A5025554387- EEG and Brain-Computer Interfaces
- Neuroscience and Neural Engineering
- Functional Brain Connectivity Studies
- Gaze Tracking and Assistive Technology
- Neural dynamics and brain function
- Advanced Memory and Neural Computing
- Muscle activation and electromyography studies
- Blind Source Separation Techniques
- Smart Grid and Power Systems
- Multimodal Machine Learning Applications
- Advanced DC-DC Converters
- Microgrid Control and Optimization
- Multilevel Inverters and Converters
- Power Systems and Renewable Energy
- Machine Learning in Bioinformatics
- Anomaly Detection Techniques and Applications
- High-Voltage Power Transmission Systems
- Topic Modeling
- Advanced Neuroimaging Techniques and Applications
- ECG Monitoring and Analysis
- Emotion and Mood Recognition
- Advanced Graph Neural Networks
- Power Quality and Harmonics
- Speech and Audio Processing
- Advanced Image and Video Retrieval Techniques
Juntendo University
2023-2025
Juntendo University Urayasu Hospital
2023-2025
RIKEN Center for Advanced Photonics
2024-2025
Nanjing University of Aeronautics and Astronautics
2024
Juntendo University Hospital
2024
Yanshan University
2017-2023
Nankai University
2023
Data Assurance and Communication Security
2023
Zhengzhou University
2023
National Clinical Research
2023
Brain-computer interface provides a new communication bridge between the human mind and devices, depending largely on accurate classification identification of non-invasive EEG signals. Recently, deep learning approaches have been widely used in many fields to extract features classify various types data successfully. However, approach requires massive train its neural networks, amount impacts greatly quality classifiers. This paper proposes novel that combines augmentation for...
Abstract Objective . Alzheimer’s disease is a progressive neurodegenerative dementia that poses significant global health threat. It imperative and essential to detect patients in the mild cognitive impairment (MCI) stage or even earlier, enabling effective interventions prevent further deterioration of dementia. This study focuses on early prediction utilizing Magnetic Resonance Imaging (MRI) data, using proposed Graph Convolutional Networks (GCNs). Approach Specifically, we developed...
Previous studies made progress in the early diagnosis of Alzheimer's disease (AD) using electroencephalography (EEG) without considering EEG connectivity. To fill this gap, we explored significant differences between AD patients and controls based on frequency domain spatial properties functional connectivity mild cognitive impairment (MCI) datasets. Four global metrics, network resilience, connection-level metrics node versatility were used to distinguish patients. The results show that...
Findings from unbiased genetic studies have consistently implicated synaptic protein networks in schizophrenia, but the molecular pathologic features within these and their contribution to circuit deficits thought underlie disease symptoms remain unknown.
Brain-computer interface (BCI) technologies have been widely used in many areas. In particular, non-invasive such as electroencephalography (EEG) or near-infrared spectroscopy (NIRS) to detect motor imagery, disease, mental state. It has already shown literature that the hybrid of EEG and NIRS better results than their respective individual signals. The fusion algorithm for sources is key implement them real-life applications. this research, we propose three methods NIRS-based brain-computer...
Treadmills are widely used to recover walking function in the rehabilitation field for those patients with gait disorders. Nevertheless, ultimate goal of recovery is walk on ground rather than treadmill. This study aims determine effect treadmill and upper trunk movement characteristics using wearable sensors. Eight healthy male subjects recruited perform 420-m straight overground (OW) 5 min (TW), wearing 3 inertial measurement units a pair insole In addition common linear features,...
Great progress has been made in diagnosing medical diseases based on deep learning. Large-scale data are expected to improve learning performance further. It is almost impossible for a single institution collect so much due the time-consuming and costly collection labeling of data. Many studies have turned attention sharing among multiple institutions. However, different acquiring processing procedures, institutions' characterized by distribution heterogeneity. Besides, protection patient...
Abstract As the Chinese Spallation Neutron Source enters Phase II, increase in proton beam power will lead to a further boost intensity of pulsed neutron beams. To address demand for higher event-rate readout electronics energy-resolved imaging detectors, we have developed high-performance system based on Timepix4 chip. The prototype comprises chip board and digital board, which are interconnected through custom FMC interface. advantage this is its ability achieve full bandwidth 160 Gbps...
In this article, we present a collection of fifteen novel contributions on machine learning methods with low-quality or imperfect datasets, which were accepted for publication in the special issue “Machine Learning Methods Noisy, Incomplete Small Datasets”, Applied Sciences (ISSN 2076-3417). These papers provide variety approaches to real-world problems where available datasets suffer from imperfections such as missing values, noise artefacts. Contributions applied sciences include medical...
Abstract Schizophrenia (Sz) is a highly polygenic disorder, with common, rare, and structural variants each contributing only small fraction of overall disease risk. Thus, there need to identify downstream points convergence that can be targeted therapeutics. Reduction microtubule-associated protein 2 (MAP2) immunoreactivity (MAP2-IR) present in individuals Sz, despite no change MAP2 levels. phosphorylated multiple receptors kinases identified as Sz risk genes, altering its function. Using...
Empirical mode decomposition (EMD) has developed into a prominent tool for adaptive, scale-based signal analysis in various fields like robotics, security and biomedical engineering. Since the dramatic increase amount of data puts forward higher requirements capability real-time analysis, it is difficult existing EMD its variants to trade off growth dimension speed analysis. In order decompose multi-dimensional signals at faster speed, we present novel signal-serialization method...
Abstract Purpose The brain–computer interface (BCI) based on motor imagery (MI) has attracted extensive interest due to its spontaneity and convenience. However, the traditional MI paradigm is limited by weak features in evoked EEG signal, which often leads lower classification performance. Methods In this paper, a novel proposed improve BCI performance, speech imaginary combined with silent reading (SR) writing (WI), instead of imagining body movements. multimodal (imaginary voices...
In many machine learning applications, measurements are sometimes incomplete or noisy resulting in missing features. other cases, and for different reasons, the datasets originally small, therefore, more data samples required to derive useful supervised unsupervised classification methods. Correct handling of incomplete, small is a fundamental classic challenge. this article, we provide unified review recently proposed methods based on signal decomposition features imputation (data...
Objective Platelet (PLT) engages in immune and inflammatory responses, all of which are related to the prognosis critically ill patients. Although thrombocytopenia at ICU admission contributes in-hospital mortality, PLT is repeatedly measured during hospitalization role longitudinal trajectory remains unclear. We aimed identify dynamic patterns evaluate their relationships with mortality risk thrombocytopenia. Methods adopted a three-phase, multi-cohort study strategy. Firstly, within first...
The classification of limb movements can provide with control commands in non-invasive brain-computer interface. Previous studies on the have focused left/right limbs; however, different types upper has often been ignored despite that it provides more active-evoked Nevertheless, few machine learning method be used as state-of-the-art multi-class movements. This work focuses and proposes filter bank task-related component analysis (mFBTRCA) method, which consists three steps: spatial...