A5083590594- AI in cancer detection
- Generative Adversarial Networks and Image Synthesis
- Radiomics and Machine Learning in Medical Imaging
- Electrochemical Analysis and Applications
- Infant Health and Development
- Semiconductor materials and devices
- Conducting polymers and applications
- Parallel Computing and Optimization Techniques
- Security and Verification in Computing
- Image Processing Techniques and Applications
- VLSI and Analog Circuit Testing
- Digital Imaging for Blood Diseases
- Cancer-related molecular mechanisms research
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Electrochemical sensors and biosensors
- Molecular Sensors and Ion Detection
- Advanced Memory and Neural Computing
- Advanced Vision and Imaging
- Cell Image Analysis Techniques
- Advanced Image Processing Techniques
- Metal-Organic Frameworks: Synthesis and Applications
- COVID-19 diagnosis using AI
- Advanced Malware Detection Techniques
- Advanced Numerical Analysis Techniques
Zhangzhou Normal University
2025
Tsinghua University
2022-2024
University Town of Shenzhen
2024
South China University of Technology
2024
Tsinghua–Berkeley Shenzhen Institute
2024
As an essential step in the pathological diagnosis, histochemical staining can show specific tissue structure information and, consequently, assist pathologists making accurate diagnoses. Clinical kidney histopathological analyses usually employ more than one type of staining: H&E, MAS, PAS, PASM, etc. However, due to interference colors among multiple stains, it is not easy perform simultaneously on biological tissue. To address this problem, we propose a network based unpaired training...
In histopathology, the tissue slides are usually stained by common H&E stain or special stains (MAS, PAS, and PASM, etc.) to clearly show specific structures. The rapid development of deep learning provides a good solution generate virtual staining images significantly reduce time labor costs associated with histochemical staining. However, most existing methods need train model for every two stains, which consumes lot computing resources increasing types. To address this problem, we propose...
Different types of staining highlight different structures in organs, thereby assisting diagnosis. However, due to the impossibility repeated staining, we cannot obtain stained slides same tissue area. Translating slide that is easy (e.g., H&E) difficult MT, PAS) a promising way solve this problem. some regions are closely connected other regions, and maintain connection, they often have complex translate, which may lead wrong translations. In paper, propose Attention-Based Varifocal...
In clinical practice, frozen section (FS) images can be utilized to obtain the immediate pathological results of patients in operation due their fast production speed. However, compared with formalin-fixed and paraffin-embedded (FFPE) images, FS greatly suffer from poor quality. Thus, it is great significance transfer image FFPE one, which enables pathologists observe high-quality operation. obtaining paired quite hard, so difficult accurate using supervised methods. Apart this, stain faces...
The common computational pathology tissue slides mainly include Fresh Frozen (FF) and Formalin-Fixed Paraffin-Embedded (FFPE) slides. FF slide has low quality but is easy to prepare, while the FFPE one opposite, which widely used in pathological preservation high-precision diagnosis. Our goal generate images based on patches order quickly obtain high-quality a short time provide doctors with convenient accurate diagnostic evidence. We propose ST-MKSC, an FF2FFPE image translation network,...
Histochemical staining is a critical step in the diagnosis of cancer, where hematoxylin-eosin (H&E) stain used most commonly clinical practice. However, H&E images often cannot be for making accurate diagnoses. To this end, pathologists must perform immunohistochemical (IHC) stain, which time-consuming and costly. In field computer-aided diagnosis, existing models can virtually generate IHC images, but they require pixel-aligned data annotations from pathologists, are difficult to obtained....