A5076674630- AI in cancer detection
- Radiomics and Machine Learning in Medical Imaging
- Medical Image Segmentation Techniques
- Gene expression and cancer classification
- HIV Research and Treatment
- Bioinformatics and Genomic Networks
- Functional Brain Connectivity Studies
- Cell Image Analysis Techniques
- HIV/AIDS drug development and treatment
- Machine Learning in Bioinformatics
- Colorectal Cancer Screening and Detection
- Digital Imaging for Blood Diseases
- Advanced Neuroimaging Techniques and Applications
- Robotics and Sensor-Based Localization
- Single-cell and spatial transcriptomics
- Brain Tumor Detection and Classification
- Cancer-related molecular mechanisms research
- Image Retrieval and Classification Techniques
- HIV/AIDS Research and Interventions
- Cancer Genomics and Diagnostics
- EEG and Brain-Computer Interfaces
- Advanced Neural Network Applications
- Medical Imaging and Analysis
- Ferroptosis and cancer prognosis
- Alzheimer's disease research and treatments
Nanjing University of Aeronautics and Astronautics
2015-2025
Anhui Medical University
2019-2025
Shanghai University
2025
Qingdao University of Science and Technology
2014-2024
Wuhan No.1 Hospital
2022-2024
Hôpital de Morges
2024
Indiana University School of Medicine
2019-2023
Ministry of Industry and Information Technology
2022-2023
Indiana University – Purdue University Indianapolis
2019-2023
University of Electronic Science and Technology of China
2023
Chest computed tomography (CT) becomes an effective tool to assist the diagnosis of coronavirus disease-19 (COVID-19). Due outbreak COVID-19 worldwide, using computed-aided technique for classification based on CT images could largely alleviate burden clinicians. In this paper, we propose <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b> daptive xmlns:xlink="http://www.w3.org/1999/xlink">F</b> eature...
Advances in computational algorithms and tools have made the prediction of cancer patient outcomes using pathology feasible. However, predicting clinical from pre-treatment histopathologic images remains a challenging task, limited by poor understanding tumor immune micro-environments. In this study, an automatic, accurate, comprehensive, interpretable, reproducible whole slide image (WSI) feature extraction pipeline known as, IMage-based Pathological REgistration Segmentation Statistics...
To fully utilize the power of single-cell RNA sequencing (scRNA-seq) technologies for identifying cell lineages and bona fide transcriptional signals, it is necessary to combine data from multiple experiments. We present BERMUDA (Batch Effect ReMoval Using Deep Autoencoders), a novel transfer-learning-based method batch effect correction in scRNA-seq data. effectively combines different batches with vastly population compositions amplifies biological signals by transferring information among...
BACKGROUNDHIV-1 viremia that is not suppressed by combination antiretroviral therapy (ART) generally attributed to incomplete medication adherence and/or drug resistance. We evaluated individuals referred clinicians for nonsuppressible (plasma HIV-1 RNA above 40 copies/mL) despite reported ART and the absence of resistance current regimen.METHODSSamples were collected from at least 2 time points 8 donors who had more than 6 months. Single templates obtained plasma viral outgrowth cultured...
The integrative analysis of histopathological images and genomic data has received increasing attention for studying the complex mechanisms driving cancers. However, most image-genomic studies have been restricted to combining with single modality (e.g., mRNA transcription or genetic mutation), thus neglect fact that molecular architecture cancer is manifested at multiple levels, including genetic, epigenetic, transcriptional, post-transcriptional events. To address this issue, we propose a...
Tumor-infiltrating lymphocytes (TILs) and their spatial characterizations on whole-slide images (WSIs) of histopathology sections have become crucial in diagnosis, prognosis, treatment response prediction for different cancers. However, fully automatic assessment TILs WSIs currently remains a great challenge because the heterogeneity large size WSIs. We present an pipeline based cascade-training U-net to generate high-resolution TIL maps WSIs.We global cell-level 43 quantitative image...
Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal cancer that typically demonstrates resistance to chemotherapy. Tumor-associated macrophages (TAMs) are essential in tumor microenvironment (TME) regulation, including promoting chemoresistance. However, the specific TAM subset and mechanisms behind this promotion remain unclear. We employ multi-omics strategies, single-cell RNA sequencing (scRNA-seq), transcriptomics, multicolor immunohistochemistry (mIHC), flow cytometry,...
In the domain of AI-driven healthcare, deep learning models have markedly advanced pneumonia diagnosis through X-ray image analysis, thus indicating a significant stride in efficacy medical decision systems. This paper presents novel approach utilizing convolutional neural network that effectively amalgamates strengths EfficientNetB0 and DenseNet121, it is enhanced by suite attention mechanisms for refined classification. Leveraging pre-trained models, our employs multi-head, self-attention...
Brain region-of-interest (ROI) segmentation based on structural magnetic resonance imaging (MRI) scans is an essential step for many computer-aid medical image analysis applications. Due to low intensity contrast around ROI boundary and large inter-subject variance, it has been remaining a challenging task effectively segment brain ROIs from MR images. Even though several deep learning methods have developed, most of them do not incorporate shape priors take advantage the regularity...
Multi-modal fusion has become an important data analysis technology in Alzheimer's disease (AD) diagnosis, which is committed to effectively extract and utilize complementary information among different modalities. However, most of the existing methods focus on pursuing common feature representation by transformation, ignore discriminative structural samples. In addition, use high-order extraction, such as deep neural network, it difficult identify biomarkers. this paper, we propose a novel...
Abstract Schizophrenia is a highly heritable psychiatric disorder characterized by widespread functional and structural brain abnormalities. However, previous association studies between MRI polygenic risk were mostly ROI-based single modality analyses, rather than identifying brain-based multimodal predictive biomarkers. Based on schizophrenia scores (PRS) from healthy white people within the UK Biobank dataset ( N = 22,459), we discovered robust PRS-associated pattern with smaller gray...
We propose DEGAS (Diagnostic Evidence GAuge of Single cells), a novel deep transfer learning framework, to disease information from patients cells. call such transferrable "impressions," which allow individual cells be associated with attributes like diagnosis, prognosis, and response therapy. Using simulated data ten diverse single-cell patient bulk tissue transcriptomic datasets glioblastoma multiforme (GBM), Alzheimer's (AD), multiple myeloma (MM), we demonstrate the feasibility,...
Multi-modal imaging data fusion has attracted much attention in medical analysis because it can provide complementary information for more accurate analysis. Integrating functional and structural multi-modal been increasingly used the diagnosis of brain diseases, such as epilepsy. Most existing methods focus on feature space different modalities but ignore valuable high-order relationships among samples discriminative fused features classification. In this paper, we propose a novel framework...
The NCI Retrovirus Integration Database is a MySql-based relational database created for storing and retrieving comprehensive information about retroviral integration sites, primarily, but not exclusively, HIV-1. accessible to the public submission or extraction of data originating from experiments aimed at collecting related sites including: site into host genome, virus family subtype, origin sample, gene exons/introns associated with integration, proviral orientation. Information...
Multi-atlas based segmentation methods have shown their effectiveness in brain regions-of-interesting (ROIs) segmentation, by propagating labels from multiple atlases to a target image on the similarity between patches and atlas images. Most of existing multiatlas use intensity features calculate pair for label fusion. In particular, using only low-level cannot adequately characterize complex appearance patterns (e.g., high-order relationship voxels within patch) magnetic resonance (MR) To...
Sleep staging is a vital process for evaluating sleep quality and diagnosing sleep-related diseases. Most of the existing automatic methods focus on time-domain information often ignore transformation relationship between stages. To deal with above problems, we propose Temporal-Spectral fused Attention-based deep neural Network model (TSA-Net) staging, using single-channel electroencephalogram (EEG) signal. The TSA-Net composed two-stream feature extractor, context learning, conditional...
Multi-modal brain networks characterize the complex connectivities among different regions from structure and function aspects, which have been widely used in analysis of diseases. Although many multi-modal network fusion methods proposed, most them are unable to effectively extract spatio-temporal topological characteristics while fusing modalities. In this paper, we develop an adaptive multi-channel graph convolution (GCN) framework with contrast learning, not only can mine both...
Survival analysis is to estimate the survival time for an individual or a group of patients, which valid solution cancer treatments. Recent studies suggested that integrative histopathological images and genomic data can better predict patients than simply using single bio-marker, different bio-markers may provide complementary information. However, given multi-modal contain irrelevant redundant features, it still challenge design distance metric simultaneously discover significant features...