A5101477602- Radiomics and Machine Learning in Medical Imaging
- Liver Disease Diagnosis and Treatment
- Advanced X-ray and CT Imaging
- Advanced Steganography and Watermarking Techniques
- Medical Imaging Techniques and Applications
- Advanced Research in Systems and Signal Processing
- Spondyloarthritis Studies and Treatments
- Photoacoustic and Ultrasonic Imaging
- Organ Transplantation Techniques and Outcomes
- Advanced Image and Video Retrieval Techniques
- Liver Disease and Transplantation
- Orthopedic Infections and Treatments
- Bone and Joint Diseases
- Hip disorders and treatments
- Induction Heating and Inverter Technology
- Digital Media Forensic Detection
- Chaos-based Image/Signal Encryption
- Text and Document Classification Technologies
- Sports Performance and Training
- Bone Tumor Diagnosis and Treatments
- Image Retrieval and Classification Techniques
- Sports injuries and prevention
- Lower Extremity Biomechanics and Pathologies
- AI in cancer detection
- Spectroscopy Techniques in Biomedical and Chemical Research
Third Affiliated Hospital of Southern Medical University
2023-2025
Guangdong Provincial People's Hospital
2025
Guangdong Academy of Medical Sciences
2025
Beijing Jiaotong University
2008
Photoacoustic microscopy imaging could be used to evaluate the severity of liver fibrosis by means an analysis structural and functional characteristics lobules.
Liver fibrosis represents a progressive pathological condition that can culminate in severe hepatic dysfunction, potentially advancing to cirrhosis and liver cancer. The extent of is intrinsically associated with the quantity collagen fibers. Although biopsy ultrasound imaging are standard diagnostic tools, their application constrained by risks significant complications variability different investigators, respectively. In this study, we utilized linear dichroism photoacoustic microscopy...
Abstract Objectives To develop a machine learning (ML)-based model using MRI and clinical risk factors to enhance diagnostic accuracy for axial spondyloarthritis (axSpA). Methods We retrospectively analyzed datasets from four centers (A–D), focusing on patients with chronic low back pain. A subset center was used prospective validation. deep (DL) based ResNet50 constructed sacroiliac joint MRI. Clinical variables were integrated DL scores in ML algorithms distinguish axSpA non-axSpA...
Abstract Background Accurately classifying primary bone tumors is crucial for guiding therapeutic decisions. The National Comprehensive Cancer Network guidelines recommend multimodal images to provide different perspectives the comprehensive evaluation of tumors. However, in clinical practice, most patients’ medical are often incomplete. This study aimed build a deep learning model using incomplete from X-ray, CT, and MRI alongside characteristics classify as benign, intermediate, or...
To develop a deep learning (DL) model for segmenting fat metaplasia (FM) on sacroiliac joint (SIJ) MRI and further DL classifying axial spondyloarthritis (axSpA) non-axSpA. This study retrospectively collected 706 patients with FM who underwent SIJ from center 1 (462 axSpA 186 non-axSpA) 2 (37 21 non-axSpA). Patients were divided into the training, validation, internal test sets (n = 455, 64, 129). used as external set. We developed UNet-based to segment FM. Based segmentation results,...
We present the first explicit construction of two-sided lossless expanders in unbalanced setting (bipartite graphs that have many more nodes on left than right). Prior to our work, all known constructions achieved only one-sided expansion. Specifically, we show constructed by Kalev and Ta-Shma (RANDOM'22)--that are based multiplicity codes introduced Kopparty, Saraf, Yekhanin (STOC'11)--are, fact, expanders. Using bipartite expander, easily obtain (non-bipartite) expander $N$ vertices with...
A reversible data hiding scheme based on the companding technique and difference expansion (DE) of triplets is proposed in this paper. The employed to increase number expandable triplets. capacity consumed by location map recording expanded positions largely decreased. As a result, considerably increased. experimental results reveal that high can be achieved at low embedding distortion.
A novel automatic image annotation (AIA) scheme is proposed based on multiple-instance learning (MIL). For a given concept, manifold ranking (MR) first employed to MIL (referred as MR-MIL) for effectively mining the positive instances (i. e. regions in images) embedded bags images). With mined instances, semantic model of concept built by probabilistic output SVM classifier. The experimental results reveal that high accuracy can be achieved at region-level.