Tohru Yagi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tohru Yagi has a total of 101 co-authors. Here are the top co-authors by number of shared publications:
E. Sawaguchi
10 shared publicationsK. Gesi
10 shared publicationsK. Hasebe
10 shared publications- Kazuhiko Deguchi
Nagoya University
10 shared publications M. Adachi
10 shared publicationsM. Komukae
10 shared publications
Latest Publications
- Open Access
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 34 (15344320) , 3041-3052
Auditory Perception as a Surrogate for Auditory Imagery in EEG-Based BCI Training: Neural Evidence and a Feasibility Study
Brain-computer interface (BCI) systems are commonly classified as reactive or active. Reactive BCIs rely on responses to external stimuli, simplifying decoding but limiting direct user control. In contrast, active BCIs enable intuitive control through spontaneous mental activity but impose greater signal-processing challenges. Speech BCIs, a subset of active systems, hold promise for individuals with severe neurological impairments such as locked-in syndrome (LIS), potentially enabling communication through imagined speech. However, their adoption is hindered by the demanding pre-training required to collect imagery data, which often induces fatigue, and the difficulty many users face in generating consistent, high-quality imagery. To address these limitations, we propose a hybrid BCI framework that uses passive listening tasks as a surrogate for auditory imagery tasks during model training. As a first step, this proof-of-concept feasibility study examined neural responses to five Japanese vowels (/a/, /i/, /u/, /e/, /o/) under listening and imagery conditions in a limited experimental setting with healthy male participants, using event-related potentials (ERPs), topographical mapping, and event-related spectral perturbations (ERSPs), followed by statistical analysis of shared and distinct neural patterns. In the primary three-class and binary classification analyses, classification accuracies for imagery signals obtained from models trained solely on listening data were comparable to, and in some settings higher than, those trained directly on imagery data, with both outperforming chance level. These findings indicate that passive listening may provide effective training data for active auditory imagery BCIs within the tested setting, potentially reducing cognitive demands without compromising performance. Further validation in larger and more diverse cohorts is required before clinical application.
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 34 (15344320) , 3041-3052
Open AccessAuditory Perception as a Surrogate for Auditory Imagery in EEG-Based BCI Training: Neural Evidence and a Feasibility Study
Brain-computer interface (BCI) systems are commonly classified as reactive or active. Reactive BCIs rely on responses to external stimuli, simplifying decoding but limiting direct user control. In contrast, active BCIs enable intuitive control through spontaneous mental activity but impose greater signal-processing challenges. Speech BCIs, a subset of active systems, hold promise for individuals with severe neurological impairments such as locked-in syndrome (LIS), potentially enabling communication through imagined speech. However, their adoption is hindered by the demanding pre-training required to collect imagery data, which often induces fatigue, and the difficulty many users face in generating consistent, high-quality imagery. To address these limitations, we propose a hybrid BCI framework that uses passive listening tasks as a surrogate for auditory imagery tasks during model training. As a first step, this proof-of-concept feasibility study examined neural responses to five Japanese vowels (/a/, /i/, /u/, /e/, /o/) under listening and imagery conditions in a limited experimental setting with healthy male participants, using event-related potentials (ERPs), topographical mapping, and event-related spectral perturbations (ERSPs), followed by statistical analysis of shared and distinct neural patterns. In the primary three-class and binary classification analyses, classification accuracies for imagery signals obtained from models trained solely on listening data were comparable to, and in some settings higher than, those trained directly on imagery data, with both outperforming chance level. These findings indicate that passive listening may provide effective training data for active auditory imagery BCIs within the tested setting, potentially reducing cognitive demands without compromising performance. Further validation in larger and more diverse cohorts is required before clinical application.
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (20) , 42573-42589
EEGMeNet: End-to-End Multitask Neural Network for Brain-Based Mental Workload Classification
High mental workload poses significant challenges across various domains, often impairing performance and decision-making. Recent advancements in Internet of Things (IoT) technologies have enabled continuous monitoring of brain activity via electroencephalography (EEG), facilitating real-time assessment of mental workload during task execution. Early detection of elevated workload levels is essential for mitigating potential adverse outcomes. This paper presents EEGMeNet, an end-to-end, multi-task neural network designed to classify mental workload levels from EEG signals. EEGMeNet consists of four key components: (1) a local feature learner that extracts spectral and spatial characteristics, (2) a global feature learner employing attention-based mechanisms to model temporal dependencies, (3) a feature preserver that enhances representation learning, and (4) a supervised classifier for mental workload classification. The proposed model is evaluated across five mental task EEG datasets—including one acquired in clinical settings—under both within-subject and cross-subject, as well as within-session and cross-session classification paradigms. EEGMeNet outperforms nine state-of-the-art baselines, achieving an accuracy of 87.49±2.66% and an F1 score of 85.26±2.63% in cross-subject evaluations, showing at least a 9% improvement in accuracy and a 12% gain in F1 score. Moreover, EEGMeNet demonstrates low prediction time complexity, making it well-suited for deployment on resource-constrained IoT devices. These results underscore EEGMeNet’s robustness to inter-subject and multi-session EEG variability, validating its effectiveness for real-world EEG-IoT applications. This work establishes EEGMeNet as a promising foundation for future mental workload classification research in EEG-IoT environments.
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (20) , 42573-42589
EEGMeNet: End-to-End Multitask Neural Network for Brain-Based Mental Workload Classification
High mental workload poses significant challenges across various domains, often impairing performance and decision-making. Recent advancements in Internet of Things (IoT) technologies have enabled continuous monitoring of brain activity via electroencephalography (EEG), facilitating real-time assessment of mental workload during task execution. Early detection of elevated workload levels is essential for mitigating potential adverse outcomes. This paper presents EEGMeNet, an end-to-end, multi-task neural network designed to classify mental workload levels from EEG signals. EEGMeNet consists of four key components: (1) a local feature learner that extracts spectral and spatial characteristics, (2) a global feature learner employing attention-based mechanisms to model temporal dependencies, (3) a feature preserver that enhances representation learning, and (4) a supervised classifier for mental workload classification. The proposed model is evaluated across five mental task EEG datasets—including one acquired in clinical settings—under both within-subject and cross-subject, as well as within-session and cross-session classification paradigms. EEGMeNet outperforms nine state-of-the-art baselines, achieving an accuracy of 87.49±2.66% and an F1 score of 85.26±2.63% in cross-subject evaluations, showing at least a 9% improvement in accuracy and a 12% gain in F1 score. Moreover, EEGMeNet demonstrates low prediction time complexity, making it well-suited for deployment on resource-constrained IoT devices. These results underscore EEGMeNet’s robustness to inter-subject and multi-session EEG variability, validating its effectiveness for real-world EEG-IoT applications. This work establishes EEGMeNet as a promising foundation for future mental workload classification research in EEG-IoT environments.
- 15 April 2025
- Published by Royal Society of Chemistry (RSC) in Nanoscale
- Vol. 17 (18) , 11668-11678
Unraveling the time course of interaction between DNA nanopores and lipid bilayers using QCM-D: role of cholesterol anchors and bilayer supporting substrates
We captured the time course of interactions between DNA nanopores and lipid bilayers using dissipation vs. frequency plots obtained from quartz crystal microbalance with energy dissipation monitoring (QCM-D).
- 15 April 2025
- Published by Royal Society of Chemistry (RSC) in Nanoscale
- Vol. 17 (18) , 11668-11678
Unraveling the time course of interaction between DNA nanopores and lipid bilayers using QCM-D: role of cholesterol anchors and bilayer supporting substrates
We captured the time course of interactions between DNA nanopores and lipid bilayers using dissipation vs. frequency plots obtained from quartz crystal microbalance with energy dissipation monitoring (QCM-D).
- 26 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Letters
- Vol. 9 (2) , 1-4
EEG-BBNet: A Hybrid Framework for Brain Biometric Using Graph Connectivity
Most EEG-based biometrics rely on either convolutional neural networks (CNNs) or graph convolutional neural networks (GCNNs) for personal authentication, potentially overlooking the limitations of each approach. To address this, we propose EEG-BBNet, a hybrid network that combines CNNs and GCNNs. EEG-BBNet leverages CNN's capability for automatic feature extraction and the GCNN's ability to learn connectivity patterns between EEG electrodes through graph representation. We evaluate its performance against solely CNN-based and graph-based models across three brain–computer interface tasks, focusing on daily motor and sensory activities. The results show that while EEG-BBNet with Rho index functional connectivity metric outperforms graph-based models, it initially lags behind CNN-based models. However, with additional fine-tuning, EEG-BBNet surpasses CNN-based models, achieving a correct recognition rate of approximately 90%. This improvement enables EEG-BBNet to adapt its learning in new sessions and to acquire different domain knowledge across various BCI tasks (e.g., motor imagery to steady-state visually evoked potentials), demonstrating promise for practical authentication.
- 26 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Letters
- Vol. 9 (2) , 1-4
EEG-BBNet: A Hybrid Framework for Brain Biometric Using Graph Connectivity
Most EEG-based biometrics rely on either convolutional neural networks (CNNs) or graph convolutional neural networks (GCNNs) for personal authentication, potentially overlooking the limitations of each approach. To address this, we propose EEG-BBNet, a hybrid network that combines CNNs and GCNNs. EEG-BBNet leverages CNN's capability for automatic feature extraction and the GCNN's ability to learn connectivity patterns between EEG electrodes through graph representation. We evaluate its performance against solely CNN-based and graph-based models across three brain–computer interface tasks, focusing on daily motor and sensory activities. The results show that while EEG-BBNet with Rho index functional connectivity metric outperforms graph-based models, it initially lags behind CNN-based models. However, with additional fine-tuning, EEG-BBNet surpasses CNN-based models, achieving a correct recognition rate of approximately 90%. This improvement enables EEG-BBNet to adapt its learning in new sessions and to acquire different domain knowledge across various BCI tasks (e.g., motor imagery to steady-state visually evoked potentials), demonstrating promise for practical authentication.
- 4 June 2024
- Published by Royal Society of Chemistry (RSC) in Nanoscale
- Vol. 16 (25) , 12174-12183
Synthetic DNA nanopores for direct molecular transmission between lipid vesicles
We designed a pair of DNA nanopores that can connect and form a direct molecular pathway between lipid vesicles.
- 4 June 2024
- Published by Royal Society of Chemistry (RSC) in Nanoscale
- Vol. 16 (25) , 12174-12183
Synthetic DNA nanopores for direct molecular transmission between lipid vesicles
We designed a pair of DNA nanopores that can connect and form a direct molecular pathway between lipid vesicles.
- 5 November 2023
- Published by Springer Nature
- p. 931-940
Behavior Control of Drone Triggered by Object Detection
- 5 November 2023
- Published by Springer Nature
- p. 931-940
Behavior Control of Drone Triggered by Object Detection
- Open Access
- 11 October 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-12
Depressive states in healthy subjects lead to biased processing in frontal-parietal ERPs during emotional stimuli
Subthreshold depressive (sD) states and major depression are considered to occur on a continuum, and there are only quantitative and not qualitative differences between depressive states in healthy individuals and patients with depression. sD is showing a progressively increasing prevalence and has a lifelong impact, and the social and clinical impacts of sD are no less than those of major depressive disorder (MDD). Because depression leads to biased cognition, patients with depression and healthy individuals show different visual processing properties. However, it remains unclear whether there are significant differences in visual information recognition among healthy individuals with various depressive states. In this study, we investigated the event-related potentials (ERPs) and event-related spectrum perturbation (ERSP) of healthy individuals with various depressive states during the perception of emotional visual stimulation. We show that different neural activities can be detected even among healthy individuals. We divided healthy participants into high, middle, and low depressive state groups and found that participants in a high depressive state had a lower P300 amplitude and significant differences in fast and slow neural responses in the frontal and parietal lobes. We anticipate our study to provide useful parameters for assessing the evaluation of depressive states in healthy individuals.
- 11 October 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-12
Open AccessDepressive states in healthy subjects lead to biased processing in frontal-parietal ERPs during emotional stimuli
Subthreshold depressive (sD) states and major depression are considered to occur on a continuum, and there are only quantitative and not qualitative differences between depressive states in healthy individuals and patients with depression. sD is showing a progressively increasing prevalence and has a lifelong impact, and the social and clinical impacts of sD are no less than those of major depressive disorder (MDD). Because depression leads to biased cognition, patients with depression and healthy individuals show different visual processing properties. However, it remains unclear whether there are significant differences in visual information recognition among healthy individuals with various depressive states. In this study, we investigated the event-related potentials (ERPs) and event-related spectrum perturbation (ERSP) of healthy individuals with various depressive states during the perception of emotional visual stimulation. We show that different neural activities can be detected even among healthy individuals. We divided healthy participants into high, middle, and low depressive state groups and found that participants in a high depressive state had a lower P300 amplitude and significant differences in fast and slow neural responses in the frontal and parietal lobes. We anticipate our study to provide useful parameters for assessing the evaluation of depressive states in healthy individuals.
- 14 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 10, 1-5
Effects of Trial-Adjusted Neurofeedback Training on Motor-Imagery Based Brain-Computer Interface Performance
Motor imagery (MI) classification based on electroencephalography (EEG) has been extensively studied and recently used more in brain-computer interfaces (BCI). This study uses left and right hand MI tasks for the BCI system. A common obstacle for MI-BCI is the inability of some participants to perform the BCI task, called BCI illiteracy. Various training protocols have been investigated to improve the performance of BCI but are designed with a balanced dataset. Similarly to how people show a bias towards a side (e.g. left or right) for motor execution tasks, it has been seen that participants also show a performance bias in MI tasks as well. To address this MI bias in participants, a novel neurofeedback protocol was designed to adjust the number of trials each condition has. Trials will be adjusted to increase the number of times participants have to perform their weak MI task. This study aims to investigate the overall effect that the trial-adjusted neurofeedback had on participant’s cognitive performance on the MI-BCI system. The effects were investigated through time-frequency and band power analysis. The time-frequency analysis showed improvement in key MI feature and band power analysis results had an improvement on the alpha and beta frequency bands. In the analysis results, trial-adjusted neurofeedback was seen to have an effect on participant’s cognitive performance on the MI-BCI task.
- 14 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 10, 1-5
Effects of Trial-Adjusted Neurofeedback Training on Motor-Imagery Based Brain-Computer Interface Performance
Motor imagery (MI) classification based on electroencephalography (EEG) has been extensively studied and recently used more in brain-computer interfaces (BCI). This study uses left and right hand MI tasks for the BCI system. A common obstacle for MI-BCI is the inability of some participants to perform the BCI task, called BCI illiteracy. Various training protocols have been investigated to improve the performance of BCI but are designed with a balanced dataset. Similarly to how people show a bias towards a side (e.g. left or right) for motor execution tasks, it has been seen that participants also show a performance bias in MI tasks as well. To address this MI bias in participants, a novel neurofeedback protocol was designed to adjust the number of trials each condition has. Trials will be adjusted to increase the number of times participants have to perform their weak MI task. This study aims to investigate the overall effect that the trial-adjusted neurofeedback had on participant’s cognitive performance on the MI-BCI system. The effects were investigated through time-frequency and band power analysis. The time-frequency analysis showed improvement in key MI feature and band power analysis results had an improvement on the alpha and beta frequency bands. In the analysis results, trial-adjusted neurofeedback was seen to have an effect on participant’s cognitive performance on the MI-BCI task.
- Open Access
Functional Evaluation of 3D Liver Models Labeled with Polysaccharide Functionalized Magnetic Nanoparticles
Establishing a rapid in vitro evaluation system for drug screening is essential for the development of new drugs. To reproduce tissues/organs with functions closer to living organisms, in vitro three-dimensional (3D) culture evaluation using microfabrication technology has been reported in recent years. Culture on patterned substrates with controlled hydrophilic and hydrophobic regions (Cell-ableTM) can create 3D liver models (miniature livers) with liver-specific Disse luminal structures and functions. MRI contrast agents are widely used as safe and minimally invasive diagnostic methods. We focused on anionic polysaccharide magnetic iron oxide nanoparticles (Resovist®) and synthesized the four types of nanoparticle derivatives with different properties. Cationic nanoparticles (TMADM) can be used to label target cells in a short time and have been successfully visualized in vivo. In this study, we examined the morphology of various nanoparticles. The morphology of various nanoparticles showed relatively smooth-edged spherical shapes. As 3D liver models, we prepared primary hepatocyte–endothelial cell heterospheroids. The toxicity, CYP3A, and albumin secretory capacity were evaluated in the heterospheroids labeled with various nanoparticles. As the culture period progressed, the heterospheroids labeled with anionic and cationic nanoparticles showed lower liver function than non-labeled heterospheroids. In the future, there is a need to improve the method of creation of artificial 3D liver or to design a low-invasive MRI contrast agent to label the artificial 3D liver.
Open AccessFunctional Evaluation of 3D Liver Models Labeled with Polysaccharide Functionalized Magnetic Nanoparticles
Establishing a rapid in vitro evaluation system for drug screening is essential for the development of new drugs. To reproduce tissues/organs with functions closer to living organisms, in vitro three-dimensional (3D) culture evaluation using microfabrication technology has been reported in recent years. Culture on patterned substrates with controlled hydrophilic and hydrophobic regions (Cell-ableTM) can create 3D liver models (miniature livers) with liver-specific Disse luminal structures and functions. MRI contrast agents are widely used as safe and minimally invasive diagnostic methods. We focused on anionic polysaccharide magnetic iron oxide nanoparticles (Resovist®) and synthesized the four types of nanoparticle derivatives with different properties. Cationic nanoparticles (TMADM) can be used to label target cells in a short time and have been successfully visualized in vivo. In this study, we examined the morphology of various nanoparticles. The morphology of various nanoparticles showed relatively smooth-edged spherical shapes. As 3D liver models, we prepared primary hepatocyte–endothelial cell heterospheroids. The toxicity, CYP3A, and albumin secretory capacity were evaluated in the heterospheroids labeled with various nanoparticles. As the culture period progressed, the heterospheroids labeled with anionic and cationic nanoparticles showed lower liver function than non-labeled heterospheroids. In the future, there is a need to improve the method of creation of artificial 3D liver or to design a low-invasive MRI contrast agent to label the artificial 3D liver.
- Open Access
- 20 April 2022
- Published by Fuji Technology Press Ltd. in Journal of Robotics and Mechatronics
- Vol. 34 (2) , 316-318
Neural Interface for Biohybrid Prosthetic Hands to Realize Sensory and Motor Functions
A neural interface is a technology that facilitates communication between the brain and external devices. One potential clinical application of a neural interface is in prosthetic arms. These devices can realize motor and sensory functions, enabling amputee patients to perform daily tasks. However, such prosthetic arms are still challenging because of the poor resolution of conventional neural electrodes and the difficulty in sorting the motor and sensory nerve cells. In this study, we attempt to utilize deoxyribonucleic acid (DNA) nanotubes to develop a novel intracellular electrode with high resolution. Our results indicate that DNA nanotubes can transport ions, such as Ca2+. Moreover, a microchannel device for sorting motor and sensory nerves is introduced.
- 20 April 2022
- Published by Fuji Technology Press Ltd. in Journal of Robotics and Mechatronics
- Vol. 34 (2) , 316-318
Open AccessNeural Interface for Biohybrid Prosthetic Hands to Realize Sensory and Motor Functions
A neural interface is a technology that facilitates communication between the brain and external devices. One potential clinical application of a neural interface is in prosthetic arms. These devices can realize motor and sensory functions, enabling amputee patients to perform daily tasks. However, such prosthetic arms are still challenging because of the poor resolution of conventional neural electrodes and the difficulty in sorting the motor and sensory nerve cells. In this study, we attempt to utilize deoxyribonucleic acid (DNA) nanotubes to develop a novel intracellular electrode with high resolution. Our results indicate that DNA nanotubes can transport ions, such as Ca2+. Moreover, a microchannel device for sorting motor and sensory nerves is introduced.