Improving motor imagery classification during induced motor perturbations

C Vidaurre,T Jorajuría,V V Nikulin,K-R Müller,M Gómez,A Ramos-Murguialday

doi:10.1088/1741-2552/ac123f

Abstract

Objective. Motor imagery is the mental simulation of movements. It is a common paradigm to design brain-computer interfaces (BCIs) that elicits the modulation of brain oscillatory activity similar to real, passive and induced movements. In this study, we used peripheral stimulation to provoke movements of one limb during the performance of motor imagery tasks. Unlike other works, in which induced movements are used to support the BCI operation, our goal was to test and improve the robustness of motor imagery based BCI systems to perturbations caused by artificially generated movements. Approach. We performed a BCI session with ten participants who carried out motor imagery of three limbs. In some of the trials, one of the arms was moved by neuromuscular stimulation. We analysed 2-class motor imagery classifications with and without movement perturbations. We investigated the performance decrease produced by these disturbances and designed different computational strategies to attenuate the observed classification accuracy drop. Main results. When the movement was induced in a limb not coincident with the motor imagery classes, extracting oscillatory sources of the movement imagination tasks resulted in BCI performance being similar to the control (undisturbed) condition; when the movement was induced in a limb also involved in the motor imagery tasks, the performance drop was significantly alleviated by spatially filtering out the neural noise caused by the stimulation. We also show that the loss of BCI accuracy was accompanied by weaker power of the sensorimotor rhythm. Importantly, this residual power could be used to predict whether a BCI user will perform with sufficient accuracy under the movement disturbances. Significance. We provide methods to ameliorate and even eliminate motor related afferent disturbances during the performance of motor imagery tasks. This can help improving the reliability of current motor imagery based BCI systems.

Highlights

Non-invasive brain-computer interfaces (BCIs) allow generating a control signal by inferring a user’s intention from the ongoing electroencephalography (EEG)
When the movement was induced in a limb not coincident with the motor imagery classes, extracting oscillatory sources of the movement imagination tasks resulted in BCI performance being similar to the control condition; when the movement was induced in a limb involved in the motor imagery tasks, the performance drop was significantly alleviated by spatially filtering out the neural noise caused by the stimulation
Signal to noise ratio of motor imagery In order to understand performance drops due to movement stimulation, we evaluated the amount of event related desychronization/synchronization (ERD/ERS) observed for each subject (Gastaut and Bert 1954, Pfurtscheller and da Silva 1999)

Summary

Introduction

Non-invasive brain-computer interfaces (BCIs) allow generating a control signal by inferring a user’s intention from the ongoing electroencephalography (EEG) (see Vidal 1973, Kuebler et al 2001, Birbaumer 2006, Neuper et al 2006, Sejnowski et al 2007, Müller et al 2008, Blankertz et al 2010b, Millan et al 2010, McFarland and Wolpaw 2017, Scherer and Vidaurre 2018, Stevenson et al 2019, Mane et al 2020). In BCIs based on the modulation of the sensorimotor rhythms (SMR), the imagination of movements elicits changes in the motor cortex that are reflected in the EEG. Other possible interventions that cause SMR modulations are passive and induced movements. The stimulation of muscles, even under the threshold of movement, excites neural networks in the brain (Chatterjee et al 2007, Cho et al 2011, Vidaurre et al 2013, 2019, Ahn et al 2014, Yi et al 2017, Corbet et al 2018) and stimulated (induced) movements cause SMR modulations through the activation of the afferent pathways (e.g. stimulating one nerve or some muscles directly). Electrical stimulation is based on a very broad variety of stimulation parameters (e.g. pulse width, voltage amplitude, current, electrodes shape and size, frequency) and empirically found standard stimulation parameters are normally chosen to achieve effective motor or sensory stimulation (Doucet et al 2012)

Methods

Results

Discussion

Conclusion