The steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) has received extensive attention in research for the less training time, excellent recognition performance, and high information translate rate. At present, most of the powerful SSVEPs detection methods are similarity measurements based on spatial filters and Pearson's correlation coefficient. Among them, the task-related component analysis (TRCA)-based method and its variant, the ensemble TRCA (eTRCA)-based method, are two methods with high performance and great potential. However, they have a defect, that is, they can only suppress certain kinds of noise, but not more general noises. To solve this problem, a novel time filter was designed by introducing the temporally local weighting into the objective function of the TRCA-based method and using the singular value decomposition. Based on this, the time filter and (e)TRCA-based similarity measurement methods were proposed, which can perform a robust similarity measure to enhance the detection ability of SSVEPs. A benchmark dataset recorded from 35 subjects was used to evaluate the proposed methods and compare them with the (e)TRCA-based methods. The results indicated that the proposed methods performed significantly better than the (e)TRCA-based methods. Therefore, it is believed that the proposed time filter and the similarity measurement methods have promising potential for SSVEPs detection.

Objective: Tactile brain-computer interface (BCI) systems can provide new communication and control options for patients with impairments of eye movements or vision. One of the most common modalities used in these BCIs is the P300 potential. Until now, tactile P300 BCIs have been successfully constructed by situating tactile stimuli at various parts of the human body. This study proposed a novel tactile P300 BCI paradigm for further expanding the tactile stimulation methods. Methods: In our proposed paradigm, the spatial target vibrotactile stimuli were delivered to subject's left and right cheeks. To validate the feasibility of our proposed paradigm, a traditional tactile P300 BCI paradigm employing spatial target vibrotactile stimuli to subject's left and right wrists was used for comparison. Results: The experimental results of nine healthy subjects demonstrated that the proposed paradigm could obtain significantly higher classification accuracy and information transfer rate than the traditional paradigm (both for p < 0.05). Furthermore, the subjective feedback showed that our proposed paradigm was more favored by the subjects compared to the traditional paradigm, and most subjects reported that the new paradigm helped them easily distinguish between targets and non-targets. Conclusion: The proposed tactile P300 BCI paradigm is feasible, and can bring about superior performance and use-evaluation. Significance: The new paradigm might lead to many promising applications of such BCIs.

Motor imagery, one of the main brain-computer interface (BCI) paradigms, has been extensively utilized in numerous BCI applications, such as the interaction between disabled people and external devices. Precise decoding, one of the most significant aspects of realizing efficient and stable interaction, has received a great deal of intensive research. However, the current decoding methods based on deep learning are still dominated by single-scale serial convolution, which leads to insufficient extraction of abundant information from motor imagery signals. To overcome such challenges, we propose a new end-to-end convolutional neural network based on multiscale spatial-temporal feature fusion (MSTFNet) for EEG classification of motor imagery. The architecture of MSTFNet consists of four distinct modules: feature enhancement module, multiscale temporal feature extraction module, spatial feature extraction module and feature fusion module, with the latter being further divided into the depthwise separable convolution block and efficient channel attention block. Moreover, we implement a straightforward yet potent data augmentation strategy to bolster the performance of MSTFNet significantly. To validate the performance of MSTFNet, we conduct cross-session experiments and leave-one-subject-out experiments. The cross-session experiment is conducted across two public datasets and one laboratory dataset. On the public datasets of BCI Competition IV 2a and BCI Competition IV 2b, MSTFNet achieves classification accuracies of 83.62% and 89.26%, respectively. On the laboratory dataset, MSTFNet achieves 86.68% classification accuracy. Besides, the leave-one-subject-out experiment is performed on the BCI Competition IV 2a dataset, and MSTFNet achieves 66.31% classification accuracy. These experimental results outperform several state-of-the-art methodologies, indicate the proposed MSTFNet's robust capability in decoding EEG signals associated with motor imagery.

Motor imagery-based brain-computer interfaces (MI-BCIs) features are generally extracted from a wide fixed frequency band and time window of EEG signal. The performance suffers from individual differences in corresponding time to MI tasks. In order to solve the problem, in this study, we propose a novel method named Riemannian sparse optimization and Dempster-Shafer fusion of multi-time-frequency patterns (RSODSF) to enhance the decoding efficiency. First, we effectively combine the Riemannian geometry of the spatial covariance matrix with sparse optimization to extract more robust and distinct features. Second, the Dempster-Shafer theory is introduced and used to fuse each time window after sparse optimization of Riemannian features. Besides, the probabilistic values of the support vector machine (SVM) are obtained and transformed to effectively fuse multiple classifiers to leverage potential soft information of multiple trained SVM. The openaccess BCI Competition IV dataset IIa and Competition III dataset IIIa are employed to evaluate the performance of the proposed RSODSF. It achieves higher average accuracy (89.7% and 96.8%) than state-of-the-art methods. The improvement over the common spatial patterns (SFBCSP) are respectively 9.9% and 12.4% (p<0.01, paired t-test). These results show that our proposed RSODSF method is a promising candidate for the performance improvement of MI-BCI.

The evaluation of the level of consciousness in patients with disorders of consciousness (DOC) is primarily based on behavioural assessments. Patients with unresponsive wakefulness syndrome (UWS) do not show any sign of awareness of their environment, while minimally conscious state (MCS) patients show reproducible but fluctuating signs of awareness. Some patients, although with remaining cognitive abilities, are not able to exhibit overt voluntary responses at the bedside and may be misdiagnosed as UWS. Several studies investigated functional neuroimaging and neurophysiology as an additional tool to evaluate the level of consciousness and to detect covert command following in DOC. Most of these studies are based on auditory stimulation, neglecting patients suffering from decreased or absent hearing abilities. In the present study, we aim to assess the response to a P3-based paradigm in 40 patients with DOC and 12 healthy participants using auditory (AEP) and vibrotactile (VTP) stimulation. To this end, an EEG-based brain-computer interface was used at DOC patient’s bedside. We compared the significance of the P3 performance (i.e., the interpretation of significance of the evoked P3 response) as obtained by ‘direct processing’ (i.e., theoretical-based significance threshold) and ‘offline processing’ (i.e., permutation-based single subject level threshold). We evaluated whether the P3 performances were dependent on clinical variables such as diagnosis (UWS and MCS), aetiology and time since injury. Last we tested the dependency of AEP and VTP performances at the single subject level. Direct processing tends to overestimate P3 performance. We did not find any difference in the presence of a P3 performance according to the level of consciousness (UWS vs. MCS) or the aetiology (traumatic vs. non-traumatic brain injury). The performance achieved at the AEP paradigm was independent from what was achieved at the VTP paradigm, indicating that some patients performed better on the AEP task while others performed better on the VTP task. Our results support the importance of using multimodal approaches in the assessment of DOC patients in order to optimise the evaluation of patient’s abilities.

In this study, we present a novel multi-class brain-computer interface (BCI) for communication and control. In this system, the information processing is shared by the algorithm (computer) and the user (human). Specifically, an electro-tactile cycle was presented to the user, providing the choice (class) by delivering timely sensory input. The user discriminated these choices by his/her endogenous sensory ability and selected the desired choice with an intuitive motor task. This selection was detected by a fast brain switch based on real-time detection of movement-related cortical potentials from scalp EEG. We demonstrated the feasibility of such a system with a four-class BCI, yielding a true positive rate of ~ 80% and ~ 70%, and an information transfer rate of ~ 7 bits/min and ~ 5 bits/min, for the movement and imagination selection command, respectively. Furthermore, when the system was extended to eight classes, the throughput of the system was improved, demonstrating the capability of accommodating a large number of classes. Combining the endogenous sensory discrimination with the fast brain switch, the proposed system could be an effective, multi-class, gaze-independent BCI system for communication and control applications.