A5100676870- Functional Brain Connectivity Studies
- Neural dynamics and brain function
- Advanced Neural Network Applications
- Visual perception and processing mechanisms
- Machine Learning in Healthcare
- Domain Adaptation and Few-Shot Learning
- Artificial Intelligence in Healthcare
- Machine Learning and Data Classification
- Advanced MRI Techniques and Applications
- Adversarial Robustness in Machine Learning
- Cancer-related gene regulation
- Speech and Audio Processing
- Speech Recognition and Synthesis
- Neural and Behavioral Psychology Studies
- Mental Health Research Topics
- Remote Sensing and LiDAR Applications
- Linguistics and Cultural Studies
- Sparse and Compressive Sensing Techniques
- RNA modifications and cancer
- Music and Audio Processing
- Autonomous Vehicle Technology and Safety
- Tensor decomposition and applications
- Higher Education and Teaching Methods
- Epigenetics and DNA Methylation
- Nonlinear Differential Equations Analysis
Dalian Medical University
2025
First Affiliated Hospital of Dalian Medical University
2025
China Medical University
2025
Jilin University
2024-2025
The Ohio State University
2012-2024
Nanjing University of Science and Technology
2017-2024
Jilin Medical University
2024
Chinese Academy of Sciences
2024
Guangzhou Regenerative Medicine and Health Guangdong Laboratory
2024
Wayne State University
2016-2023
The development of magnetic resonance imaging (MRI) techniques has defined modern neuroimaging. Since its inception, tens thousands studies using such as functional MRI and diffusion weighted have allowed for the non-invasive study brain. Despite fact that is routinely used to obtain data neuroscience research, there been no widely adopted standard organizing describing collected in an experiment. This renders sharing reusing (within or between labs) difficult if not impossible unnecessarily...
Abstract The Iowa Gambling Task (IGT) is a sensitive test for the detection of decision‐making impairments in several neurological and psychiatric populations. Very few studies have employed IGT functional magnetic resonance imaging (fMRI) investigations, part, because task cognitively complex. Here we report method exploring brain activity using fMRI during performance IGT. Decision‐making was associated with regions group healthy individuals. activated were consistent neural circuitry...
In this paper, we propose a new deep feature selection method based on architecture. Our uses stacked auto-encoders for representation in higher-level abstraction. We developed and applied novel learning approach to specific precision medicine problem, which focuses assessing prioritizing risk factors hypertension (HTN) vulnerable demographic subgroup (African-American). is use identify significant affecting left ventricular mass indexed body surface area (LVMI) as an indicator of heart...
In the rapidly evolving landscape of Internet Things (IoT), concerns about privacy and security have become significant as interconnected devices communicate collaborate. Fingerprints, serving unique biometric identifiers, play a crucial role in authentication identification processes within this exchanged network. However, attention is often directed towards disclosure visible fingerprints, overlooking latent fingerprints. This primarily due to challenges involved extracting especially...
5-Methylcytosine (m 5 C) is one of the post-transcriptional modifications in mRNA and involved pathogenesis various diseases. However, capacity existing assays for accurately comprehensively transcriptome-wide m C mapping still needs improvement. Here, we develop a detection method named DRAM (deaminase reader protein assisted RNA methylation analysis), which deaminases (APOBEC1 TadA-8e) are fused with proteins (ALYREF YBX1) to identify sites through deamination events neighboring sites....
Early screening of lung nodules is mainly done manually by reading the patient's CT. This approach time-consuming laborious and prone to leakage misdiagnosis. Current methods for nodule detection face limitations such as high cost obtaining large-scale, high-quality annotated datasets poor robustness when dealing with data varying quality. The challenges include accurately detecting small irregular nodules, well ensuring model generalization across different sources. Therefore, this paper...
Abstract Adaptation of visual cortical cells' responses is observed following repeated presentation grating stimuli. Grating adaptation believed to exist only at the level. The purpose this study was see if also occurs in lateral geniculate nucleus. We studied 164 relay cells layer A and A1 dorsal nucleus (LGNd) Normal cats, as well cats which cortex ablated, were studied. investigated using gratings different contrasts orientations. results showed following: (1) reduced 46% LGNd recorded....
This work proposes a simple but effective attention mechanism, namely Skip Attention (SA), for monaural singing voice separation (MSVS).First, the SA, embedded in convolutional encoder-decoder network (CEDN), realizes an attentiondriven and dependency modeling repetitive structures of music source.Second, replacing popular skip connection CEDN, effectively controls flow lowlevel (vocal musical) features to output improves feature sensitivity accuracy MSVS.Finally, we implement proposed SA on...
Latent factor collaborative filtering (CF) has been a widely used technique for recommender system by learning the semantic representations of users and items. Recently, explainable recommendation attracted much attention from research community. However, trade-off exists between explainability performance where metadata is often needed to alleviate dilemma. We present novel feature mapping approach that maps uninterpretable general features onto interpretable aspect features, achieving both...
Deep convolutional neural networks (CNNs) trained with logistic and softmax losses have made significant advancement in visual recognition tasks computer vision. When training data exhibit class imbalances, the class-wise reweighted version of are often used to boost performance unweighted version. In this paper, motivated explain reweighting mechanism, we explicate learning property those two loss functions by analyzing necessary condition (e.g., gradient equals zero) after CNNs converge a...
Previous research has found that functional connectivity (FC) can accurately predict the identity of a subject performing task and type being performed. These results are replicated using large data set collected at Ohio State University Center for Cognitive Behavioral Brain Imaging. This work introduces novel perspective on prediction: blood-oxygen-level-dependent variability (BV). Conceptually, BV is region-specific measure based variance within each brain region. simple to compute,...
Abstract The development of magnetic resonance imaging (MRI) techniques has defined modern neuroimaging. Since its inception, tens thousands studies using such as functional MRI and diffusion weighted have allowed for the non-invasive study brain. Despite fact that is routinely used to obtain data neuroscience research, there been no widely adopted standard organizing describing collected in an experiment. This renders sharing reusing (within or between labs) difficult if not impossible...
Abstract Introduction Maddox suggested that there were four convergence subtypes, each driven by a different stimulus. The purpose of this study was to assess the neural correlates for accommodative convergence, proximal (convergence stimulus provided), disparity and voluntary (no specific provided) using functional magnetic resonance imaging (fMRI). Methods Ten subjects (mean age = 24.4 years) with normal binocular vision participated. blood oxygenation level‐dependent (BOLD) signals brain...
Accurate predictive modeling in clinical research enables effective early intervention that patients are most likely to benefit from. However, due the complex biological nature of disease progression, capturing highly non-linear information from low-level input features is quite challenging. This requires models with high-capacity. In practice, datasets often limited size, bringing danger overfitting for high-capacity models. To address these two challenges, we propose a deep multi-task...
Status characteristics theory provides a theoretical explanation for why social status promotes influence in collectively oriented task groups. It argues that differences produce expectation states, which are anticipations of task-related contributions. Those with an advantage more influential, contribute often to group discussions, and so on. The authors conducted the first experimental test while participants were magnetic resonance imaging machine. This permitted measurement neural...
The reconstruction of gene regulatory network from time course microarray data can help us comprehensively understand the biological system and discover pathogenesis cancer other diseases. But how to correctly efficiently decifer high-throughput expression is a big challenge due relatively small amount observations curse dimensionality. Computational biologists have developed many statistical inference machine learning algorithms analyze data. In previous studies, correctness an inferred...
Convolutional neural networks (CNNs) have achieved state-of-the-art performance on various tasks in computer vision. However, recent studies demonstrate that these models are vulnerable to carefully crafted adversarial samples and suffer from a significant drop when predicting them. Many methods been proposed improve robustness (e.g., training new loss functions learn adversarially robust feature representations). Here we offer unique insight into the predictive behavior of CNNs they tend...