A5061759611- AI in cancer detection
- Nonmelanoma Skin Cancer Studies
- Advanced Neural Network Applications
- Cutaneous Melanoma Detection and Management
- Parallel Computing and Optimization Techniques
- Advanced Image Processing Techniques
- Visual Attention and Saliency Detection
- Metabolism, Diabetes, and Cancer
- Advanced Image and Video Retrieval Techniques
- Medical Image Segmentation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Image and Object Detection Techniques
- Pancreatic function and diabetes
- Face recognition and analysis
- Colorectal Cancer Screening and Detection
- Image and Video Quality Assessment
Lund University
2023
Anhui University
2018-2023
Colorectal cancer is one of the leading causes death worldwide. Polyps are early symptoms colorectal and prone to malignant transformation. Polyp segmentation colonoscopy images can help diagnosis. However, existing studies on polyp face two main difficulties: blurry boundaries, close resemblances between polyps surrounding tissues. The former may lead partial segmentations, while latter result in false positive segmentations. This paper proposes a new framework tackle challenges. In this...
This article proposes a novel spectral domain based solution to the challenging polyp segmentation. The main contribution is on an interesting finding of significant existence middle frequency sub-band during CNN process. Consequently, Sub-Band Attention (SBA) module proposed, which uniformly adopts either high or sub-bands encoder features boost decoder and thus concretely improve feature discrimination. A strong supplying informative also very important, while we highly value...
The pancreatic islet is a highly structured micro-organ that produces insulin in response to rising blood glucose. Here we develop label-free and automatic imaging approach visualize the islets situ diabetic rodents by synchrotron radiation X-ray phase-contrast microtomography (SRμCT) at ID17 station of European Synchrotron Radiation Facility. large-size images (3.2 mm × 15.97 mm) were acquired pancreas STZ-treated mice GK rats. Each was dissected 3000 reconstructed images. image datasets...
Abstract The large variations of polyp sizes and shapes the close resemblances polyps to their surroundings call for features with long‐range information in rich scales strong discrimination. This article proposes two parallel structured modules building those features. One is Transformer Inception module (TI) which applies Transformers different reception fields input thus enriches them more scales. other Local‐Detail Augmentation (LDA) spatial channel attentions each block locally augments...
Robust skin lesion segmentation of dermoscopic images is still very difficult. Recent methods often take the combinations CNN and Transformer for feature abstraction multi-scale features further classification. Both types combination in general rely on some forms fusion. This paper considers these fusions from two novel points view. For abstraction, viewed as affinity exploration different patch tokens can be applied to attend multiple scales. Consequently, a new fusion module,...
Automatic segmentation of skin lesion is still challenging due to ambiguous boundary and noise interference regions. Recent exiting Transformer-based methods often directly apply Transformer obtain long-range dependency overcome these problems. However, they generally do not consider that patch partitioning strategy could lead the loss local details around boundaries. Furthermore, dependencies across windows only represent global information at a coarse level. Therefore, limitations, two...