An Impending Paradigm Shift in Motor Imagery Based Brain-Computer Interfaces.

Sotirios Papadopoulos,Jérémie Mattout,James Bonaiuto

doi:10.3389/fnins.2021.824759

Abstract

The development of reliable assistive devices for patients that suffer from motor impairments following central nervous system lesions remains a major challenge in the field of non-invasive Brain-Computer Interfaces (BCIs). These approaches are predominated by electroencephalography and rely on advanced signal processing and machine learning methods to extract neural correlates of motor activity. However, despite tremendous and still ongoing efforts, their value as effective clinical tools remains limited. We advocate that a rather overlooked research avenue lies in efforts to question neurophysiological markers traditionally targeted in non-invasive motor BCIs. We propose an alternative approach grounded by recent fundamental advances in non-invasive neurophysiology, specifically subject-specific feature extraction of sensorimotor bursts of activity recorded via (possibly magnetoencephalography-optimized) electroencephalography. This path holds promise in overcoming a significant proportion of existing limitations, and could foster the wider adoption of online BCIs in rehabilitation protocols.

Highlights

Central nervous system (CNS) lesions have a major socioeconomic impact on modern societies (World Health Organization, 2011)
A prominent direction lies in the development of Brain-Computer Interfaces (BCIs) whose output can be translated into motor commands for rehabilitation protocols (Raffin and Hummel, 2017; Coscia et al, 2019), or devices such as wheelchairs, spellers, exoskeletons, and prostheses (Rosenfeld and Wong, 2017), in order to assist patients in overcoming motor disabilities following stroke, spinal cord injuries, and other CNS pathologies or peripheral deficits, respectively
Despite the sophisticated methods employed, still about a third of subjects have great difficulty or are totally unable to control BCIs, a phenomenon known as BCI illiteracy (Vidaurre and Blankertz, 2010). We argue that this focus on advanced signalprocessing and feature extraction techniques for motor imagery (MI) based BCIs has effectively rendered event-related desynchronization (ERD) and event-related synchronization (ERS) the only signals of interest, creating a gap between the BCI community and recent advances in neurophysiology

Summary

INTRODUCTION

Central nervous system (CNS) lesions have a major socioeconomic impact on modern societies (World Health Organization, 2011). We argue that this focus on advanced signalprocessing and feature extraction techniques for MI based BCIs has effectively rendered ERD and ERS the only signals of interest, creating a gap between the BCI community and recent advances in neurophysiology In light of these considerations we believe that significant progress toward the development of high-fidelity non-invasive BCIs, suitable for online motor decoding, will be achieved only if we attempt to bridge the gap between basic neuroscience and applied neuroengineering. The major challenges for non-invasive BCIs are to deal with highly variable and noisy single trial activity, and to overcome the low SNR and limited spatial resolution in order to improve online feature extraction and decoding. We anticipate that these challenges will be even more acute when targeting short-lasting, potentially non-sinusoidal bursts of frequency specific activity. Optically-pumped the neural features most informative of imaged movements magnetometers (OPMs) may be able to replace hybrid M/EEG (Boto et al, 2019, 2021; Iivanainen et al, 2019; Roberts et al, 2019; Borna et al, 2020; Hill et al, 2020; Paek et al, 2020; Wittevrongel et al, 2021), as they promise to be more portable and cheaper while improving SNR

A Paradigm Shift Toward Motor Rehabilitation

CONCLUSION

DATA AVAILABILITY STATEMENT