Can a highly accurate multi-class SSMVEP BCI induce sensory-motor rhythm in sensorimotor area?

Cortical activation of neuromuscular electrical stimulation synchronized mirror neuron rehabilitation strategies: an fNIRS study

Background

The mirror neuron system (MNS) plays a key role in the neural mechanism underlying motor learning and neural plasticity. Action observation (AO), action execution (AE), and a combination of both, known as action imitation (AI), are the most commonly used rehabilitation strategies based on MNS. It is possible to enhance the cortical activation area and amplitude by combining traditional neuromuscular electrical stimulation (NMES) with other top-down and active rehabilitation strategies based on the MNS theory.

Objective

This study aimed to explore the cortical activation patterns induced by NMES synchronized with rehabilitation strategies based on MNS, namely NMES+AO, NMES+AE, and NMES+AI. In addition, the study aimed to assess the feasibility of these three novel rehabilitative treatments in order to provide insights and evidence for the design, implementation, and application of brain-computer interfaces.

Methods

A total of 70 healthy adults were recruited from July 2022 to February 2023, and 66 of them were finally included in the analysis. The cortical activation patterns during NMES+AO, NMES+AE, and NMES+AI were detected using the functional Near-Infrared Spectroscopy (fNIRS) technique. The action to be observed, executed, or imitated was right wrist and hand extension, and two square-shaped NMES electrodes were placed on the right extensor digitorum communis. A block design was adopted to evaluate the activation intensity of the left MNS brain regions.

Results

General linear model results showed that compared with the control condition, the number of channels significantly activated (PFDR < 0.05) in the NMES+AO, NMES+AE, and NMES+AI conditions were 3, 9, and 9, respectively. Region of interest (ROI) analysis showed that 2 ROIs were significantly activated (PFDR < 0.05) in the NMES+AO condition, including BA6 and BA44; 5 ROIs were significantly activated in the NMES+AE condition, including BA6, BA40, BA44, BA45, and BA46; and 6 ROIs were significantly activated in the NMES+AI condition, including BA6, BA7, BA40, BA44, BA45, and BA46.

Conclusion

The MNS was activated during neuromuscular electrical stimulation combined with an AO, AE, and AI intervention. The synchronous application of NMES and mirror neuron rehabilitation strategies is feasible in clinical rehabilitation. The fNIRS signal patterns observed in this study could be used to develop brain-computer interface and neurofeedback therapy rehabilitation devices.

Bio-robotics research for non-invasive myoelectric neural interfaces for upper-limb prosthetic control: a 10-year perspective review

A decade ago, a group of researchers from academia and industry identified a dichotomy between the industrial and academic state-of-the-art in upper-limb prosthesis control, a widely used bio-robotics application. They proposed that four key technical challenges, if addressed, could bridge this gap and translate academic research into clinically and commercially viable products. These challenges are unintuitive control schemes, lack of sensory feedback, poor robustness and single sensor modality. Here, we provide a perspective review on the research effort that occurred in the last decade, aiming at addressing these challenges. In addition, we discuss three research areas essential to the recent development in upper-limb prosthetic control research but were not envisioned in the review 10 years ago: deep learning methods, surface electromyogram decomposition and open-source databases. To conclude the review, we provide an outlook into the near future of the research and development in upper-limb prosthetic control and beyond.

Age-related differences in the transient and steady state responses to different visual stimuli

Objective

Brain-computer interface (BCI) has great potential in geriatric applications. However, most BCI studies in the literature used data from young population, and dedicated studies investigating the feasibility of BCIs among senior population are scarce. The current study, we analyzed the age-related differences in the transient electroencephalogram (EEG) response used in visual BCIs, i.e., visual evoked potential (VEP)/motion onset VEP (mVEP), and steady state-response, SSVEP/SSMVEP, between the younger group (age ranges from 22 to 30) and senior group (age ranges from 60 to 75).

Methods

The visual stimulations, including flicker, checkerboard, and action observation (AO), were designed with a periodic frequency. Videos of several hand movement, including grasping, dorsiflexion, the thumb opposition, and pinch were utilized to generate the AO stimuli. Eighteen senior and eighteen younger participants were enrolled in the experiments. Spectral-temporal characteristics of induced EEG were compared. Three EEG algorithms, canonical correlation analysis (CCA), task-related component analysis (TRCA), and extended CCA, were utilized to test the performance of the respective BCI systems.

Results

In the transient response analysis, the motion checkerboard and AO stimuli were able to elicit prominent mVEP with a specific P1 peak and N2 valley, and the amplitudes of P1 elicited in the senior group were significantly higher than those in the younger group. In the steady-state analysis, SSVEP/SSMVEP could be clearly elicited in both groups. The CCA accuracies of SSVEPs/SSMVEPs in the senior group were slightly lower than those in the younger group in most cases. With extended CCA, the performance of both groups improved significantly. However, for AO targets, the improvement of the senior group (from 63.1 to 71.9%) was lower than that of the younger group (from 63.6 to 83.6%).

Conclusion

Compared with younger subjects, the amplitudes of P1 elicited by motion onset is significantly higher in the senior group, which might be a potential advantage for seniors if mVEP-based BCIs is used. This study also shows for the first time that AO-based BCI is feasible for the senior population. However, new algorithms for senior subjects, especially in identifying AO targets, are needed.

A Novel Online Action Observation-Based Brain-Computer Interface That Enhances Event-Related Desynchronization

Brain-computer interface (BCI)-based stroke rehabilitation is an emerging field in which different studies have reported variable outcomes. Among the BCI paradigms, motor imagery (MI)-based closed-loop BCI is still the main pattern in rehabilitation training. It can estimate a patient' motor intention and provide corresponding feedback. However, the individual difference in the ability to generate event-related desynchronization (ERD) and the low classification accuracy of the multi-class scenario restrict the application of MI-based BCI. In the current study, a novel online action observation (AO)-based BCI was proposed. The visual stimuli of four types of hand movements were designed to simultaneously induce steady-state motion visual evoked potential (SSMVEP) in the occipital region and to activate the sensorimotor region. Task-related component analysis was performed to identify the SSMVEP. Results showed that the amplitude of the induced frequency in the SSMVEP had a negative relationship with the stimulus frequency. The classification accuracy in the four-class scenario reached 72.81 ± 13.55% within 2.5s. Importantly, the AO-based closed-loop BCI, which provided visual feedback based on the SSMVEP, could enhance ERD compared with AO-alone. The increased attentiveness might be one key factor for the enhancement of the ERD in the designed AO-based BCI. In summary, the proposed AO-based BCI provides a new insight for BCI-based rehabilitation.

Enhancing Detection of SSMVEP Induced by Action Observation Stimuli Based on Task-Related Component Analysis

Action observation (AO)-based brain-computer interface (BCI) is an important technology in stroke rehabilitation training. It has the advantage of simultaneously inducing steady-state motion visual evoked potential (SSMVEP) and activating sensorimotor rhythm. Moreover, SSMVEP could be utilized to perform classification. However, SSMVEP is composed of complex modulation frequencies. Traditional canonical correlation analysis (CCA) suffers from poor recognition performance in identifying those modulation frequencies at short stimulus duration. To address this issue, task-related component analysis (TRCA) was utilized to deal with SSMVEP for the first time. An interesting phenomenon was found: different modulated frequencies in SSMVEP distributed in different task-related components. On this basis, a multi-component TRCA method was proposed. All the significant task-related components were utilized to construct multiple spatial filters to enhance the detection of SSMVEP. Further, a combination of TRCA and CCA was proposed to utilize both advantages. Results showed that the accuracies using the proposed methods were significant higher than that using CCA at all window lengths and significantly higher than that using ensemble-TRCA at short window lengths (≤2 s). Therefore, the proposed methods further validate the induced modulation frequencies and will speed up the application of the AO-based BCI in rehabilitation.

Comparison of Modern Highly Interactive Flicker-Free Steady State Motion Visual Evoked Potentials for Practical Brain-Computer Interfaces

Motion-based visual evoked potentials (mVEP) is a new emerging trend in the field of steady-state visual evoked potentials (SSVEP)-based brain-computer interfaces (BCI). In this paper, we introduce different movement-based stimulus patterns (steady-state motion visual evoked potentials-SSMVEP), without employing the typical flickering. The tested movement patterns for the visual stimuli included a pendulum-like movement, a flipping illusion, a checkerboard pulsation, checkerboard inverse arc pulsations, and reverse arc rotations, all with a spelling task consisting of 18 trials. In an online experiment with nine participants, the movement-based BCI systems were evaluated with an online four-target BCI-speller, in which each letter may be selected in three steps (three trials). For classification, the minimum energy combination and a filter bank approach were used. The following frequencies were utilized: 7.06 Hz, 7.50 Hz, 8.00 Hz, and 8.57 Hz, reaching an average accuracy between 97.22% and 100% and an average information transfer rate (ITR) between 15.42 bits/min and 33.92 bits/min. All participants successfully used the SSMVEP-based speller with all types of stimulation pattern. The most successful SSMVEP stimulus was the SSMVEP1 (pendulum-like movement), with the average results reaching 100% accuracy and 33.92 bits/min for the ITR.