A5003321058- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Lung Cancer Diagnosis and Treatment
- Medical Image Segmentation Techniques
- Advanced Radiotherapy Techniques
- Atomic and Subatomic Physics Research
- Advanced MRI Techniques and Applications
- Advanced Neural Network Applications
- Cardiac Imaging and Diagnostics
- Domain Adaptation and Few-Shot Learning
- AI in cancer detection
- Ultrasound and Hyperthermia Applications
- Advanced Image and Video Retrieval Techniques
- Cerebrovascular and Carotid Artery Diseases
- COVID-19 diagnosis using AI
- Acute Ischemic Stroke Management
- Robotics and Sensor-Based Localization
- Ultrasound Imaging and Elastography
- Artificial Intelligence in Healthcare and Education
- Cardiovascular Function and Risk Factors
- Dental Radiography and Imaging
- Intracerebral and Subarachnoid Hemorrhage Research
- Forensic Anthropology and Bioarchaeology Studies
- Digital Imaging for Blood Diseases
- Soft Robotics and Applications
Compliment Corporation (United States)
2018-2020
Seattle University
2018-2020
Chinese Academy of Sciences
2020
National University of Defense Technology
2020
University of Oulu
2020
Toa Pharmaceutical (Japan)
2020
Tencent (China)
2020
General Electric (United States)
2012-2019
Chengdu University of Information Technology
2019
Sichuan University
2019
Background Coronavirus disease 2019 (COVID-19) has widely spread all over the world since beginning of 2020. It is desirable to develop automatic and accurate detection COVID-19 using chest CT. Purpose To a fully framework detect CT evaluate its performance. Materials Methods In this retrospective multicenter study, deep learning model, neural network (COVNet), was developed extract visual features from volumetric scans for COVID-19. community-acquired pneumonia (CAP) other non-pneumonia...
EMPIRE10 (Evaluation of Methods for Pulmonary Image REgistration 2010) is a public platform fair and meaningful comparison registration algorithms which are applied to database intrapatient thoracic CT image pairs. Evaluation nonrigid techniques nontrivial task. This compounded by the fact that researchers typically test only on their own data, varies widely. For this reason, reliable assessment different has been virtually impossible in past. In work we present results launch phase...
To evaluate the performance of a novel three-dimensional (3D) joint convolutional and recurrent neural network (CNN-RNN) for detection intracranial hemorrhage (ICH) its five subtypes (cerebral parenchymal, intraventricular, subdural, epidural, subarachnoid) in non-contrast head CT.A total 2836 subjects (ICH/normal, 1836/1000) from three institutions were included this ethically approved retrospective study, with 76,621 slices CT scans. ICH annotated by independent experienced radiologists,...
Characterization of lung nodules as benign or malignant is one the most important tasks in cancer diagnosis, staging and treatment planning. While variation appearance remains large, there a need for fast robust computer aided system. In this work, we propose an end-to-end trainable multi-view deep Convolutional Neural Network (CNN) nodule characterization. First, use median intensity projection to obtain 2D patch corresponding each dimension. The three images are then concatenated form...
Acute lung injury is characterized by heterogeneity of regional mechanical properties, which thought to be correlated with disease severity. The feasibility using respiratory input impedance (Z(rs)) and computed tomographic (CT) image registration for assessing parenchymal was evaluated. In six dogs, measurements Z(rs) before after oleic acid at various distending pressures were obtained, followed whole CT scans. Each spectrum fit a model incorporating variable distributions compliances....
Purpose: Lung function depends on lung expansion and contraction during the respiratory cycle. Respiratory-gated CT imaging 3D image registration can be used to locally estimate tissue (regional volume change) by computing determinant of Jacobian matrix deformation field. In this study, authors examine reproducibility Jacobian-based measures in two repeat 4DCT acquisitions mechanically ventilated sheep free-breathing humans. Methods: data from three white nine human subjects were for...
Purpose: Regional lung volume change as a function of inflation serves an index parenchymal and airway status well regional ventilation can be used to detect pathologic changes over time. In this article, we propose new measure mechanics --- the specific air by corrected Jacobian. Methods: 4DCT Xe-CT data sets from four adult sheep are in study. Nonlinear, 3D image registration is applied register acquired near end inspiration expiration. Approximately 200 annotated anatomical points...
Computerized automatic methods have been employed to boost the productivity as well objectiveness of hand bone age assessment. These approaches make predictions according whole X-ray images, which include other objects that may introduce distractions. Instead, our framework is inspired by clinical workflow (Tanner-Whitehouse) assessment, focuses on key components hand. The proposed composed two components: a Mask R-CNN subnet pixelwise segmentation and residual attention network for segments...
Accurate pulmonary image registration is a challenging problem when the lungs have deformation with large distance. In this work, we present nonrigid volumetric algorithm to track lung motion between pair of intrasubject CT images acquired at different inflation levels and introduce new vesselness similarity cost that improves intensity-only registration. Volumetric datasets from six human subjects were used in study. The performance four algorithms was compared without adding function....
We have previously developed a robotic ultrasound imaging system for motion monitoring in abdominal radiation therapy. Owing to the slow speed of image processing, our previous could only track motions under breath-hold. To overcome this limitation, novel 2D-based processing method tracking intra-fraction respiratory is proposed. Fifty-seven different anatomical features acquired from 27 sets 2D sequences were used study. Three with three healthy volunteers. The remaining datasets provided...
Ultrasound guided catheter insertion is a common procedure in current clinical practice, but it requires skilled ultrasound practitioner to correctly acquire the images so that and tip are well visualized. Automated detection of location can help procedural guidance as surveillance position post placement, especially when imaging being performed by non-sonographer. Accurate fast localization very challenging task because poor observability images. In this paper, we present novel algorithm...
Multi-modal medical image registration takes an essential role in image-based clinical diagnosis and surgical planning. It is not trivial due to appearance variations across different modalities. Rigidly aligning two images used register rigid body structure, it also usually the first step for deformable with a large discrepancy. In field of computer vision, one well-established method alignment find corresponding points from images, based on identified points. Our lies this category....
Intracranial hemorrhage (ICH) is a critical disease that requires immediate diagnosis and treatment. Accurate detection, subtype classification volume quantification of ICH are aspects in diagnosis. Previous studies have applied deep learning techniques for analysis but usually tackle the aforementioned tasks separate manner without taking advantage information sharing between tasks. In this paper, we propose multi-task fully convolutional network, ICHNet, simultaneous segmentation ICH. The...
This paper focuses on the semi-supervised object detection (SSOD) which makes good use of unlabeled data to boost performance. We face following obstacles when adapting knowledge distillation (KD) framework in SSOD. (1) The teacher model serves a dual role as and student, such that predictions images may limit upper bound student. (2) imbalance issue caused by large quantity consistent between student hinders an efficient transfer them. To mitigate these issues, we propose novel SSOD called...