EEG changes during passive movements improve the motor imagery feature extraction in BCIs-based sensory feedback calibration

Abstract

Objective. This work proposes a method for two calibration schemes based on sensory feedback to extract reliable motor imagery (MI) features, and provide classification outputs more correlated to the user's intention. Method. After filtering the raw EEG, a two-step method for spatial feature extraction by using the Riemannian Covariance Matrices (RCM) method and Common Spatial Patterns (CSP) is proposed here. It uses electroencephalogram (EEG) data from trials providing feedback, in an intermediate step composed of both kth nearest neighbors and probability analyses, to find periods of time in which the user probably performed well the MI task without feedback. These periods are then used to extract features with better separability, and train a classifier for MI recognition. For evaluation, an in-house dataset with eight healthy volunteers and two post-stroke patients that performed lower-limb MI, and consequently received passive movements as feedback was used. Other popular public EEG datasets (such as BCI Competition IV dataset IIb, among others) from healthy subjects that executed upper-and lower-limbs MI tasks under continuous visual sensory feedback were further used. Results. The proposed system based on the Riemannian geometry method in two-steps (RCM-RCM) outperformed significantly baseline methods, reaching average accuracy up to 82.29%. These findings show that EEG data on periods providing passive movement can be used to contribute greatly during MI feature extraction. Significance. Unconscious brain responses elicited over the sensorimotor areas may be avoided or greatly reduced by applying our approach in MI-based brain-computer interfaces (BCIs). Therefore, BCI's outputs more correlated to the user's intention can be obtained.

Gait Training-Based Motor Imagery and EEG Neurofeedback in Lokomat: A Clinical Intervention With Complete Spinal Cord Injury Individuals

“…This approach tested in a total of 12 sessions each one of 19 min in length, undoubtedly may be attractive for clinical practice, as it contributes optimizing therapies addressed to physically provide gait training by a robotic system, and also neurofeedback or BCI training. In our approach, gait MI and passive gait training are combined in Lokomat®, which agree with some studies suggesting that the combination of MI with passive movements may be attractive for upper limb or lower limb rehabilitation of individuals with severe motor impairments [29], [30]. In fact, Delisle et al, [30] observed that better lower-limb motor imagery features can be extracted using as a reference EEG changes from passive movements.…”

“…In our approach, gait MI and passive gait training are combined in Lokomat®, which agree with some studies suggesting that the combination of MI with passive movements may be attractive for upper limb or lower limb rehabilitation of individuals with severe motor impairments [29], [30]. In fact, Delisle et al, [30] observed that better lower-limb motor imagery features can be extracted using as a reference EEG changes from passive movements. Therefore, similar brain regions around the foot area (Cz) may be activated when an individuals received gait passive training in Lokomat®.…”

Evaluation of temporal, spatial and spectral filtering in CSP-based methods for decoding pedaling-based motor tasks using EEG signals

“…However, some related studies can be mentioned. For example, some studies have reported the use of CSP-based methods to classify upper-and lower-limb MI tasks, where Accs close to 0.75 have been achieved [23,36]. However, little is known about these algorithmic strategies in pedaling AM tasks, where it is possible to highlight that, in our study, it was possible to obtain an Acc close to 0.81.…”

“…Therefore, our findings make a contribution to the neuroscience community where we observed that pedaling tasks can be detected using EEG and machine learning-based techniques. Recently, our team has been interested in the study of neuroscience and BCIs with promising applicators to lower limbs rehabilitation using MMEBs [5,18,28,36]. However, one of the limitations identified from this work was that a real-time BCI was not implemented to classify pedaling tasks with an end-effector, which may limit the scope of our results.…”

Unraveling Transformative Effects after tDCS and BCI Intervention in Chronic Post-Stroke Patient Rehabilitation—An Alternative Treatment Design Study

“…The proposed method integrates cutting-edge technologies including tDCS, MI-BCI, VR, and a customized MP to facilitate the lower-limb rehabilitation of post-stroke patients. This innovative system also introduces a novel approach to BCI calibration, combining EEG signals from both MI and actual movements [19], fostering a real-time closed-loop system for neural rehabilitation. Furthermore, the study highlights the potential benefits and cost-effectiveness of tDCS in inducing cortical plasticity for motor recovery in lower-limb rehabilitation, aiming to mitigate neuromuscular disabilities and associated societal costs.…”