Byoung-Hee Kwon

Byeong-Hoo Lee, Jeong-Hyun Cho, Byoung-Hee Kwon et al.

Recently, advanced technologies have unlimited potential in solving various problems with a large amount of data. However, these technologies have yet to show competitive performance in brain-computer interfaces (BCIs) which deal with brain signals. Basically, brain signals are difficult to collect in large quantities, in particular, the amount of information would be sparse in spontaneous BCIs. In addition, we conjecture that high spatial and temporal similarities between tasks increase the prediction difficulty. We define this problem as sparse condition. To solve this, a factorization approach is introduced to allow the model to obtain distinct representations from latent space. To this end, we propose two feature extractors: A class-common module is trained through adversarial learning acting as a generator; Class-specific module utilizes loss function generated from classification so that features are extracted with traditional methods. To minimize the latent space shared by the class-common and class-specific features, the model is trained under orthogonal constraint. As a result, EEG signals are factorized into two separate latent spaces. Evaluations were conducted on a single-arm motor imagery dataset. From the results, we demonstrated that factorizing the EEG signal allows the model to extract rich and decisive features under sparse condition.

Kang Yin, Byeong-Hoo Lee, Byoung-Hee Kwon et al.

Data augmentation approaches are widely explored for the enhancement of decoding electroencephalogram signals. In subject-independent brain-computer interface system, domain adaption and generalization are utilized to shift source subjects' data distribution to match the target subject as an augmentation. However, previous works either introduce noises (e.g., by noise addition or generation with random noises) or modify target data, thus, cannot well depict the target data distribution and hinder further analysis. In this paper, we propose a target-centered subject transfer framework as a data augmentation approach. A subset of source data is first constructed to maximize the source-target relevance. Then, the generative model is applied to transfer the data to target domain. The proposed framework enriches the explainability of target domain by adding extra real data, instead of noises. It shows superior performance compared with other data augmentation methods. Extensive experiments are conducted to verify the effectiveness and robustness of our approach as a prosperous tool for further research.

Byeong-Hoo Lee, Byoung-Hee Kwon, Seong-Whan Lee

Deep learning has shown promise in decoding brain signals, such as electroencephalogram (EEG), in the field of brain-computer interfaces (BCIs). However, the non-stationary characteristics of EEG signals pose challenges for training neural networks to acquire appropriate knowledge. Inconsistent EEG signals resulting from these non-stationary characteristics can lead to poor performance. Therefore, it is crucial to investigate and address sample inconsistency to ensure robust performance in spontaneous BCIs. In this study, we introduce the concept of sample dominance as a measure of EEG signal inconsistency and propose a method to modulate its effect on network training. We present a two-stage dominance score estimation technique that compensates for performance degradation caused by sample inconsistencies. Our proposed method utilizes non-parametric estimation to infer sample inconsistency and assigns each sample a dominance score. This score is then aggregated with the loss function during training to modulate the impact of sample inconsistency. Furthermore, we design a curriculum learning approach that gradually increases the influence of inconsistent signals during training to improve overall performance. We evaluate our proposed method using public spontaneous BCI dataset. The experimental results confirm that our findings highlight the importance of addressing sample dominance for achieving robust performance in spontaneous BCIs.

Si-Hyun Kim, Heon-Gyu Kwak, Byoung-Hee Kwon et al.

Brain-computer interface (BCI) aims to decode motor intent from noninvasive neural signals to enable control of external devices, but practical deployment remains limited by noise and variability in motor imagery (MI)-based electroencephalogram (EEG) signals. This work investigates a hierarchical and meta-cognitive decoding framework for four-class MI classification. We introduce a multi-scale hierarchical signal processing module that reorganizes backbone features into temporal multi-scale representations, together with an introspective uncertainty estimation module that assigns per-cycle reliability scores and guides iterative refinement. We instantiate this framework on three standard EEG backbones (EEGNet, ShallowConvNet, and DeepConvNet) and evaluate four-class MI decoding using the BCI Competition IV-2a dataset under a subject-independent setting. Across all backbones, the proposed components improve average classification accuracy and reduce inter-subject variance compared to the corresponding baselines, indicating increased robustness to subject heterogeneity and noisy trials. These results suggest that combining hierarchical multi-scale processing with introspective confidence estimation can enhance the reliability of MI-based BCI systems.

Jeong-Hyun Cho, Byoung-Hee Kwon, Byeong-Hoo Lee et al.

Brain-computer interface (BCI) is a practical pathway to interpret users' intentions by decoding motor execution (ME) or motor imagery (MI) from electroencephalogram (EEG) signals. However, developing a BCI system driven by ME or MI is challenging, particularly in the case of containing continual and compound muscles movements. This study analyzes three grasping actions from EEG under both ME and MI paradigms. We also investigate the classification performance in offline and pseudo-online experiments. We propose a novel approach that uses muscle activity pattern (MAP) images for the convolutional neural network (CNN) to improve classification accuracy. We record the EEG and electromyogram (EMG) signals simultaneously and create the MAP images by decoding both signals to estimate specific hand grasping. As a result, we obtained an average classification accuracy of 63.6($\pm$6.7)% in ME and 45.8($\pm$4.4)% in MI across all fifteen subjects for four classes. Also, we performed pseudo-online experiments and obtained classification accuracies of 60.5($\pm$8.4)% in ME and 42.7($\pm$6.8)% in MI. The proposed method MAP-CNN, shows stable classification performance, even in the pseudo-online experiment. We expect that MAP-CNN could be used in various BCI applications in the future.

Byoung-Hee Kwon, Jeong-Hyun Cho, Byeong-Hoo Lee

An electroencephalogram is an effective approach that provides a bidirectional pathway between user and computer in a non-invasive way. In this study, we adopted the visual perception data for training the visual imagery decoding network. We proposed a visual perception-guided network training approach for decoding visual imagery. Visual perception decreases the power of the alpha frequency range of the visual cortex over time when the user performed the task, and visual imagery increases the power of the alpha frequency range of the visual cortex over time as the user performed with the task. Generated brain signals when the user performing visual imagery and visual perception have opposite brain activity tendencies, and we used these characteristics to design the proposed network. When using the proposed method, the average classification performance of visual imagery with the visual perception data was 0.7008. Our results provide the possibility of using the visual perception data as a guide of the visual imagery classification network training.

Byoung-Hee Kwon, Ji-Hoon Jeong, Seong-Whan Lee

Brain-computer interfaces (BCIs) use brain signals such as electroencephalography to reflect user intention and enable two-way communication between computers and users. BCI technology has recently received much attention in healthcare applications, such as neurorehabilitation and diagnosis. BCI applications can also control external devices using only brain activity, which can help people with physical or mental disabilities, especially those suffering from neurological and neuromuscular diseases such as stroke and amyotrophic lateral sclerosis. Motor imagery (MI) has been widely used for BCI-based device control, but we adopted intuitive visual motion imagery to overcome the weakness of MI. In this study, we developed a three-dimensional (3D) BCI training platform to induce users to imagine upper-limb movements used in real-life activities (picking up a cell phone, pouring water, opening a door, and eating food). We collected intuitive visual motion imagery data and proposed a deep learning network based on functional connectivity as a mind-reading technique. As a result, the proposed network recorded a high classification performance on average (71.05%). Furthermore, we applied the leave-one-subject-out approach to confirm the possibility of improvements in subject-independent classification performance. This study will contribute to the development of BCI-based healthcare applications for rehabilitation, such as robotic arms and wheelchairs, or assist daily life.

Byoung-Hee Kwon, Byeong-Hoo Lee, Ji-Hoon Jeong

The electroencephalogram, a type of non-invasive-based brain signal that has a user intention-related feature provides an efficient bidirectional pathway between user and computer. In this work, we proposed a deep learning framework based on corresponding frequency empahsize method to decode the motor imagery (MI) data from 2020 International BCI competition dataset. The MI dataset consists of 3-class, namely 'Cylindrical', 'Spherical', and 'Lumbrical'. We utilized power spectral density as an emphasize method and a convolutional neural network to classify the modified MI data. The results showed that MI-related frequency range was activated during MI task, and provide neurophysiological evidence to design the proposed method. When using the proposed method, the average classification performance in intra-session condition was 69.68% and the average classification performance in inter-session condition was 52.76%. Our results provided the possibility of developing a BCI-based device control system for practical applications.

Byoung-Hee Kwon, Ji-Hoon Jeong, Jeong-Hyun Cho et al.

In this study, we adopted visual motion imagery, which is a more intuitive brain-computer interface (BCI) paradigm, for decoding the intuitive user intention. We developed a 3-dimensional BCI training platform and applied it to assist the user in performing more intuitive imagination in the visual motion imagery experiment. The experimental tasks were selected based on the movements that we commonly used in daily life, such as picking up a phone, opening a door, eating food, and pouring water. Nine subjects participated in our experiment. We presented statistical evidence that visual motion imagery has a high correlation from the prefrontal and occipital lobes. In addition, we selected the most appropriate electroencephalography channels using a functional connectivity approach for visual motion imagery decoding and proposed a convolutional neural network architecture for classification. As a result, the averaged classification performance of the proposed architecture for 4 classes from 16 channels was 67.50 % across all subjects. This result is encouraging, and it shows the possibility of developing a BCI-based device control system for practical applications such as neuroprosthesis and a robotic arm.

Byoung-Hee Kwon, Ji-Hoon Jeong, Dong-Joo Kim

In this study, 3D brain-computer interface (BCI) training platforms were used to stimulate the subjects for visual motion imagery and visual perception. We measured the activation brain region and alpha-band power activity when the subjects perceived and imagined the stimuli. Based on this, 4-class were classified in visual stimuli session and visual motion imagery session respectively. The results showed that the occipital region is involved in visual perception and visual motion imagery, and alpha-band power is increased in visual motion imagery session and decreased in visual motion stimuli session. Compared with the performance of visual motion imagery and motor imagery, visual motion imagery has higher performance than motor imagery. The binary class was classified using one versus rest approach as well as analysis of brain activation to prove that visual-related brain wave signals are meaningful, and the results were significant.