A5008751186- Functional Brain Connectivity Studies
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- Advanced Neuroimaging Techniques and Applications
- Mental Health Research Topics
- Medical Imaging and Analysis
- Neural dynamics and brain function
- Brain Tumor Detection and Classification
- Advanced X-ray and CT Imaging
- Domain Adaptation and Few-Shot Learning
- Advanced Graph Neural Networks
- EEG and Brain-Computer Interfaces
- Image and Signal Denoising Methods
- Radiomics and Machine Learning in Medical Imaging
- Advanced Image Fusion Techniques
- Magnetic and transport properties of perovskites and related materials
- Advanced MRI Techniques and Applications
- Electronic and Structural Properties of Oxides
- Advanced Image and Video Retrieval Techniques
- Anomaly Detection Techniques and Applications
- Image Retrieval and Classification Techniques
- Adversarial Robustness in Machine Learning
- Image and Video Quality Assessment
- Neural Networks and Applications
- Forensic Anthropology and Bioarchaeology Studies
Southeast University
2016-2025
Jiangnan University
2025
Inserm
2018-2024
Université de Rennes
2018-2024
Nanjing Medical University
2024
Zhongda Hospital Southeast University
2024
Lumir Research Institute (United States)
2024
Ministry of Education
2021
Ministry of Education of the People's Republic of China
2021
Chinese University of Hong Kong
1998-2015
Can we train the computer to beat experienced traders for financial assert trading? In this paper, try address challenge by introducing a recurrent deep neural network (NN) real-time signal representation and trading. Our model is inspired two biological-related learning concepts of (DL) reinforcement (RL). framework, DL part automatically senses dynamic market condition informative feature learning. Then, RL module interacts with representations makes trading decisions accumulate ultimate...
Deep learning (DL) is an emerging and powerful paradigm that allows large-scale task-driven feature from big data. However, typical DL a fully deterministic model sheds no light on data uncertainty reductions. In this paper, we show how to introduce the concepts of fuzzy into overcome shortcomings fixed representation. The bulk proposed system hierarchical deep neural network derives information both representations. Then, knowledge learnt these two respective views are fused altogether...
Renal cancer is one of ten most common cancers in human beings. The laparoscopic partial nephrectomy (LPN) becomes the main therapeutic approach treating renal cancer. Accurate kidney and tumor segmentation CT images a prerequisite step surgery planning. However, automatic accurate remains challenge. In this paper, we propose new method to perform precise angiography images. This relies on three-dimensional (3D) fully convolutional network (FCN) which combines pyramid pooling module (PPM)....
Deep learning models such as convolutional neural network has been widely used in 3D biomedical image segmentation. However, most of them neither consider the correlations between different modalities, nor fully exploit depth information. To better leverage multi-modalities and information, we proposed an architecture for brain tumor segmentation multi- modal magnetic resonance images (MRI), named LSTM UNet. Experiments results on BRATS-2015 show that our method outperforms state-of-the-art...
The pathophysiology of major depressive disorder (MDD) has been explored to be highly associated with the dysfunctional integration brain networks. It is therefore imperative explore neuroimaging biomarkers aid diagnosis and treatment. In this study, we developed a spatiotemporal graph convolutional network (STGCN) framework learn discriminative features from functional connectivity for automatic treatment response prediction MDD. Briefly, dynamic networks were first obtained resting-state...
The loss of locus coeruleus (LC)-norepinephrine system may contribute to freezing gait (FOG) in Parkinson's disease (PD), but free-water (FW) imaging has not been applied investigate LC microstructural degeneration FOG. This study was the role LC-norepinephrine FOG pathophysiology using FW and resting-state functional magnetic resonance imaging. metrics were analyzed 52 healthy controls, 79 PD patients without (Non-FOG), 110 with (48 "Off-period" 62 "Levodopa unresponsive" FOG). Correlation...
To investigate the differences in gait parameters and cortical activity during a single-task walking (STW) cognitive dual-task (DTW) between multiple system atrophy with predominant parkinsonism (MSA-P) Parkinson's disease (PD). 24 MSA-P patients, 20 PD 22 healthy controls (HCs) were enrolled. Gait collected using portable inertial measurement unit system, relative change of oxyhemoglobin (ΔHbO2) bilateral frontal sensorimotor cortex was obtained by functional near-infrared spectroscopy...
The financial industry has witnessed an exceptionally fast progress of incorporating information processing techniques in designing knowledge-based automated systems for high-frequency trading (HFT). This paper proposes a sparse coding-inspired optimal (SCOT) system real-time signal representation and trading. Mathematically, SCOT simultaneously learns the dictionary, features, strategy joint optimization, yielding feature representations specific objective. learning process is modeled as...
Alzheimer's disease (AD) is the most common reason for dementia in elderly population. Its neuropathological features include senile plaques, neurofibril tangles and neuronal death. Scientists have established many AD animal models, including yeast, Caenorhabditis elegans, Drosophila melanogaster, mice, rats non-human primates. models are much more efficient genetic manipulation screening assay than mammals. microRNAs (miRNAs) ~22nt small RNA molecules that fine-tune gene expression at...
Deformable medical image registration estimates corresponding deformation to align the regions of interest (ROIs) two images a same spatial coordinate system. However, recent unsupervised models only have correspondence ability without perception, making misalignment on blurred anatomies and distortion task-unconcerned backgrounds. Label-constrained (LC) embed perception via labels, but lack texture constraints in labels expensive labeling costs causes internal ROIs overfitted perception. We...
Background: Previous studies have demonstrated that cognitive impairment is linked with neurophysiological alterations in chronic tinnitus. This study aimed to investigate the intrinsic functional connectivity (FC) pattern within default mode network (DMN) and its associations tinnitus patients using a resting-state magnetic resonance imaging (rs-fMRI). Methods: Thirty-five unilateral patients, 50 healthy controls were recruited for rsfMRI scanning. Both groups age, gender education level...
This paper presents a robust information theoretic (RIT) model to reduce the uncertainties, i.e., missing and noisy labels, in general discriminative data representation tasks. The fundamental pursuit of our is simultaneously learn transformation function classifier that maximize mutual their labels latent space. In this paradigm, we, respectively, discuss three types RIT implementations with linear subspace embedding, deep transformation, structured sparse learning. practice, are exploited...
The recently proposed principal component analysis network (PCANet) has performed well with respect to the classification of 2-D images. However, feature extraction may perform less when dealing multi-dimensional images, since spatial relationships within structures images are not fully utilized. In this paper, we develop a multilinear (MPCANet), which is tensor extension PCANet, extract high-level semantic features from extracted largely minimize intraclass invariance objects by making...