A5076412381- Medical Image Segmentation Techniques
- Advanced Neural Network Applications
- Medical Imaging Techniques and Applications
- Radiomics and Machine Learning in Medical Imaging
- Cell Image Analysis Techniques
- Medical Imaging and Analysis
- Advanced X-ray and CT Imaging
- Cerebrospinal fluid and hydrocephalus
- AI in cancer detection
- Functional Brain Connectivity Studies
- Advanced Image Processing Techniques
- Advanced MRI Techniques and Applications
- Gene expression and cancer classification
- Generative Adversarial Networks and Image Synthesis
- Anatomy and Medical Technology
- Molecular Biology Techniques and Applications
- Brain Tumor Detection and Classification
- MRI in cancer diagnosis
- Single-cell and spatial transcriptomics
- Glioma Diagnosis and Treatment
- COVID-19 diagnosis using AI
- Fetal and Pediatric Neurological Disorders
- Dental Radiography and Imaging
- Metabolomics and Mass Spectrometry Studies
- Spinal Fractures and Fixation Techniques
Shanghai Jiao Tong University
2019-2022
University of North Carolina at Chapel Hill
2020-2022
Imaging Center
2021-2022
ShanghaiTech University
2021
Guangdong University of Foreign Studies
2018
In medical imaging such as PET-MR attenuation correction and MRI-guided radiation therapy, synthesizing CT images from MR plays an important role in obtaining tissue density properties. Recently deep-learning-based image synthesis techniques have attracted much attention because of their superior ability for mapping. However, most the current methods require large scales paired data, which greatly limits usage. Efforts been made to relax a restriction, cycle-consistent adversarial networks...
Deformable registration is fundamental to longitudinal and population-based image analyses. However, it challenging precisely align infant brain MR images of the same subject, as well cross-sectional different subjects, due fast development during infancy. In this paper, we propose a recurrently usable deep neural network for images. There are three main highlights our proposed method. (i) We use tissue segmentation maps registration, instead intensity images, tackle issue rapid contrast...
Brain magnetic resonance (MR) segmentation for hydrocephalus patients is considered as a challenging work. Encoding the variation of brain anatomical structures from different individuals cannot be easily achieved. The task becomes even more difficult especially when image data are considered, which often have large deformations and differ significantly normal subjects. Here, we propose novel strategy with hard soft attention modules to solve problems MR images. Our main contributions...
The human brain is not only efficiently but also "redundantly" connected. redundancy design could help the maintain resilience to disease attacks. This paper explores subnetwork-level dynamics and potential of such metrics in studies. As such, we looked into specific functional subnetworks, including those associated with high-level functions. We investigated how subnetwork change along Alzheimer's (AD) progression major depressive disorder (MDD), two disorders that share similar...
The amount of available biological data has exploded since the emergence high-throughput technologies, which is not only revolting way we recognize molecules and diseases but also bringing novel analytical challenges to bioinformatics analysis. In recent years, deep learning become a dominant technique in science. However, classification accuracy plagued with domain discrepancy. Notably, presence multiple batches, discrepancy typically happens between individual batches. Most pairwise...
Deformable image registration is fundamental to longitudinal and population analysis. Geometric alignment of the infant brain MR images challenging, owing rapid changes in appearance association with development. In this paper, we propose an infant-dedicated deep network that uses auto-context strategy gradually refine deformation fields obtain highly accurate correspondences. Instead training multiple networks, our method estimates by invoking a single times for iterative refinement. The...
Motion estimation is a fundamental step in dynamic medical image processing for the assessment of target organ anatomy and function. However, existing image-based motion methods, which optimize field by evaluating local similarity, are prone to produce implausible estimation, especially presence large motion. In addition, correct anatomical topology difficult be preserved as global context not well incorporated into estimation. this study, we provide novel framework dense-sparse-dense (DSD),...
Objective: Deformable brain MR image registration is challenging due to large inter-subject anatomical variation. For example, the highly complex cortical folding pattern makes it hard accurately align corresponding structures of individual images. In this paper, we propose a novel deep learning way simplify difficult problem Methods: We train morphological simplification network (MS-Net), which can generate "simple" with less details based on "complex" input. With MS-Net, complexity fixed...
After the economic development has entered new normal stage, manufacturing industry in Guangzhou faces challenges and opportunities, it urgently needs to change way of its economy development. This paper uses RSCA index analyze Guangzhou’s problems existing industry, compare situation six different areas along B&R. The results show that, a long steady comparative advantage garment textile sector, metal products leather sector while is disadvantage smelting processing chemical...
Thermal ablation is a minimally invasive procedure for treat-ing small or unresectable tumors. Although CT widely used guiding procedures, the contrast of tumors against surrounding normal tissues in images often poor, aggravating difficulty accurate thermal ablation. In this paper, we propose fast MR-CT image registration method to overlay pre-procedural MR (pMR) onto an intra-procedural (iCT) liver By first using Cycle-GAN model with mutual information constraint generate synthesized (sCT)...
Brain magnetic resonance (MR) segmentation for hydrocephalus patients is considered as a challenging work. Encoding the variation of brain anatomical structures from different individuals cannot be easily achieved. The task becomes even more difficult especially when image data are considered, which often have large deformations and differ significantly normal subjects. Here, we propose novel strategy with hard soft attention modules to solve problems MR images. Our main contributions...