Brain Processes Involved in Motor Planning Are a Dominant Factor for Inducing Event-Related Desynchronization

Event-related desynchronization (ERD) is a relative attenuation in the spectral power of an electroencephalogram (EEG) observed over the sensorimotor area during motor execution and motor imagery. It is a well-known EEG feature and is commonly employed in brain-computer interfaces. However, its underlying neural mechanisms are not fully understood, as ERD is a single variable correlated with external events involving numerous pathways, such as motor intention, planning, and execution. In this study, we aimed to identify a dominant factor for inducing ERD. Participants were instructed to grasp their right hand with three different (10, 25, or 40%MVF: maximum voluntary force) levels under two distinct experimental conditions: a closed-loop condition involving real-time visual force feedback (VF) or an open-loop condition in a feedforward (FF) manner. In each condition, participants were instructed to repeat the grasping task a certain number of times with a timeline of Rest (10.0 s), Preparation (1.0 s), and Motor Execution (4.0 s) periods, respectively. EEG signals were recorded simultaneously with the motor task to evaluate the time-course of the event-related spectrum perturbation for each condition and dissect the modulation of EEG power. We performed statistical analysis of mu and beta-ERD under the instructed grasping force levels and the feedback conditions. In the FF condition (i.e., no force feedback), mu and beta-ERD were significantly attenuated in the contralateral motor cortex during the middle of the motor execution period, while ERD in the VF condition was maintained even during keep grasping. Only mu-ERD at the somatosensory cortex tended to be slightly stronger in high load conditions. The results suggest that the extent of ERD reflects neural activity involved in the motor planning process for changing virtual equilibrium point rather than the motor control process for recruiting motor neurons to regulate grasping force.

Deep Heterogeneous Dilation of LSTM for Transient-phase Gesture Prediction through High-density Electromyography: Application in Neurorobotics

Deep networks have been recently proposed to estimate motor intention using conventional bipolar surface electromyography (sEMG) signals for myoelectric control of neurorobots. In this regard, deepnets are generally challenged by long training times (affecting the practicality and calibration), complex model architectures (affecting the predictability of the outcomes), a large number of trainable parameters (increasing the need for big data), and possibly overfitting. Capitalizing on our recent work on homogeneous temporal dilation in a Recurrent Neural Network (RNN) model, this paper proposes, for the first time, heterogeneous temporal dilation in an LSTM model and applies that to high-density surface electromyography (HD-sEMG), allowing for decoding dynamic temporal dependencies with tunable temporal foci. In this paper, a 128-channel HD-sEMG signal space is considered due to the potential for enhancing the spatiotemporal resolution of human-robot interfaces. Accordingly, this paper addresses a challenging motor intention decoding problem of neurorobots, namely, transient intention identification. The aforementioned problem only takes into account the dynamic and transient phase of gesture movements when the signals are not stabilized or plateaued, addressing which can significantly enhance the temporal resolution of human-robot interfaces. This would eventually enhance seamless real-time implementations. Additionally, this paper introduces the concept of dilation foci to modulate the modeling of temporal variation in transient phases. In this work a high number (i.e. 65) of gestures is included, which adds to the complexity and significance of the understudied problem. Our results show state-of-the-art performance for gesture prediction in terms of accuracy, training time, and model convergence.

Motor Intention Decoding from the Upper Limb by Graph Convolutional Network Based on Functional Connectivity

Decoding brain intention from noninvasively measured neural signals has recently been a hot topic in brain-computer interface (BCI). The motor commands about the movements of fine parts can increase the degrees of freedom under control and be applied to external equipment without stimulus. In the decoding process, the classifier is one of the key factors, and the graph information of the EEG was ignored by most researchers. In this paper, a graph convolutional network (GCN) based on functional connectivity was proposed to decode the motor intention of four fine parts movements (shoulder, elbow, wrist, hand). First, event-related desynchronization was analyzed to reveal the differences between the four classes. Second, functional connectivity was constructed by using synchronization likelihood (SL), phase-locking value (PLV), H index (H), mutual information (MI), and weighted phase-lag index (WPLI) to acquire the electrode pairs with a difference. Subsequently, a GCN and convolutional neural networks (CNN) were performed based on functional topological structures and time points, respectively. The results demonstrated that the proposed method achieved a decoding accuracy of up to 92.81% in the four-class task. Besides, the combination of GCN and functional connectivity can promote the development of BCI.

A Virtual Reality Algorithm for the Study of Clinical Efficacy of Sports Injury Rehabilitation Training

This paper aims to explore the influence of virtual reality technology interactive rehabilitation training system and PT and OT operation therapy on the exercise function, daily life activity ability (ADL), and the quality of life in patients with a sports injury. In this context, this paper mainly designed three experiments to test the virtual rehabilitation system: two action experiments (experiment 1), two experiments with actions in 3 different positions (experiment 2), and five different actions (experiment 3), and the motion intention recognition rate, average total time, and task completion degree of the three experiments were calculated. The virtual scene and hardware equipment were kept stable, and the human-machine interaction effect was good. The effectiveness of the proposed virtual reality rehabilitation training system is demonstrated from other aspects. The results showed that the average completion time of 5 volunteers was 57.72 seconds, with an average offline accuracy of 89.03%. In experiment 2, the five volunteers averaged 54.98 seconds, with an average offline accuracy of 91.73%. The average recognition accuracy of the training system reached 90%, demonstrating the effectiveness of the virtual reality rehabilitation training system in terms of motor intention recognition rate, average total use time, and task completion.

Increasing Motor Cortex Activation During Grasping Via Novel Robotic Mirror Hand Therapy: A Pilot fNIRS Study

Background: Mirror therapy (MT) has been used for functional recovery of the affected hand by providing the mirrored image of the unaffected hand movement, which induces neural activation of the contralateral cortical hemisphere. Recently, many wearable robots assisting the movement of the hand have been developed, and several studies have proposed robotic mirror therapy (RMT) that provides mirrored movements of the unaffected hand on the affected hand with the robot controlled by electromyography or posture of the unaffected hand. There have been limited evaluations of the cortical activity during RMT compared to MT and robotic therapy (RT) providing passive movements despite the difference in the modality of sensory feedback and the involvement of motor intention, respectively. Methods: This paper analyzes bilateral motor cortex activation in nine healthy subjects and five chronic stroke survivors during a pinching task performed in MT, RT, and RMT conditions using functional near infrared spectroscopy (fNIRS). In the MT condition, the person moved the unaffected hand and observed it in a mirror while the affected hand remained still. In RT condition passive movements were provided to the affected hand with a cable-driven soft robotic glove, while, in RMT condition, the posture of the unaffected hand was measured by a sensing glove and the soft robotic glove mirrored its movement on the affected hand. Results: For both groups, the RMT condition showed the greatest mean cortical activation on the contralateral motor cortex compared to other conditions. Individual results indicate that RMT induces similar or greater neural activation on the motor cortex compared to MT and RT conditions. The interhemispheric activations of both groups were balanced in RMT condition. In MT condition, significantly greater activation was shown on the ipsilateral side for both subject groups, while the contralateral side showed significantly greater activation for healthy in RT condition. Conclusion: The experimental results indicate that combining visual feedback, somatosensory feedback, and motor intention are important for greater stimulation on the contralateral motor cortex of the affected hand. RMT that includes these factors is hypothesized to achieve a more effective functional rehabilitation due to greater and more balanced cortical activation.

Brain-Computer Interfaces Systems for Upper and Lower Limb Rehabilitation: A Systematic Review

In recent years, various studies have demonstrated the potential of electroencephalographic (EEG) signals for the development of brain-computer interfaces (BCIs) in the rehabilitation of human limbs. This article is a systematic review of the state of the art and opportunities in the development of BCIs for the rehabilitation of upper and lower limbs of the human body. The systematic review was conducted in databases considering using EEG signals, interface proposals to rehabilitate upper/lower limbs using motor intention or movement assistance and utilizing virtual environments in feedback. Studies that did not specify which processing system was used were excluded. Analyses of the design processing or reviews were excluded as well. It was identified that 11 corresponded to applications to rehabilitate upper limbs, six to lower limbs, and one to both. Likewise, six combined visual/auditory feedback, two haptic/visual, and two visual/auditory/haptic. In addition, four had fully immersive virtual reality (VR), three semi-immersive VR, and 11 non-immersive VR. In summary, the studies have demonstrated that using EEG signals, and user feedback offer benefits including cost, effectiveness, better training, user motivation and there is a need to continue developing interfaces that are accessible to users, and that integrate feedback techniques.