A5080719369- Advanced Neuroimaging Techniques and Applications
- Advanced MRI Techniques and Applications
- Cell Image Analysis Techniques
- Advanced Neural Network Applications
- AI in cancer detection
- Fetal and Pediatric Neurological Disorders
- Image Processing Techniques and Applications
- Multimodal Machine Learning Applications
- Radiomics and Machine Learning in Medical Imaging
- 3D Shape Modeling and Analysis
- COVID-19 diagnosis using AI
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Digital Imaging for Blood Diseases
- MRI in cancer diagnosis
- Domain Adaptation and Few-Shot Learning
- Human Pose and Action Recognition
- Brain Tumor Detection and Classification
- Generative Adversarial Networks and Image Synthesis
- Anomaly Detection Techniques and Applications
- Medical Image Segmentation Techniques
- Video Analysis and Summarization
- Single-cell and spatial transcriptomics
- Cooperative Communication and Network Coding
- Medical Imaging Techniques and Applications
The University of Sydney
2018-2025
Fudan University Shanghai Cancer Center
2023-2025
Nanjing Jiangning Hospital
2025
China Pharmaceutical University
2025
Sichuan University
2020-2024
West China Hospital of Sichuan University
2024
University at Buffalo, State University of New York
2016-2023
Shanghai Medical College of Fudan University
2023
Shanghai Cancer Institute
2023
Northwestern Polytechnical University
2023
Discrete point cloud objects lack sufficient shape descriptors of 3D geometries. In this paper, we present a novel method for aggregating hypothetical curves in clouds. Sequences connected points (curves) are initially grouped by taking guided walks the clouds, and then subsequently aggregated back to augment their pointwise features. We provide an effective implementation proposed aggregation strategy including curve grouping operator followed operator. Our was benchmarked on several...
Recent advances in unsupervised domain adaptation (UDA) techniques have witnessed great success cross-domain computer vision tasks, enhancing the generalization ability of data-driven deep learning architectures by bridging distribution gaps. For UDA-based object detection methods, majority them alleviate bias inducing domain-invariant feature generation via adversarial strategy. However, their discriminators limited classification due to unstable training process. Therefore, extracted...
Radiography imaging protocols focus on particular body regions, therefore producing images of great similarity and yielding recurrent anatomical structures across patients. To exploit this structured information, we propose the use Space-aware Memory Queues for In-painting Detecting anomalies from radiography (abbreviated as SQUID). We show that SQUID can taxonomize ingrained into patterns; in inference, it identify (unseen/modified patterns) image. surpasses 13 state-of-the-art methods...
Purpose: To propose a deep learning-based reconstruction framework for ultrafast and robust diffusion tensor imaging fiber tractography. Methods: We SuperDTI to learn the nonlinear relationship between diffusion-weighted images (DWIs) corresponding tensor-derived quantitative maps as well Super DTI bypasses fitting procedure, which is known be highly susceptible noise motion in DWIs. The network trained tested using datasets from Human Connectome Project patients with ischemic stroke....
Automated detection and segmentation of individual nuclei in histopathology images is important for cancer diagnosis prognosis. Due to the high variability appearances numerous overlapping objects, this task still remains challenging. Deep learning based semantic instance models have been proposed address challenges, but these methods tend concentrate on either global or local features hence suffer from information loss. In work, we propose a panoptic model which incorporates an auxiliary...
Scene understanding is a critical problem in computer vision. In this paper, we propose 3D point-based scene graph generation (SGG <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">point</inf> ) framework to effectively bridge perception and reasoning achieve under-standing via three sequential stages, namely construction, reasoning, inference. Within the stage, an EDGE-oriented Graph Convolutional Network (EdgeGCN) created exploit...
Dense captioning in 3D point clouds is an emerging vision-and-language task involving object-level scene understanding. Apart from coarse semantic class prediction and bounding box regression as traditional object detection, dense aims at producing a further finer instance-level label of natural language description on visual appearance spatial relations for each interest. To detect describe objects scene, following the spirit neural machine translation, we propose transformer-based...
White matter fiber clustering is an important strategy for white parcellation, which enables quantitative analysis of brain connections in health and disease. In combination with expert neuroanatomical labeling, data-driven a powerful tool creating atlases that can model anatomy across individuals. While widely used approaches have shown good performance using classical unsupervised machine learning techniques, recent advances deep reveal promising direction toward fast effective clustering....
Automated segmentation of multiple sclerosis (MS) lesions in brain imaging is challenging due to the high variability lesion characteristics. Based on generative adversarial network (GAN), we propose a semantic framework MS-GAN localize MS multimodal magnetic resonance (MRI), which consists one encoder-decoder generator G and discriminators D corresponding input modalities. For design generator, adopt an deep learning architecture with bypass spatial information from encoder decoder, helps...
Endoplasmic reticulum (ER) perturbations are novel subcellular effectors involved in the ischaemia-reperfusion injury. As an ER stress-inducible protein, mesencephalic astrocyte-derived neurotrophic factor (MANF) has been proven to be increased during ischaemic brain However, role of MANF liver ischaemia reperfusion (I/R) injury not yet studied.To investigate process ischaemia-reperfusion, Hepatocyte-specific knockout (MANFhep-/- ) mice and their wild-type (WT) littermates were used our...
We present a novel weakly-supervised framework for classifying whole slide images (WSIs). WSIs, due to their gigapixel resolution, are commonly processed by patch-wise classification with patch-level labels. However, labels require precise annotations, which is expensive and usually unavailable on clinical data. With image-level only, would be sub-optimal inconsistency between the patch appearance label. To address this issue, we posit that WSI analysis can effectively conducted integrating...
Deep learning methods have been successfully used in various computer vision tasks. Inspired by that success, deep has explored magnetic resonance imaging (MRI) reconstruction. In particular, integrating and model-based optimization shown considerable advantages. However, a large amount of labeled training data is typically needed for high reconstruction quality, which challenging some MRI applications. this paper, we propose novel method, named DURED-Net, enables interpretable...
Hand‑foot syndrome (HFS) is defined as a major adverse reaction to capecitabine; however, the underlying mechanisms remain unclear. In total, 85 patients who were taking oral capecitabine included in present study and these divided into HFS‑positive HFS‑negative groups. Serum samples collected from an untargeted metabolomics analysis was conducted using ultra‑high performance liquid chromatography‑mass spectrometry/mass spectrometry. The aimed investigate presence of metabolites serum that...
Abstract Motivation The mutations of cancers can encode the seeds their own destruction, in form T-cell recognizable immunogenic peptides, also known as neoantigens. It is computationally challenging, however, to accurately prioritize potential neoantigen candidates according ability activating immunoresponse, especially when somatic are abundant. Although a few prioritization methods have been proposed address this issue, advanced machine learning model that specifically designed tackle...