A5088498169- Radiomics and Machine Learning in Medical Imaging
- Head and Neck Cancer Studies
- Pulmonary Hypertension Research and Treatments
- Cancer-related molecular mechanisms research
- Venous Thromboembolism Diagnosis and Management
- Lung Cancer Diagnosis and Treatment
- Gastric Cancer Management and Outcomes
- MRI in cancer diagnosis
- Circular RNAs in diseases
- MicroRNA in disease regulation
- Colorectal Cancer Screening and Detection
- Medical Image Segmentation Techniques
- Medical Imaging and Analysis
- Tracheal and airway disorders
- Inflammatory Biomarkers in Disease Prognosis
- Advanced Neural Network Applications
- AI in cancer detection
- PI3K/AKT/mTOR signaling in cancer
- Web Data Mining and Analysis
- Pelvic and Acetabular Injuries
- Ocular Surface and Contact Lens
- Cardiovascular Effects of Exercise
- Gastrointestinal Bleeding Diagnosis and Treatment
- Brain Tumor Detection and Classification
- Clostridium difficile and Clostridium perfringens research
Sun Yat-sen University
2019-2025
Sun Yat-sen University Cancer Center
2019-2025
Yunnan University
2020-2024
State Key Laboratory of Oncology in South China
2020-2023
Second People's Hospital of Yunnan Province
2019-2022
State Key Laboratory of Ophthalmology
2022
Children's Hospital of Chongqing Medical University
2022
Chongqing Medical University
2022
Kunming Medical University
2019-2020
Hebei Medical University
2016
Accurate diagnosis and prognosis prediction are conducive to early intervention improvement of medical care for natural killer/T cell lymphoma (NKTCL). Artificial intelligence (AI)-based systems developed based on nasopharynx magnetic resonance imaging. The diagnostic achieve areas under the curve 0.905–0.960 in detecting malignant nasopharyngeal lesions distinguishing NKTCL from carcinoma independent validation datasets. In comparison human radiologists, show higher accuracies than resident...
Multilabel pathological tissue segmentation is a vital task in computational pathology that aims to semantically segment different tissues within images. Fully and weakly supervised models have demonstrated impressive performances this regard. However, still face challenges, such as the poor performance of nondominant samples limited effectiveness aggregation functions conveying supervisory signals. To address these issues, we propose two key contributions: introduction graph attention...
Accurate tumor diagnosis by pathologists relies on identifying specific morphological characteristics. However, summarizing these unique features in classifications can be challenging. Although deep learning models have been extensively studied for classification, their indirect and subjective interpretation obstructs from comprehending the model discerning accountable classifications. In this study, we introduce a new approach utilizing Style Generative Adversarial Networks, which enables...
Abstract Background To investigate the feasibility of detecting local recurrent nasopharyngeal carcinoma (rNPC) using unenhanced magnetic resonance images (MRI) and optimize a layered management strategy for follow-up with deep learning model. Methods Deep models based on 3D DenseNet or ResNet frames unique sequence (T1WI, T2WI, T1WIC) combination T1WI T2WI sequences (T1_T2) were developed to detect rNPC. A d eep-learning- ssisted r ecurrent N PC de tecting s imultaneous t actic (DARNDEST)...
Abstract The incidence of hypoxic pulmonary hypertension (HPH) is increasing. Accumulating evidence suggests that long noncoding RNAs (lncRNAs) play an important role in HPH, but the functions and mechanism have yet to be fully elucidated. In present study, we established a HPH rat model with 8 h hypoxia exposure (10% O2) per day for 21 days. High-throughput sequencing identified 60 differentially expressed (DE) lncRNAs, 20 DE miRNAs 695 mRNAs lung tissue. qRT-PCR verified accuracy results....
Delineating the lesion area is an important task in image-based diagnosis. Pixel-wise classification a popular approach to segmenting region of interest. However, at fuzzy boundaries, such methods usually result glitches, discontinuity or disconnection, inconsistent with fact that lesions are solid and smooth. To overcome these problems provide efficient, accurate, robust concise solution simplifies whole segmentation pipeline AI-assisted applications, we propose BézierSeg model which...
This study aims to evaluate the value of a serum metabolomics-based metabolic signature for locoregionally advanced nasopharyngeal carcinoma (LA-NPC) patients, thereby assisting clinical decisions.In this retrospective study, total 320 LA-NPC patients were randomly divided into training set (ca. 70%; n = 224) and validation 30%; 96). Serum samples analyzed using widely targeted metabolomics. Univariate multivariate Cox regression analyses used identify candidate metabolites related...
To determine the diagnostic accuracy of techniques with chronic thromboembolic pulmonary hypertension (CTEPH) patients via a protocol for systemic review and network meta-analysis.We will search PubMed, EMBASE, Web Science, Google Scholar from inception to October 1, 2018. The reference lists retrieved articles are also consulted. Quality Assessment Diagnostic Accuracy Studies 2 (QUADAS-2) be used assess risk bias in each study. direct meta-analyses, ranking competing tests by STATA 12.0...
Objective: To investigate the effects of Streptococcuspneumoniae (Spn) colonization and change upper airway microbiome on clinical manifestations in children with respiratory syncytial virus (RSV) infection. Methods: A retrospective cohort included 508 RSV-infected pneumonia hospitalized Respiratory Department Children's Hospital Chongqing Medical University from July 2009 to 2018. total cases (RSV non-sequencing group) were divided into 2 groups: Spn (RSV+Spn without according detection for...
It is imperative to optimally utilize virtues and obviate defects of fully automated analysis expert knowledge in new paradigms healthcare. We present a deep learning-based semiautomated workflow (RAINMAN) with 12,809 follow-up scans among 2,172 patients treated nasopharyngeal carcinoma from three centers (ChiCTR.org.cn, Chi-CTR2200056595). A boost diagnostic performance reduced workload was observed RAINMAN compared the original manual interpretations (internal vs. external: sensitivity,...