High Information Transfer Rate Research Articles

Optimizing Stimulus Frequency Ranges for Building a High-Rate High Frequency SSVEP-BCI.

The brain-computer interfaces (BCIs) based on steady-state visual evoked potential (SSVEP) have been extensively explored due to their advantages in terms of high communication speed and smaller calibration time. The visual stimuli in the low- and medium-frequency ranges are adopted in most of the existing studies for eliciting SSVEPs. However, there is a need to further improve the comfort of these systems. The high-frequency visual stimuli have been used to build BCI systems and are generally considered to significantly improve the visual comfort, but their performance is relatively low. The distinguishability of 16-class SSVEPs encoded by the three frequency ranges, i.e., 31-34.75 Hz with an interval of 0.25 Hz, 31-38.5 Hz with an interval of 0.5 Hz, 31-46 Hz with an interval of 1 Hz, is explored in this study. We compare classification accuracy and information transfer rate (ITR) of the corresponding BCI system. According to the optimized frequency range, this study builds an online 16-target high frequency SSVEP-BCI and verifies the feasibility of the proposed system based on 21 healthy subjects. The BCI based on visual stimuli with the narrowest frequency range, i.e., 31-34.5 Hz, have the highest ITR. Therefore, the narrowest frequency range is adopted to build an online BCI system. An averaged ITR obtained from the online experiment is 153.79 ± 6.39 bits/min. These findings contribute to the development of more efficient and comfortable SSVEP-based BCIs.

Representative-based Cold Start for Adaptive SSVEP-BCI.

The tradeoff between calibration effort and model performance still hinders the user experience for steady-state visual evoked brain-computer interfaces (SSVEP-BCI). To address this issue and improve model generalizability, this work investigated the adaptation from the cross-dataset model to avoid the training process, while maintaining high prediction ability. When a new subject enrolls, a group of user-independent (UI) models is recommended as the representative from a multi-source data pool. The representative model is then augmented with online adaptation and transfer learning techniques based on user-dependent (UD) data. The proposed method is validated on both offline (N=55) and online (N=12) experiments. Compared with the UD adaptation, the recommended representative model relieved approximately 160 trials of calibration efforts for a new user. In the online experiment, the time window decreased from 2 s to 0.56±0.2 s, while maintaining high prediction accuracy of 0.89-0.96. Finally, the proposed method achieved the average information transfer rate (ITR) of 243.49 bits/min, which is the highest ITR ever reported in a complete calibration-free setting. The results of the offline result were consistent with the online experiment. Representatives can be recommended even in a cross-subject/device/session situation. With the help of represented UI data, the proposed method can achieve sustained high performance without a training process. This work provides an adaptive approach to the transferable model for SSVEP-BCIs, enabling a more generalized, plug-and-play and high-performance BCI free of calibrations.

An efficient CNN-LSTM network with spectral normalization and label smoothing technologies for SSVEP frequency recognition

Objective. Steady-state visual evoked potentials (SSVEPs) based brain–computer interface (BCI) has received great interests owing to the high information transfer rate and available large number of targets. However, the performance of frequency recognition methods heavily depends on the amount of the calibration data for intra-subject classification. Some research adopted the deep learning (DL) algorithm to conduct the inter-subject classification, which could reduce the calculation procedure, but the performance still has large room to improve compared with the intra-subject classification. Approach. To address these issues, we proposed an efficient SSVEP DL NETwork (termed SSVEPNET) based on one-dimensional convolution and long short-term memory (LSTM) module. To enhance the performance of SSVEPNET, we adopted the spectral normalization and label smoothing technologies during implementing the network architecture. We evaluated the SSVEPNET and compared it with other methods for the intra- and inter-subject classification under different conditions, i.e. two datasets, two time-window lengths (1 s and 0.5 s), three sizes of training data. Main results. Under all the experimental settings, the proposed SSVEPNET achieved the highest average accuracy for the intra- and inter-subject classification on the two SSVEP datasets, when compared with other traditional and DL baseline methods. Significance. The extensive experimental results demonstrate that the proposed DL model holds promise to enhance frequency recognition performance in SSVEP-based BCIs. Besides, the mixed network structures with convolutional neural network and LSTM, and the spectral normalization and label smoothing could be useful optimization strategies to design efficient models for electroencephalography data.

Bidirectional Siamese correlation analysis method for enhancing the detection of SSVEPs

Objective. Steady-state visual evoked potential (SSVEP)-based brain–computer interfaces (BCIs) have attracted increasing attention due to their high information transfer rate. To improve the performance of SSVEP detection, we propose a bidirectional Siamese correlation analysis (bi-SiamCA) model. Approach. In this model, an long short-term memory-based Siamese architecture is designed to measure the similarity between the SSVEP signal and the template in each frequency and obtain the probability that the SSVEP signal belongs to each frequency. Additionally, a maximize agreement module with a designed contrastive loss is adopted in the Siamese architecture to increase the similarity between the SSVEP signal and the reference signal in the same frequency. Moreover, a two-way signal processing mechanism is built to effectively integrate complementary information from two temporal directions of the input signals. Our model uses raw SSVEPs as inputs and can be trained end-to-end. Main results. Experimental results on a 40-class dataset and a 12-class dataset indicate that bi-SiamCA can significantly improve the classification accuracy compared with the prominent traditional and deep learning methods, especially under short data lengths. Feature visualizations show that the similarity between the SSVEP signal and the reference signal in the same frequency gradually improved in our model. Conclusion. The proposed bi-SiamCA model enhances the performance of SSVEP detection and outperforms the compared methods. Significance. Due to its high decoding accuracy under short signals, our approach has great potential to implement a high-speed SSVEP-based BCI.

Advances in Wearable Brain-Computer Interfaces From an Algorithm-Hardware Co-Design Perspective

Brain-computer interface (BCI), a communication technology between brain and computer developed for a long time since the 1970s, can be incorporated into wearable devices by developing powerful signal processing algorithms and semiconductor technologies. For a satisfactory user experience based on BCI, high information transfer rate and low power consumption should be considered together without losing accuracy. Although many existing BCI algorithms have been mainly focused solely on the accuracy, their deployment on wearable devices is not straightforward due to the limited hardware resources and computational capabilities. This tutorial summarizes recent advances in wearable BCI algorithms and hardware implementations from an algorithm-hardware co-design perspective and discusses future directions.

Cross-subject fusion based on time-weighting canonical correlation analysis in SSVEP-BCIs

Brain–computer interface technology provides new possibilities for medical rehabilitation and human–computer interaction. The steady-state visual evoked potential based brain–computer interface (SSVEP-BCI) is the preferred solution for controlling equipment because of its ease of operation, low training requirements, and high information transfer rate (ITR). However, the low recognition accuracy of SSVEP in short time limits its further improvement in ITR. To address this issue, this paper proposes time-weighting canonical correlation analysis (TWCCA), a time-domain enhancement CCA-based method to improve the recognition accuracy of SSVEP in short recognition time. The TWCCA method evaluates the SSVEP components of EEG signals at different time periods and performs a time-dimension weighted CCA on the original signal. The method integrates the features of all stimulus targets based on CCA and is insensitive to the specific number of targets, so it does not require a large amount of calibration data. To further shorten the calibration time for a certain user, this paper also proposes a cross-subject fusion method to calibrate the parameters of the target user using the acquired data from other users. The effectiveness of the proposed method is evaluated using EEG signals from the occipital and ear regions. The experiment results show that the TWCCA method significantly improves the average recognition accuracy by 3.86% and the cross-subject fusion method shortens the calibration time by 78.5% without compromising recognition accuracy. The proposed combination of cross-subject fusion method and TWCCA (CF-TWCCA) increases recognition accuracy by 13.16% on average (p<0.001).

SSVEP-based brain–computer interface for music using a low-density EEG system

ABSTRACT In this paper, we present a bespoke brain–computer interface (BCI), which was developed for a person with severe motor-impairments, who was previously a Violinist, to allow performing and composing music at home. It uses steady-state visually evoked potential (SSVEP) and adopts a dry, low-density, and wireless electroencephalogram (EEG) headset. In this study, we investigated two parameters: (1) placement of the EEG headset and (2) inter-stimulus distance and found that the former significantly improved the information transfer rate (ITR). To analyze EEG, we adopted canonical correlation analysis (CCA) without weight-calibration. The BCI for musical performance realized a high ITR of 37.59 ± 9.86 bits min−1 and a mean accuracy of 88.89 ± 10.09%. The BCI for musical composition obtained an ITR of 14.91 ± 2.87 bits min−1 and a mean accuracy of 95.83 ± 6.97%. The BCI was successfully deployed to the person with severe motor-impairments. She regularly uses it for musical composition at home, demonstrating how BCIs can be translated from laboratories to real-world scenarios.

Overview of the winning approaches in BCI Controlled Robot Contest in World Robot Contest 2021: Calibration-free SSVEP

Recently, steady-state visual evoked potential (SSVEP) has become one of the most popular electroencephalography paradigms due to its high information transfer rate. Several approaches have been proposed to improve the performance of SSVEP. The calibration-free scenario is significant in SSVEP-based brain–computer interface systems, where the subject is the first time to use the system. The participating teams proposed several effective calibration-free algorithm frameworks in the SSVEP competition (calibration-free) of the BCI Controlled Robot Contest in World Robot Contest 2021. This paper introduces the approaches used in the algorithms of the top five teams in the final. The results of the five subjects in the final proved the effectiveness of the approaches. This paper discusses the effectiveness of each approach in improving the system performance in the calibration-free scenario and gives suggestions on how to use these approaches in a real-world system.

The effect of stimulus number on the recognition accuracy and information transfer rate of SSVEP–BCI in augmented reality

Objective. The biggest advantage of steady-state visual evoked potential (SSVEP)-based brain–computer interface (BCI) lies in its large command set and high information transfer rate (ITR). Almost all current SSVEP–BCIs use a computer screen (CS) to present flickering visual stimuli, which limits its flexible use in actual scenes. Augmented reality (AR) technology provides the ability to superimpose visual stimuli on the real world, and it considerably expands the application scenarios of SSVEP–BCI. However, whether the advantages of SSVEP–BCI can be maintained when moving the visual stimuli to AR glasses is not known. This study investigated the effects of the stimulus number for SSVEP–BCI in an AR context. Approach. We designed SSVEP flickering stimulation interfaces with four different numbers of stimulus targets and put them in AR glasses and a CS to display. Three common recognition algorithms were used to analyze the influence of the stimulus number and stimulation time on the recognition accuracy and ITR of AR–SSVEP and CS–SSVEP. Main results. The amplitude spectrum and signal-to-noise ratio of AR–SSVEP were not significantly different from CS–SSVEP at the fundamental frequency but were significantly lower than CS–SSVEP at the second harmonic. SSVEP recognition accuracy decreased as the stimulus number increased in AR–SSVEP but not in CS–SSVEP. When the stimulus number increased, the maximum ITR of CS–SSVEP also increased, but not for AR–SSVEP. When the stimulus number was 25, the maximum ITR (142.05 bits min−1) was reached at 400 ms. The importance of stimulation time in SSVEP was confirmed. When the stimulation time became longer, the recognition accuracy of both AR–SSVEP and CS–SSVEP increased. The peak value was reached at 3 s. The ITR increased first and then slowly decreased after reaching the peak value. Significance. Our study indicates that the conclusions based on CS–SSVEP cannot be simply applied to AR–SSVEP, and it is not advisable to set too many stimulus targets in the AR display device.

Classification of code-modulated visual evoked potentials using adaptive modified covariance beamformer and EEG signals

Objective: In general, brain computer interface (BCI) studies based on code-modulated Visual Evoked Potentials (c-VEP) use m-sequences to decode EEG responses to visual stimuli. BCI systems based on the c-VEP paradigm can simultaneously present a large number of commands, which results in a significantly high information transfer rate (ITR). Spatiotemporal beamforming (STB) is one of the commonly used approaches in c-VEP-based BCI systems. Approach: In the current work, a novel STB-based technique is proposed to detect the gazed targets. The proposed method improves the performance of conventional STB-based techniques by providing a robust estimation of the covariance matrix in short stimulation times. Different user parameter-free methods, including the convex combination (CC), the general linear combination (GLC), and the modified versions of these techniques, are used to estimate a reliable and robust covariance matrix when a small number of repetitions are available. Main results: The stimulus presentation rate of 120 Hz is used to assess the performance of the proposed structures. Our proposed methods improved the classification accuracy by an average of 20% compared to the conventional STB method at the shortest stimulation time. The proposed method achieves an average ITR of 157.07 bits/min by using only two repetitions of the m-sequences. Significance: The results show that our proposed methods perform significantly better than the conventional STB technique in all stimulation times.

Investigating the Effect of Flickering Frequency Pair and Mother Wavelet Selection in Steady-State Visually-Evoked Potentials on Two-Command Brain-Computer Interfaces

IntroductionSteady-state visually evoked potentials (SSVEPs) have become popular in brain-computer interface (BCI) applications in addition to many other applications on clinical neuroscience (neurodegenerative disorders, schizophrenia, epilepsy, etc.), cognitive (visual attention, working memory, brain rhythms, etc.), and use of engineering researches. Among available methods to measure brain activities, SSVEPs have advantages like higher information transfer rate, simplicity in structure, and short training time. SSVEP-based BCIs use flickering stimuli at different frequencies to discriminate distinct commands in real life. Some features are extracted from the SSVEP signals before these commands are classified. The wavelet transform (WT) has attracted researchers among feature extraction methods since it utilizes the non-stationary signals well. In the WT, a sample function (named mother wavelet) represents the SSVEP signal in both time and frequency domains. Unfortunately, there is no universal mother wavelet function that fits all the signals. Therefore, choosing an appropriate mother wavelet function may be a challenge in WT-related studies. Although there are such studies in three- and seven-command SSVEP-based studies, there is no study for two-command systems in our knowledge. Materials and MethodsIn this study, two user commands flickered at the combinations of seven different frequencies were tested to determine which frequency pairs give the highest performance. For this purpose, three well-known wavelet features (energy, entropy, and variance) were calculated for each of derived EEG frequency bands from the discrete WT coefficients of SSVEP signals. The WT was repeated for six different mother wavelet functions (Haar, Db4, Sym4, Coif1, Bior3.5, and Rbior2.8). Then, four feature sets (every three features, and all together) were applied to seven commonly-used machine learning algorithms (Decision Tree, Discriminant Analysis, Logistic Regression, Naive Bayes, Support Vector Machines, Nearest Neighbors, and Ensemble Classifiers). Results and DiscussionWe achieved 100% accuracies among these 3,528 runs (7 classifiers x 4 feature sets x 6 mother wavelets x 21 flickering frequency pairs) using the mother wavelet function of Haar and the Ensemble Learner classifier. The highest classifier performances are 100% when two commands have the flickering frequency pairs of (6.0 and 10 Hz), (6.5 and 8.2 Hz), or (6.5 and 10.0 Hz). ConclusionWe obtained three main outcomes from this study. First, the most representative mother wavelet function was Haar, while the worst one was Symlet 4. Second, the Ensemble Learner classifier gave the maximum classifier performance in a two-command SSVEP-based BCI system. Besides, two user commands from SSVEP should be one of the frequency pairs of (6.0 and 10.0 Hz), (6.5 and 8.2 Hz), and (6.5 and 10.0 Hz) to achieve the maximum accuracy.

A 120-target brain-computer interface based on code-modulated visual evoked potentials

BackgroundIn recent years, numerous studies on the brain-computer interface (BCI) have been published. However, the number of targets in most of the existing studies was not enough for many practical applications. New methodTo achieve highly efficient communications, this study proposed a 120-target BCI system based on code-modulated visual evoked potentials (c-VEPs). Four 31-bit pseudorandom codes were used, and each code generated 30 targets by cyclic shift with a lag of 1 bit. ResultsIn the online experiments, subjects could select one target in 1.04 s (0.52 s for stimulation and 0.52 s for gaze shifting) with an average information transfer rate (ITR) of 265.74 bits/min. Comparison with existing methodThe proposed system achieved more targets and higher ITR than other recent c-VEP based studies. which attributes to the optimal code combination and the 1-bit lag. ConclusionThe results illustrate that the proposed BCI system can achieve a high ITR with a short stimulation time. In addition, the c-VEP paradigm can shorten the training time, which ensures practicality in real applications.

A comparative survey of SSVEP recognition algorithms based on template matching of training trials

PurposeSteady-state visual evoked potential (SSVEP) has been widely used in the application of electroencephalogram (EEG) based non-invasive brain computer interface (BCI) due to its characteristics of high accuracy and information transfer rate (ITR). To recognize the SSVEP components in collected EEG trials, a lot of recognition algorithms based on template matching of training trials have been proposed and applied in recent years. In this paper, a comparative survey of SSVEP recognition algorithms based on template matching of training trails has been done.Design/methodology/approachTo survey and compare the recently proposed recognition algorithms for SSVEP, this paper regarded the conventional canonical correlated analysis (CCA) as the baseline, and selected individual template CCA (ITCCA), multi-set CCA (MsetCCA), task related component analysis (TRCA), latent common source extraction (LCSE) and a sum of squared correlation (SSCOR) for comparison.FindingsFor the horizontal comparative of the six surveyed recognition algorithms, this paper adopted the “Tsinghua JFPM-SSVEP” data set and compared the average recognition performance on such data set. The comparative contents including: recognition accuracy, ITR, correlated coefficient and R-square values under different time duration of the SSVEP stimulus presentation. Based on the optimal time duration of stimulus presentation, the author has also compared the efficiency of the six compared algorithms. To measure the influence of different parameters, the number of training trials, the number of electrodes and the usage of filter bank preprocessing were compared in the ablation study.Originality/valueBased on the comparative results, this paper analyzed the advantages and disadvantages of the six compared SSVEP recognition algorithms by considering application scenes, real-time and computational complexity. Finally, the author gives the algorithms selection range for the recognition of real-world online SSVEP-BCI.

Evaluation of an Ecological Interface Design–Driven Augmentative and Alternative Communication Interface

This study evaluated the change in usability, mental workload and information transfer rate associated with an augmentative and alternative communication (AAC) interface designed through ecological interface design (EID). The design and development process is detailed in Shea et al. (2021) . Digital AAC interfaces are considered high-tech interventions for individuals who experience complex communication needs (e.g., from etiologies such as cerebral palsy) and enable users to select language options from a visual display. Interface usability, mental workload and information transfer rate collectively influence users’ communication. Ten AAC-naïve participants engaged in three semi–scripted conversations (verbal, AAC-mediated commercial interface, and EID interfaces) with an actor. Augmentative and alternative communication interfaces were accessed through a single switch pathway. Information transfer rate, error rate, heart rate variability and subjective workload performance measures were recorded for every trial. During AAC-mediated trials, interface interactions were also documented. The EID AAC interface presented improved communication in 5 out of 7 performance measures ( p &lt; .05). The EID AAC interface was associated with a significantly higher information transfer rate, lower error rate, less time elapsed between switch activations, less switch activation per word communicated and lower subjective workload. The application of EID to an AAC interface can lead to a significantly improved communication experience.

Feasibility of a Wireless Implantable Multi-electrode System for High-bandwidth Prosthetic Interfacing: Animal and Cadaver Study.

Currently used prosthetic solutions in upper extremity amputation have limited functionality, owing to low information transfer rates of neuromuscular interfacing. Although surgical innovations have expanded the functional potential of the residual limb, available interfaces are inefficacious in translating this potential into improved prosthetic control. There is currently no implantable solution for functional interfacing in extremity amputation which offers long-term stability, high information transfer rates, and is applicable for all levels of limb loss. In this study, we presented a novel neuromuscular implant, the the Myoelectric Implantable Recording Array (MIRA). To our knowledge, it is the first fully implantable system for prosthetic interfacing with a large channel count, comprising 32 intramuscular electrodes. The purpose of this study was to evaluate the MIRA in terms of biocompatibility, functionality, and feasibility of implantation to lay the foundations for clinical application. This was achieved through small- and large-animal studies as well as test surgeries in a human cadaver. We evaluated the biocompatibility of the system's intramuscular electromyography (EMG) leads in a rabbit model. Ten leads as well as 10 pieces of a biologically inert control material were implanted into the paravertebral muscles of four animals. After a 3-month implantation, tissue samples were taken and histopathological assessment performed. The probes were scored according to a protocol for the assessment of the foreign body response, with primary endpoints being inflammation score, tissue response score, and capsule thickness in µm. In a second study, chronic functionality of the full system was evaluated in large animals. The MIRA was implanted into the shoulder region of six dogs and three sheep, with intramuscular leads distributed across agonist and antagonist muscles of shoulder flexion. During the observation period, regular EMG measurements were performed. The implants were removed after 5 to 6 months except for one animal, which retained the implant for prolonged observation. Primary endpoints of the large-animal study were mechanical stability, telemetric capability, and EMG signal quality. A final study involved the development of test surgeries in a fresh human cadaver, with the goal to determine feasibility to implant relevant target muscles for prosthetic control at all levels of major upper limb amputation. Evaluation of the foreign body reaction revealed favorable biocompatibility and a low-grade tissue response in the rabbit study. No differences regarding inflammation score (EMG 4.60 ± 0.97 [95% CI 4.00 to 5.20] versus control 4.20 ± 1.48 [95% CI 3.29 to 5.11]; p = 0.51), tissue response score (EMG 4.00 ± 0.82 [95% CI 3.49 to 4.51] versus control 4.00 ± 0.94 [95% CI 3.42 to 4.58]; p > 0.99), or thickness of capsule (EMG 19.00 ± 8.76 µm [95% CI 13.57 to 24.43] versus control 29.00 ± 23.31 µm [95% CI 14.55 to 43.45]; p = 0.29) were found compared with the inert control article (high-density polyethylene) after 3 months of intramuscular implantation. Throughout long-term implantation of the MIRA in large animals, telemetric communication remained unrestricted in all specimens. Further, the implants retained the ability to record and transmit intramuscular EMG data in all animals except for two sheep where the implants became dislocated shortly after implantation. Electrode impedances remained stable and below 5 kΩ. Regarding EMG signal quality, there was little crosstalk between muscles and overall average signal-to-noise ratio was 22.2 ± 6.2 dB. During the test surgeries, we found that it was possible to implant the MIRA at all major amputation levels of the upper limb in a human cadaver (the transradial, transhumeral, and glenohumeral levels). For each level, it was possible to place the central unit in a biomechanically stable environment to provide unhindered telemetry, while reaching the relevant target muscles for prosthetic control. At only the glenohumeral level, it was not possible to reach the teres major and latissimus dorsi muscles, which would require longer lead lengths. As assessed in a combination of animal model and cadaver research, the MIRA shows promise for clinical research in patients with limb amputation, where it may be employed for all levels of major upper limb amputation to provide long-term stable intramuscular EMG transmission. In our study, the MIRA provided high-bandwidth prosthetic interfacing through intramuscular electrode sites. Its high number of individual EMG channels may be combined with signal decoding algorithms for accessing spinal motor neuron activity after targeted muscle reinnervation, thus providing numerous degrees of freedom. Together with recent innovations in amputation surgery, the MIRA might enable improved control approaches for upper limb amputees, particularly for patients with above-elbow amputation where the mismatch between available control signals and necessary degrees of freedom for prosthetic control is highest.

A classification algorithm of an SSVEP brain-Computer interface based on CCA fusion wavelet coefficients

BackgroundIn the study of brain-computer interfaces (BCIs) based on steady-state visual evoked potentials (SSVEPs), how to improve the classification accuracies of BCIs has always been the focus of researchers. Canonical correlation analysis (CCA) is widely used in BCI systems of SSVEPs because of its rapidity and scalability. However, the classical CCA algorithm always encounters the difficulty of low accuracy in a short time. New methodFor targetless stimuli, this paper proposes a fusion algorithm (CCA-CWT-SVM) that is combined with CCA, a continuous wavelet transform, and a support vector machine (SVM) to improve the low classification accuracies when a single feature extraction method is used. ResultsThis fusion algorithm achieves high accuracies and information transfer rates (ITRs) in the SSVEP paradigm with few targets. Comparison with existing methods and conclusionsThrough the study of 400 groups of experimental data from 10 subjects, the results show that CCA-CWT-SVM has a classification accuracy of 91.76% within 2 s and an ITR of 48.92 bits/min, which are 10.88% and 13.18 bits/min higher than those of the standard CCA. Compared with a mainstream EEG decoding algorithm, filter bank canonical correlation analysis (FBCCA), the classification accuracy and ITR of the CCA-CWT-SVM algorithm also improved (4.45% and 5.69 bit/min, respectively). Using a dataset from Tsinghua University (THU), we also showed that the fusion algorithm is better than the classical algorithms. The CCA-CWT-SVM algorithm obtained an 89.1% accuracy and a 39.91 bit/min ITR in a time window of 2 s. The results were significantly improved compared with those of CCA and the FBCCA (CCA: 79.44% and 28.23 bits/min, FBCCA: 84.03% and 33.4 bits/min). Hence, this work provides an experimental basis for designing an SSVEP-based BCI system with a high task classification accuracy in some crucial biomedical applications.

Design considerations for the auditory brain computer interface speller

Brain-computer interface (BCI) systems are designed to translate the measured brain signals into a set of instructions, with a purpose of interaction with outside world. Most common application of such systems is to provide aid to patients with limited physical capabilities, where the visual stimulation based BCI are most commonly used. However, in the most extreme situations, e.g. for patients with amyotrophic lateral sclerosis, visual stimulation approach is insufficient for high-level interaction due to the patient's limited eye movement capabilities. To overcome these limitations, we propose a novel design of the two-stage auditory BCI system which exposes a subject to a stream of letter utterances, such that subject is directly stimulated with letter pronunciation, decreasing the mental effort required by the subject. This is contrary to frequently used auditory spellers, which use various non-related sounds representing a specific letter (i.e. various instruments, natural sounds, etc.). Letter discrimination between targets and non-targets is performed with convolutional neural network, with convolutional spatial and temporal filters capable of efficient extraction of event related brain activity features. Second stage of the target letter discrimination is proposed in order to increase the accuracy and usability in the real-world situations, including the reduced stimuli set based on results from the first classification stage. The proposed BCI has been tested on ten healthy subjects, achieving on average spelling accuracy of 30% (max. 100%) and high information transfer rate of 2.38 bits/min (max. 8.14 bits/min), outperforming the state-of-the-art auditory BCI spellers.

A Hybrid Brain-Computer Interface Based on Visual Evoked Potential and Pupillary Response.

Brain-computer interface (BCI) based on steady-state visual evoked potential (SSVEP) has been widely studied due to the high information transfer rate (ITR), little user training, and wide subject applicability. However, there are also disadvantages such as visual discomfort and “BCI illiteracy.” To address these problems, this study proposes to use low-frequency stimulations (12 classes, 0.8–2.12 Hz with an interval of 0.12 Hz), which can simultaneously elicit visual evoked potential (VEP) and pupillary response (PR) to construct a hybrid BCI (h-BCI) system. Classification accuracy was calculated using supervised and unsupervised methods, respectively, and the hybrid accuracy was obtained using a decision fusion method to combine the information of VEP and PR. Online experimental results from 10 subjects showed that the averaged accuracy was 94.90 ± 2.34% (data length 1.5 s) for the supervised method and 91.88 ± 3.68% (data length 4 s) for the unsupervised method, which correspond to the ITR of 64.35 ± 3.07 bits/min (bpm) and 33.19 ± 2.38 bpm, respectively. Notably, the hybrid method achieved higher accuracy and ITR than that of VEP and PR for most subjects, especially for the short data length. Together with the subjects’ feedback on user experience, these results indicate that the proposed h-BCI with the low-frequency stimulation paradigm is more comfortable and favorable than the traditional SSVEP-BCI paradigm using the alpha frequency range.

A user-friendly SSVEP-based BCI using imperceptible phase-coded flickers at 60Hz

A brain-computer interface (BCI) system based on steady-state visual evoked potentials (SSVEP) was developed by four-class phase-coded stimuli. SSVEPs elicited by flickers at 60Hz, which is higher than the critical fusion frequency (CFF), were compared with those at 15Hz and 30Hz. SSVEP components in electroencephalogram (EEG) were detected using task related component analysis (TRCA) method. Offline analysis with 17 subjects indicated that the highest information transfer rate (ITR) was 29.80±4.65bpm with 0.5s data length for 60Hz and the classification accuracy was 70.07±4.15%. The online BCI system reached an averaged classification accuracy of 87.75±3.50% at 60Hz with 4s, resulting in an ITR of 16.73±1.63bpm. In particular, the maximum ITR for a subject was 80bpm with 0.5s at 60Hz. Although the BCI performance of 60Hz was lower than that of 15Hz and 30Hz, the results of the behavioral test indicated that, with no perception of flicker, the BCI system with 60Hz was more comfortable to use than 15Hz and 30Hz. Correlation analysis revealed that SSVEP with higher signal-to-noise ratio (SNR) corresponded to better classification performance and the improvement in comfortableness was accompanied by a decrease in performance. This study demonstrates the feasibility and potential of a user-friendly SSVEP-based BCI using imperceptible flickers.

Task-Related Component Analysis Combining Paired Character Decoding for Miniature Asymmetric Visual Evoked Potentials.

Brain-computer interface (BCI) technology based on event-related potentials (ERP) of electroencephalography (EEG) is widely used in daily life and medical treatment. However, the research of identifying the miniature and more informative asymmetric visual evoked potentials (aVEPs), which belongs to ERP, needs further exploration. Herein, a task-related component analysis combining paired character decoding (TRCA-PCD) method, which can enhance reproducibility of aVEPs in multiple trials and strengthen the features of different samples, was designed to realize fast decoding of aVEPs. The BCI performance and the influence of repetition times between the TRCA-PCD method, the discriminative canonical pattern matching (DCPM) method and traditional task-related component analysis (TRCA) method were compared using a 32-class aVEPs dataset recorded from 32 subjects. The highest average recognition accuracy and information transfer rate (ITR) of TRCA-PCD after parameter selection were 70.37 ± 2.49 % (DCPM: 64.91 ± 2.81 %, TRCA: 44.01 ± 3.25 %) with the peak value of 97.92% and 28.90 ± 3.83 bits/min (DCPM: 21.29 ± 3.35 bits/min, TRCA: 11.54 ± 2.81 bits/min) with the peak value of 94.55 bits/min respectively. Statistical analysis indicated that the highest average recognition rate could be obtained when the repetition time was six, and the highest ITR could be obtained when the repetition time was one. Overall, the results verified the effectiveness and superiority of TRCA-PCD in recognition of aVEPs and provided a reference for parameter selection. Therefore, the TRCA-PCD method can promote the further application of aVEPs in the BCI speller field.