A5089836011- Radiomics and Machine Learning in Medical Imaging
- Domain Adaptation and Few-Shot Learning
- COVID-19 diagnosis using AI
- AI in cancer detection
- Machine Learning and Algorithms
- Multiple Myeloma Research and Treatments
- Advanced X-ray and CT Imaging
- Bone health and treatments
- Hemodynamic Monitoring and Therapy
- Medical Imaging Techniques and Applications
- Electrical and Bioimpedance Tomography
- Respiratory Support and Mechanisms
- Lung Cancer Diagnosis and Treatment
- Medical Imaging and Analysis
- Interstitial Lung Diseases and Idiopathic Pulmonary Fibrosis
- Chemokine receptors and signaling
- Radiology practices and education
- Glioma Diagnosis and Treatment
- Multimodal Machine Learning Applications
- Colorectal Cancer Screening and Detection
- Sparse and Compressive Sensing Techniques
- Neonatal Respiratory Health Research
- Body Composition Measurement Techniques
- CNS Lymphoma Diagnosis and Treatment
- Artificial Intelligence in Healthcare and Education
Medical University of Vienna
2015-2023
HES-SO University of Applied Sciences and Arts Western Switzerland
2020
Klinik und Poliklinik für Nuklearmedizin
2019
TU Wien
2013
Machine learning methods offer great promise for fast and accurate detection prognostication of COVID-19 from standard-of-care chest radiographs (CXR) computed tomography (CT) images. Many articles have been published in 2020 describing new machine learning-based models both these tasks, but it is unclear which are potential clinical utility. In this systematic review, we search EMBASE via OVID, MEDLINE PubMed, bioRxiv, medRxiv arXiv papers preprints uploaded January 1, to October 3,...
Abstract Background Automated segmentation of anatomical structures is a crucial step in image analysis. For lung computed tomography, variety approaches exists, involving sophisticated pipelines trained and validated on different datasets. However, the clinical applicability these across diseases remains limited. Methods We compared four generic deep learning various datasets two readily available algorithms. performed evaluation routine imaging data with more than six disease patterns...
Medical imaging is a central part of clinical diagnosis and treatment guidance. Machine learning has increasingly gained relevance because it captures features disease response that are relevant for therapeutic decision-making. In practice, the continuous progress image acquisition technology or diagnostic procedures, diversity scanners, evolving protocols hamper utility machine learning, as prediction accuracy on new data deteriorates, models become outdated due to these domain shifts. We...
Machine learning is rapidly gaining importance in radiology. It allows for the exploitation of patterns imaging data and patient records a more accurate precise quantification, diagnosis, prognosis. Here, we outline basics machine relevant radiology, review current state art, limitations, challenges faced as these techniques become an important building block precision medicine. Furthermore, discuss roles can play clinical routine research predict how it might change field
Content based image retrieval is highly relevant in medical imaging, since it makes vast amounts of imaging data accessible for comparison during diagnosis. Finding similarity measures that reflect diagnostically relationships challenging, the overall appearance variability high compared to often subtle signatures diseases. To learn models capture relationship between semantic clinical information and elements at scale, we have rely on generated routine (images radiology reports), expert...
Abstract Objectives To identify and evaluate predictive lung imaging markers their pathways of change during progression idiopathic pulmonary fibrosis (IPF) from sequential data an IPF cohort. test if these predict outcome. Methods We studied radiological disease in 76 patients with IPF, including overall 190 computed tomography (CT) examinations the chest. An algorithm identified candidates for patterns marking by computationally clustering visual CT features. A classification selected...
Acute respiratory distress syndrome (ARDS) constitutes a major factor determining the clinical outcome in polytraumatized patients. Early prediction of ARDS is crucial for timely supportive therapy to reduce morbidity and mortality. The objective this study was develop test machine learning-based method early derived from first computed tomography scan patients after admission hospital.One hundred twenty-three (86 male 37 female, age 41.2 ± 16.4) with an injury severity score (ISS) 16 or...
The purpose of this study was to improve risk stratification smoldering multiple myeloma patients, introducing new 3D-volumetry based imaging biomarkers derived from whole-body MRI. Two-hundred twenty MRIs 63 patients with were retrospectively analyzed and all focal lesions >5mm manually segmented for volume quantification. total tumor volume, speed growth (development the over time), number lesions, development time recent biomarker '>1 lesion' International Myeloma Working Group compared,...
The prognostic roles of clinical and laboratory markers have been exploited to model risk in patients with primary CNS lymphoma, but these approaches do not fully explain the observed variation outcome. To date, neuroimaging or molecular information is used. aim this study was determine utility radiomic features capture clinically relevant phenotypes, link those profiles for enhanced stratification.
Electrical impedance tomography (EIT) is a promising imaging technique for bedside monitoring of lung function. It easily applicable, cheap and requires no ionizing radiation, but clinical interpretation EIT-images still not standardized. One the reasons this ill-posed nature EIT, allowing range possible images to be produced-rather than single explicit solution. Thus, further advance EIT technology application, thorough examinations EIT-image reconstruction settings-i.e., mathematical...
Machine learning in medical imaging during clinical routine is impaired by changes scanner protocols, hardware, or policies resulting a heterogeneous set of acquisition settings. When training deep model on an initial static set, performance and reliability suffer from characteristics as data targets may become inconsistent. Continual can help to adapt models the changing environment continuous stream. However, continual manual expert labelling requires substantial effort. Thus, ways use...
The reliable and timely stratification of bone lesion evolution risk in smoldering Multiple Myeloma plays an important role identifying prime markers the disease's advance improving patients' outcome. In this work we provide asymmetric cascade network for longitudinal prediction future lesions T1 weighted whole body MR images. proposed cascaded architecture, consisting two distinct configured U-Nets, first detects regions subsequently predicts within bones a patch based way. algorithm...
Imaging in clinical routine is subject to changing scanner protocols, hardware, or policies a typically heterogeneous set of acquisition hardware. Accuracy and reliability deep learning models suffer from those changes as data targets become inconsistent with their initial static training set. Continual can adapt continuous stream imaging environment. Here, we propose method for continual active on medical images. It recognizes shifts additions new sources - domains -, adapts accordingly,...