A Deep Classifier for Upper-Limbs Motor Anticipation Tasks in an Online BCI Setting.

Andrea Valenti,Michele Barsotti,Davide Bacciu,Luca Ascari

doi:10.3390/bioengineering8020021

Abstract

Decoding motor intentions from non-invasive brain activity monitoring is one of the most challenging aspects in the Brain Computer Interface (BCI) field. This is especially true in online settings, where classification must be performed in real-time, contextually with the user’s movements. In this work, we use a topology-preserving input representation, which is fed to a novel combination of 3D-convolutional and recurrent deep neural networks, capable of performing multi-class continual classification of subjects’ movement intentions. Our model is able to achieve a higher accuracy than a related state-of-the-art model from literature, despite being trained in a much more restrictive setting and using only a simple form of input signal preprocessing. The results suggest that deep learning models are well suited for deployment in challenging real-time BCI applications such as movement intention recognition.

Highlights

Decoding motor intentions from non-invasive brain activity monitoring is one of the most challenging aspects in the Brain ComputerInterface (BCI) field
The early detection of movement intention remains an open challenge for enhancing neuro-feedback interventions, in which the precise feedback timing is essential for promoting neuro-plasticity by contingently closing the loop between the motor information and the sensory information
With the aim of overcoming the limitations of “handmade” feature extractions and signal processing, and to provide a practical alternative to the panel of binary classifiers, in this work we propose a deep learning (DL)-based architecture for movement anticipation detection

Summary

Introduction

Decoding motor intentions from non-invasive brain activity monitoring (e.g., electroencephalogram, EEG) is one of the most challenging aspects in the Brain ComputerInterface (BCI) field. The possibility to reliably detect the planning of movements, and to anticipate the motor action, would positive affects the usage of both restorative and assistive BCI systems for rehabilitation of motor paralysis and to improve life quality of people for which brain-based communication would be the only viable option to interact. There is ample evidence that non-invasive restorative BCI can have a benefit relative to conventional neuro-rehabilitation approaches in patients with serious motor disability [1,2,3]. Regarding the assistive BCI, the detection of motor planning could pave the way for the establishment of intimate and natural interaction with external devices such as human-support robots, orthosis, wheelchair, etc., for those individual presenting highly impeded motor capabilities [4]. Together with the rapid scientific and technological growing in the fields of human sensing and Artificial Intelligence (AI), we are witnessing a growing interest of the industry production in BCI devices, which could enable the “natural” operation of external devices for both healthy subjects and especially for those subjects that would benefit more from the usage of BCI systems [5]

Methods

Results

Conclusion