A5024764575- COVID-19 diagnosis using AI
- Radiomics and Machine Learning in Medical Imaging
- Advanced Neural Network Applications
- Lung Cancer Diagnosis and Treatment
- Robotics and Sensor-Based Localization
- Head and Neck Cancer Studies
- AI in cancer detection
- Advanced Radiotherapy Techniques
- Infective Endocarditis Diagnosis and Management
- Video Surveillance and Tracking Methods
- Adversarial Robustness in Machine Learning
- Cell Image Analysis Techniques
- Advanced Image and Video Retrieval Techniques
- Aortic Disease and Treatment Approaches
- Cardiac Valve Diseases and Treatments
- Medical Image Segmentation Techniques
Beijing Institute of Technology
2019-2020
Ministry of Industry and Information Technology
2019-2020
Recently, automatic visual data understanding from drone platforms becomes highly demanding. To facilitate the study, Vision Meets Drone Object Detection in Image Challenge is held second time conjunction with 17-th International Conference on Computer (ICCV 2019), focuses image object detection drones. Results of 33 algorithms are presented. For each participating detector, a short description provided appendix. Our goal to advance state-of-the-art and provide comprehensive evaluation...
Drones or general Unmanned Aerial Vehicles (UAVs), endowed with computer vision function by on-board cameras and embedded systems, have become popular in a wide range of applications. However, real-time scene parsing through object detection running on UAV platform is very challenging, due to limited memory computing power devices. To deal these challenges, this paper we propose learn efficient deep detectors channel pruning convolutional layers. end, enforce channel-level sparsity layers...
Accurate morphological information on aortic valve cusps is critical in treatment planning. Image segmentation necessary to acquire this information, but manual tedious and time consuming. In paper, we propose a fully automatic method from CT images by combining two deep neural networks, spatial configuration-Net for detecting anatomical landmarks U-Net of components. A total 258 volumes end systolic diastolic phases, which include cases with without severe calcifications, were collected...
Computed tomography (CT) images are currently being adopted as the visual evidence for COVID-19 diagnosis in clinical practice. Automated detection of infection from CT based on deep models is important faster examination. Unfortunately, collecting large-scale training data systematically early stage difficult. To address this problem, we explore feasibility learning lung and segmentation a single radiological image by resorting to synthesizing diverse images. Specifically, propose novel...
The success of deep learning has been witnessed as a promising technique for computer-aided biomedical image analysis, due to end-to-end framework and availability large-scale labelled samples. However, in many cases techniques suffer from the small sample (SSL) dilemma caused mainly by lack annotations. To be more practical this paper we survey key SSL that help relieve suffering combining with development related computer vision applications. In order accelerate clinical usage analysis...
Automated infection measurement and COVID-19 diagnosis based on Chest X-ray (CXR) imaging is important for faster examination, where segmentation an essential step assessment quantification. However, due to the heterogeneity of difficulty annotating infected regions precisely, learning automated CXRs remains a challenging task. We propose novel approach, called DRR4Covid, learn from digitally reconstructed radiographs (DRRs). DRR4Covid consists infection-aware DRR generator, network, domain...
We propose adversarial constrained-CNN loss, a new paradigm of loss methods, for weakly supervised medical image segmentation. In the paradigm, prior knowledge is encoded and depicted by reference masks, further employed to impose constraints on segmentation outputs through learning with masks. Unlike pseudo label methods segmentation, such masks are used train discriminator rather than network, thus not required be paired specific images. Our only greatly facilitates imposing network's...
Radiation therapy is a primary and effective NasoPharyngeal Carcinoma (NPC) treatment strategy. The precise delineation of Gross Tumor Volumes (GTVs) Organs-At-Risk (OARs) crucial in radiation treatment, directly impacting patient prognosis. Previously, the GTVs OARs was performed by experienced oncologists. Recently, deep learning has achieved promising results many medical image segmentation tasks. However, for NPC segmentation, few public datasets are available model development...
Radiological image is currently adopted as the visual evidence for COVID-19 diagnosis in clinical. Using deep models to realize automated infection measurement and important faster examination based on radiological imaging. Unfortunately, collecting large training data systematically early stage difficult. To address this problem, we explore feasibility of learning from a single by resorting synthesizing diverse images. Specifically, propose novel conditional generative model, called...
Automated infection measurement and COVID-19 diagnosis based on Chest X-ray (CXR) imaging is important for faster examination. We propose a novel approach, called DRR4Covid, to learn automated segmentation CXRs from digitally reconstructed radiographs (DRRs). DRR4Covid comprises of an infection-aware DRR generator, classification and/or network, domain adaptation module. The generator able produce DRRs with adjustable strength radiological signs infection, generate pixel-level annotations...