A5028809808- Medical Imaging and Analysis
- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Dental Radiography and Imaging
- Bone health and osteoporosis research
- Bone and Joint Diseases
- Medical Image Segmentation Techniques
- Spine and Intervertebral Disc Pathology
- Spinal Fractures and Fixation Techniques
- Rheumatoid Arthritis Research and Therapies
- Bone fractures and treatments
- 3D Shape Modeling and Analysis
- Musculoskeletal synovial abnormalities and treatments
- Orthopaedic implants and arthroplasty
- MRI in cancer diagnosis
- Diabetic Foot Ulcer Assessment and Management
- Orthopedic Infections and Treatments
- Total Knee Arthroplasty Outcomes
- Medical Imaging and Pathology Studies
- Transplantation: Methods and Outcomes
- Advanced X-ray and CT Imaging
- Hepatocellular Carcinoma Treatment and Prognosis
- Advanced MRI Techniques and Applications
- Hip disorders and treatments
- Anatomy and Medical Technology
Institute for Biomedical Engineering
2020-2021
Friedrich Miescher Institute
2020-2021
University of Zurich
2021
Radboud University Nijmegen
2021
Imperial College London
2021
Technical University of Munich
2016-2020
Medical University of Vienna
2008-2020
Klinikum rechts der Isar
2016-2019
University of California, San Francisco
2012-2013
TU Wien
2006
To compare opportunistic quantitative CT (QCT) with dual energy X-ray absorptiometry (DXA) in their ability to predict incident vertebral fractures. We included 84 patients aged 50 years and older, who had routine including the lumbar spine DXA within a 12-month period (baseline) as well follow-up imaging after at least 12 months or sustained an fracture documented earlier. Patients bone disorders aside from osteoporosis were excluded. Fracture status trabecular mineral density (BMD)...
Abstract High-resolution peripheral quantitative computed tomography (HR-pQCT) has recently been introduced as a clinical research tool for in vivo assessment of bone quality. The utility this technology to address important skeletal health questions requires translation standardized multicenter data pools. Our goal was evaluate the feasibility pooling HR-pQCT imaging trials. Reproducibility experiments were performed using structure and composition-realistic phantoms constructed from...
Our study proposed an automatic pipeline for opportunistic osteoporosis screening using 3D texture features and regional vBMD multi-detector CT images. A combination of different local global outperformed the showed high discriminative power to identify patients with vertebral fractures. Many at risk undergo computed tomography (CT) scans, usable (non-dedicated) screening. We compared performance volumetric bone mineral density (vBMD) a random forest classifier based on separate without...
We hypothesized that machine learning analysis based on texture information from the preoperative MRI can predict IDH mutational status in newly diagnosed WHO grade II and III gliomas. This retrospective study included total 79 consecutive patients with a or glioma. Local binary pattern features were generated B0 fractional anisotropy (FA) diffusion tensor imaging. Using training set of 59 patients, single hidden layer neural network was then trained to status. The model validated prediction...
To introduce and validate an automated unsupervised multi-parametric method for segmentation of the subcutaneous fat muscle regions to determine adipose tissue (SAT) intermuscular (IMAT) areas based on data from a quantitative chemical shift-based water-fat separation approach.Unsupervised standard k-means clustering was used define sets similar features (k = 2) within whole multi-modal image after separation. The processing chain composed three primary stages: tissue, muscle, bone region...
Multi-class segmentation of vertebrae is a non-trivial task mainly due to the high correlation in appearance adjacent vertebrae. Hence, such calls for consideration both global and local context. Based on this motivation, we propose two-staged approach that, given computed tomography dataset spine, segments five lumbar simultaneously labels them. The first stage employs multi-layered perceptron performing non-linear regression locating region using second stage, comprised fully-convolutional...
To use and test a labeling algorithm that operates on two-dimensional reformations, rather than three-dimensional data to locate identify vertebrae.The authors improved the Btrfly Net, fully convolutional network architecture described by Sekuboyina et al, which works sagittal coronal maximum intensity projections (MIPs) augmented it with two additional components: spine localization adversarial priori learning. Furthermore, variants of training schemes incorporated anatomic knowledge into...
ABSTRACT Bone morphogenetic proteins (BMPs) are potent osteogenic that induce new bone formation in vivo. However, their effect on healing the trabecular surfaces remains challenging. We evaluated safety and efficacy of recombinant human BMP6 (rhBMP6) applied within an autologous blood coagulum (ABC) a surgically created wedge defect proximal tibia patients undergoing high tibial osteotomy (HTO) for varus deformity medial osteoarthritis knee. enrolled 20 HTO randomized, placebo-controlled,...
ABSTRACT Traumatic odontoid fractures (TOFs) have been described as the most common injury affecting C-spine in elderly. Previous studies identified degenerative changes and bone loss important predisposing factors. However, their interaction respective age-adjusted impact needs further clarification. We conducted a retrospective analysis of 5303 patients (aged ≥60 years) admitted to level I trauma center between January 2008 2016 who underwent CT imaging C-spine. Ninety-two with TOF 80...
Automatic non-invasive assessment of hepatocellular carcinoma (HCC) malignancy has the potential to substantially enhance tumor treatment strategies for HCC patients. In this work we present a novel framework automatically characterize lesions from DWI images. We predict in two steps: As first step segment using cascaded fully convolutional neural networks (CFCN). A 3D network (SurvivalNet) then predicts lesions' segmentation. formulate task as classification problem with classes being "low...
Purpose To characterize bone microarchitecture and quantify strength in lung transplant (LT) recipients by using high-resolution (HR) peripheral quantitative computed tomographic (CT) imaging of the ultradistal radius. Materials Methods After study approval local ethics committee, all participants provided written informed consent. Included were 118 (58 LT [mean age, 46.8 years ± 1.9; 30 women, 28 men] 60 control 39.9 41 19 men]) between April 2010 May 2012. HR CT radius was performed...
The microarchitecture of the trabecular bone is an highly informative feature for osteoporosis assessment. High resolution peripheral quantitative computed tomography permits its in-vivo observation at a 82 μm. In this paper we propose approach that assesses based on texture features extracted from bone. method three-dimensional as local descriptors structure in A clustering space indicates characteristic classes are repeatedly detected across subjects. distribution those allows...