A5088035307- Orthopaedic implants and arthroplasty
- Medical Imaging and Analysis
- Hip disorders and treatments
- Medical Image Segmentation Techniques
- Total Knee Arthroplasty Outcomes
- Medical Imaging Techniques and Applications
- Spinal Fractures and Fixation Techniques
- Advanced X-ray and CT Imaging
- Spine and Intervertebral Disc Pathology
- 3D Shape Modeling and Analysis
- Radiomics and Machine Learning in Medical Imaging
- Pelvic and Acetabular Injuries
- Anatomy and Medical Technology
- Surgical Simulation and Training
- Orthopedic Infections and Treatments
- Advanced Neural Network Applications
- Bone fractures and treatments
- Cardiac Valve Diseases and Treatments
- Robotics and Sensor-Based Localization
- Soft Robotics and Applications
- Scoliosis diagnosis and treatment
- Dental Radiography and Imaging
- Domain Adaptation and Few-Shot Learning
- COVID-19 diagnosis using AI
- AI in cancer detection
Shanghai Jiao Tong University
2008-2025
Nanjing Medical University
2021-2022
Hengyang Normal University
2022
University of Bern
2011-2020
Clinical Orthopaedics and Related Research
2019-2020
Swiss National Science Foundation
2019
University Hospital of Bern
2001-2018
Chinese People's Liberation Army
2014-2015
Heidelberg University
2012
Klinikum Ludwigshafen
2012
Deep convolutional networks (ConvNets) have achieved unprecedented performances on many computer vision tasks. However, their adaptations to crowd counting single images are still in infancy and suffer from severe over-fitting. Here we propose a new learning strategy produce generalizable features by way of deep negative correlation (NCL). More specifically, deeply learn pool decorrelated regressors with sound generalization capabilities through managing intrinsic diversities. Our proposed...
Both acetabular undercoverage (hip dysplasia) and overcoverage (pincer-type femoroacetabular impingement) can result in hip osteoarthritis. In contrast to undercoverage, there is a lack of information on radiographic reference values for excessive coverage.(1) How do common parameters differ hips with deficient or an acetabulum relation control group; (2) what are the determined from these data under- overcoverage?We retrospectively compared 11 describing anatomy among dysplasia (26...
International challenges have become the standard for validation of biomedical image analysis methods. Given their scientific impact, it is surprising that a critical common practices related to organization has not yet been performed. In this paper, we present comprehensive conducted up now. We demonstrate importance and show lack quality control consequences. First, reproducibility interpretation results often hampered as only fraction relevant information typically provided. Second, rank...
Knowledge of whole heart anatomy is a prerequisite for many clinical applications. Whole segmentation (WHS), which delineates substructures the heart, can be very valuable modeling and analysis functions heart. However, automating this challenging due to large variation shape, different image qualities data. To achieve goal, an initial set training data generally needed constructing priors or training. Furthermore, it difficult perform comparisons between methods, largely differences in...
Quantification of cerebral white matter hyperintensities (WMH) presumed vascular origin is key importance in many neurological research studies. Currently, measurements are often still obtained from manual segmentations on brain MR images, which a laborious procedure. Automatic WMH segmentation methods exist, but standardized comparison the performance such lacking. We organized scientific challenge, developers could evaluate their method multi-center/-scanner image dataset, giving an...
Cephalometric analysis is an essential clinical and research tool in orthodontics for the orthodontic treatment planning. This paper presents evaluation of methods submitted to Automatic X-Ray Landmark Detection Challenge, held at IEEE International Symposium on Biomedical Imaging 2014 with on-site competition. The challenge was set explore compare automatic landmark detection application cephalometric X-ray images. Methods were evaluated a common database including cephalograms 300 patients...
Accurate segmentation of infant brain magnetic resonance (MR) images into white matter (WM), gray (GM), and cerebrospinal fluid is an indispensable foundation for early studying growth patterns morphological changes in neurodevelopmental disorders. Nevertheless, the isointense phase (approximately 6-9 months age), due to inherent myelination maturation process, WM GM exhibit similar levels intensity both T1-weighted T2-weighted MR images, making tissue very challenging. Although many efforts...
Anteroposterior pelvic radiographs are the gold standard of imaging for mechanical hip problems. However, correct interpretation is difficult because projected morphologic features acetabulum and nearly all routinely used parameters depend on individual position, which can vary considerably during acquisition. We developed software that recreates acetabular rim measured as if obtained in a standardized orientation. The vertical horizontal distances between two easy identifiable points were...
The purpose of this study was to validate the accuracy, consistency, and reproducibility/reliability a new method for correction pelvic tilt rotation radiographic hip parameters pincer type femoroacetabular impingement on an anteroposterior radiograph. Thirty cadaver hips 100 randomized, blinded AP radiographs were used investigation. To detect software calculated femoral head coverage classic determined with our compared reference measurements based CT scans or conventional in neutral...
Background Acetabular anatomy on AP pelvic radiographs depends orientation during radiograph acquisition. However, not all parameters may change to a clinically relevant degree with differences in orientation. This issue influence the diagnosis of acetabular pathologies and planning corrective surgery (reorientation or rim trimming). this point, it has been well characterized. Questions/purposes We asked (1) which radiographic clinical setting when normalized neutral orientation; (2) do an...
Nonlinear regression has been extensively employed in many computer vision problems (e.g., crowd counting, age estimation, affective computing). Under the umbrella of deep learning, two common solutions exist i) transforming nonlinear to a robust loss function which is jointly optimizable with convolutional network, and ii) utilizing ensemble networks. Although some improved performance achieved, former may be lacking due intrinsic limitation choosing single hypothesis latter suffer from...
In this paper, we address the problems of fully automatic localization and segmentation 3D vertebral bodies from CT/MR images. We propose a learning-based, unified random forest regression classification framework to tackle these two problems. More specifically, in first stage, is solved with where aggregate votes set randomly sampled image patches get probability map center target body given image. The resultant then further regularized by Hidden Markov Model (HMM) eliminate potential...
Background: It remains unclear whether decreased femoral version (FV) causes anterior intra- or extra-articular femoroacetabular impingement (FAI). Therefore, we evaluated symptomatic hips with FV, and without cam pincer FAI, by using computed tomography (CT)–based virtual 3-dimensional (3D) simulation compared this group patients normal FV asymptomatic hips. Purpose: To investigate (1) the osseous range of motion, (2) acetabular zones, (3) hip is extra- intra-articular in FAI. Study Design:...
We present a method to address the challenging problem of segmentation lumbar vertebrae from CT images acquired with varying fields view. Our is based on cascaded 3D Fully Convolutional Networks (FCNs) consisting localization FCN and FCN. More specifically, in first step we train regression (we call it "LocalizationNet") find bounding box region. After that, U-net like "Segmentation-Net") then developed, which after training, can perform pixel-wise multi-class map cropped lumber region...
While surgical navigation offers the opportunity to accurately place an acetabular component, questions remain as best goal for component positioning in individual patients. Overall functional orientation of pelvis after surgery is one most important variables surgeon consider when determining proper orientation.We measured variation pelvic tilt 30 patients before THA and effect on same more than a year THA.Each patient had CT study CT-based standing supine radiographs surgery. Pelvic was...
Few studies have examined the order in which a spinal osteotomy and total hip replacement (THR) are to be performed for patients with ankylosing spondylitis. We retrospectively reviewed 28 consecutive spondylitis who underwent both THR from September 2004 November 2012. In cohort 22 had before (group 1), six 2). The mean duration of follow-up was 3.5 years (2 9). sagittal Cobb angle vertebral segment corrected pre-operative kyphosis 32.4 (SD 15.5°) post-operative lordosis 29.6 11.2°) (p <...