient, with LDA at 47.0% and LSTM at 44.6%.

Conclusions:

• Neurodegenerative diseases like MSA may present unique challenges for BCI-based communication systems.
• The study highlights discrepancies in BCI performance between human patients with neurodegenerative diseases and NHPs.
• Further research is needed to understand and overcome the mechanisms impeding BCI efficacy in MSA patients.