# Neural signal analysis in chronic stroke: advancing intracortical brain-computer interface design

## TL;DR

This paper explores how brain signals can be used to control devices in people with chronic stroke, offering insights into designing better brain-computer interfaces.

## Contribution

The study identifies new neural signal patterns and proposes decoding features specific to stroke-affected brains.

## Key findings

- Low frequency power decreases and high frequency power increases with distance from the stroke site.
- Coordinated cross-channel firing and low frequency bursts were observed during motor tasks and relaxation.
- Three features for decoding motor movements in stroke-affected brains were proposed.

## Abstract

Intracortical Brain-computer interfaces (iBCIs) are a promising technology to restore function after stroke. It remains unclear whether iBCIs will be able to use the signals available in the neocortex overlying stroke affecting the underlying white matter and basal ganglia.

Here, we decoded both local field potentials (LFPs) and spikes recorded from intracortical electrode arrays in a person with chronic cerebral subcortical stroke performing various tasks with his paretic hand, with and without a powered orthosis. Analysis of these neural signals provides an opportunity to explore the electrophysiological activities of a stroke affected brain and inform the design of medical devices that could restore function.

The frequency domain analysis showed that as the distance between an array and the stroke site increased, the low frequency power decreased, and high frequency power increased. Coordinated cross-channel firing of action potentials while attempting a motor task and cross-channel simultaneous low frequency bursts while relaxing were also observed. Using several offline analysis techniques, we propose three features for decoding motor movements in stroke-affected brains.

Despite the presence of unique activities that were not reported in previous iBCI studies with intact brain functions, it is possible to decode motor intents from the neural signals collected from a subcortical stroke-affected brain.