A5060249884- Cell Image Analysis Techniques
- AI in cancer detection
- Neuroinflammation and Neurodegeneration Mechanisms
- Molecular Biology Techniques and Applications
- T-cell and B-cell Immunology
- Single-cell and spatial transcriptomics
- Radiomics and Machine Learning in Medical Imaging
- CAR-T cell therapy research
- Cancer-related molecular mechanisms research
- Neurological Disease Mechanisms and Treatments
- Cancer Genomics and Diagnostics
- Acute Ischemic Stroke Management
- Traumatic Brain Injury and Neurovascular Disturbances
- RNA modifications and cancer
- Gene expression and cancer classification
- CRISPR and Genetic Engineering
- Immune Cell Function and Interaction
- Field-Flow Fractionation Techniques
- Catalysts for Methane Reforming
- Moyamoya disease diagnosis and treatment
- Protein Tyrosine Phosphatases
- Liver physiology and pathology
- Hepatitis B Virus Studies
- Reservoir Engineering and Simulation Methods
- Caveolin-1 and cellular processes
The University of Texas Southwestern Medical Center
2020-2024
Southwestern Medical Center
2020-2023
Northwest Women's and Children's Hospital
2022
University of Pennsylvania
2020-2021
Nanjing University
2014-2019
Nanjing General Hospital of Nanjing Military Command
2014-2019
Fudan University
2019
Nanjing Medical University
2015-2019
Zhejiang University
2018
Sir Run Run Shaw Hospital
2018
Functions of astrocytes in the rehabilitation after ischemic stroke, especially their impacts on inflammatory processes, remain controversial. This study uncovered two phenotypes astrocytes, which one was helpful, and other harmful to anoxic neurons brain ischemia. We tested levels factors including TNF-a, IL-6, IL-10, iNOS, IL-1beta, CXCL10 primary at 0 h, 6 12 24 48 h OGD, grouped hypoxia into iNOS-positive (iNOS(+)) iNOS-negative (iNOS(−)) by magnetic bead sorting, then co-cultured groups...
Microscopic examination of pathology slides is essential to disease diagnosis and biomedical research. However, traditional manual tissue laborious subjective. Tumor whole-slide image (WSI) scanning becoming part routine clinical procedures produces massive data that capture tumor histologic details at high resolution. Furthermore, the rapid development deep learning algorithms has significantly increased efficiency accuracy analysis. In light this progress, digital fast a powerful tool...
Abstract Current clustering analysis of spatial transcriptomics data primarily relies on molecular information and fails to fully exploit the morphological features present in histology images, leading compromised accuracy interpretability. To overcome these limitations, we have developed a multi-stage statistical method called iIMPACT. It identifies defines histology-based domains based AI-reconstructed images context gene expression measurements, detects domain-specific differentially...
Toll-like receptor 4 (TLR4) is a proinflammatory cascade initiator in poststroke inflammation. In this study, miR-1906, novel regulator of TLR4, was identified via silico analysis and microRNA profiling male adult mice its expression then quantitated the ischemic hemisphere. We found miR-1906 to be significantly brain enriched hemisphere even more drastically peri-infarct regions. Furthermore, vitro experiments demonstrated that, during oxygen–glucose deprivation, increased glial cells but...
Astrocytes, the most numerous cells in human brain, play a central role metabolic homeostasis following hypoxic injury. Caveolin-1 (Cav-1), transmembrane scaffolding protein, has been shown to converge prosurvival signaling nerve system. The present study aimed investigate of Cav-1 hypoxia-induced astrocyte We also examined how alleviates apoptotic death. To this end, primary astrocytes were exposed oxygen-glucose deprivation (OGD) for 6 h and subsequent 24-h reoxygenation mimic OGD...
Phosphatase PP2A expression levels are positively correlated to the clinical severity of systemic lupus erythematosus (SLE) and IL17A cytokine overproduction, indicating a potential role in controlling TH17 differentiation inflammation. By generating mouse strain with ablation catalytic subunit α peripheral mature T cells (PP2A cKO), we demonstrate that complex is essential for differentiation. These cKO mice had reduced cell numbers less severe disease an experimental autoimmune...
Intraluminal middle cerebral artery occlusion (MCAO) is the most widely used model of stroke. We aimed to predict outcome MCAO using a combination fine behavioural tests for prediction unsuccessful surgery in mice leading no infarction, haemorrhage and unexpected death. was performed on adult under guidance laser-Doppler flowmetry (LDF) warrant decrease regional blood flow (rCBF) MCA territory. Four outcomes were defined according histological analysis: death (death within 24 h...
Recent advances in foundation models have revolutionized model development digital pathology, reducing dependence on extensive manual annotations required by traditional methods. The ability of to generalize well with few-shot learning addresses critical barriers adapting diverse medical imaging tasks. This work presents the Granular Box Prompt Segment Anything Model (GB-SAM), an improved version (SAM) fine-tuned using granular box prompts limited training data. GB-SAM aims reduce dependency...
Polyploidy, the duplication of entire genome within a single cell, is significant characteristic cells in many tissues, including liver. The quantification hepatic ploidy typically relies on flow cytometry and immunofluorescence (IF) imaging, which are not widely available clinical settings due to high financial time costs. To improve accessibility for samples, we developed computational algorithm quantify using hematoxylin-eosin (H&E) histopathology images, commonly obtained during...
Abstract The spatially resolved transcriptomics (SRT) field has revolutionized our ability to comprehensively leverage image and molecular profiles elucidate spatial organization of cellular microenvironments. Current clustering analysis SRT data primarily relies on information fails fully exploit the morphological features present in histology images, leading compromised accuracy interpretability. To overcome these limitations, we have developed a multi-stage statistical method called...
Abstract Shengli lignite (SL) and Shenmu-Fugu subbituminous coal (SFSBC) were subjected to thermal dissolution in cyclohexane at 300°C for 1 h under hydrogen. The reaction mixtures extracted with petroleum ether (PE) the PE-extractable fractions (PEEFs) analyzed a gas chromatography/mass spectrometer. results show that PEEFs from thermally dissolved SL (TDSL) SFSBC (TDSFSBC) mainly consist of alkanes, alkenes, arenes, oxygen-containing organic compounds (OCOCs); yields alkanes alkenes PEEF...
A high-temperature coal tar collected from Xuzhou Coking Plant, Saint-Gobain Pipelines (Xuzhou) Co., Ltd. was extracted with petroleum ether (PE) under ultrasonic irradiation and in a Soxhlet extractor. The residue then methanol the All extracts were analyzed gas chromatography/mass spectrometry (GC/MS). More than 250 organic compounds detected PE-soluble fraction, but no GC/MS-detectable species (GC/MS-DSs) methanol-soluble PE-insoluble indicating that all GC/MS-DSs as PE. include various...
The ethanol-soluble portion from supercritical ethanolysis of Huolinguole lignite was separated by silica-gel column chromatography. Hydroxybenzenes and normal alkanes were concentrated to large extents, respectively. A series ethyl alkanoates alkenoates detected. Two alkylpyridines enriched in high concentrations.
Genetic mouse models of skeletal abnormalities have demonstrated promise in the identification phenotypes relevant to human diseases. Traditionally, are assessed by manually examining radiographs, a tedious and potentially error-prone process. In response, this study developed deep learning-based model that streamlines measurement murine bone lengths from radiographs an accurate reproducible manner. A detection pipeline utilizing Keypoint R-CNN algorithm with EfficientNet-B3 feature...
Recent advances in aligning large language models with human preferences have corroborated the growing importance of best-of-N distillation (BOND). However, iterative BOND algorithm is prohibitively expensive practice due to sample and computation inefficiency. This paper addresses problem by revealing a unified game-theoretic connection between self-play alignment, which unifies seemingly disparate algorithmic paradigms. Based on connection, we establish novel framework, WIN rate Dominance...