A5026685751- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Advanced X-ray and CT Imaging
- Lung Cancer Diagnosis and Treatment
- Lymphoma Diagnosis and Treatment
- AI in cancer detection
- Nuclear Physics and Applications
- Machine Learning in Materials Science
- Robotics and Automated Systems
- Additive Manufacturing and 3D Printing Technologies
- Robotic Locomotion and Control
- Radiopharmaceutical Chemistry and Applications
- Artificial Intelligence in Healthcare and Education
- Sarcoma Diagnosis and Treatment
- MRI in cancer diagnosis
- Advanced MRI Techniques and Applications
- Image and Signal Denoising Methods
- Radiation Dose and Imaging
- Traditional Chinese Medicine Analysis
- Advanced Image Processing Techniques
- Colorectal Cancer Screening and Detection
- Veterinary Equine Medical Research
- Prostate Cancer Diagnosis and Treatment
- Additive Manufacturing Materials and Processes
- Fractal and DNA sequence analysis
Forschungszentrum Jülich
2023-2025
University of Augsburg
2022-2025
Life & Brain (Germany)
2024
University Medical Center Groningen
2018-2023
Maastricht University
2022
University Hospital Augsburg
2021-2022
Harvard University
2022
Maastro Clinic
2022
University of Groningen
2018-2021
University Hospital Carl Gustav Carus
2020
The image biomarker standardisation initiative (IBSI) is an independent international collaboration which works towards standardising the extraction of biomarkers from acquired imaging for purpose high-throughput quantitative analysis (radiomics). Lack reproducibility and validation studies considered to be a major challenge field. Part this lies in scantiness consensus-based guidelines definitions process translating into biomarkers. IBSI therefore seeks provide nomenclature definitions,...
Abstract Purpose To propose a new quality scoring tool, METhodological RadiomICs Score (METRICS), to assess and improve research of radiomics studies. Methods We conducted an online modified Delphi study with group international experts. It was performed in three consecutive stages: Stage#1, item preparation; Stage#2, panel discussion among EuSoMII Auditing Group members identify the items be voted; Stage#3, four rounds exercise by panelists determine eligible for METRICS their weights. The...
Abstract Purpose Quantitative prostate-specific membrane antigen (PSMA) PET analysis may provide for non-invasive and objective risk stratification of primary prostate cancer (PCa) patients. We determined the ability machine learning-based quantitative [ 18 F]DCFPyL metrics to predict metastatic disease or high-risk pathological tumor features. Methods In a prospective cohort study, 76 patients with intermediate- PCa scheduled robot-assisted radical prostatectomy extended pelvic lymph node...
Background 18 F‐fluoro‐2‐deoxy‐D‐Glucose positron emission tomography ( F‐ FDG PET ) radiomics has the potential to guide clinical decision making in cancer patients, but validation is required before can be implemented setting. The aim of this study was explore how feature space reduction and repeatability radiomic features are affected by various sources variation such as underlying data (e.g., object size uptake), image reconstruction methods settings, noise, discretization method,...
Accurate prognostic markers are urgently needed to identify diffuse large B-Cell lymphoma (DLBCL) patients at high risk of progression or relapse. Our purpose was investigate the potential added value baseline radiomics features international index (IPI) in predicting outcome after first-line treatment.Three hundred seventeen newly diagnosed DLBCL were included. Lesions delineated using a semi-automated segmentation method (standardized uptake ≥ 4.0), and 490 extracted. We used logistic...
The sensitivity of radiomic features to several confounding factors, such as reconstruction settings, makes clinical use challenging. To investigate the impact harmonized image reconstructions on feature consistency, a multicenter phantom study was performed using 3-dimensionally printed inserts reflecting realistic tumor shapes and heterogeneity uptakes. Methods: Tumors extracted from real PET/CT scans patients with non-small cell lung cancer served model for three inserts. Different...
Purpose The widely known field 'Radiomics' aims to provide an extensive image based phenotyping of e.g. tumors using a wide variety feature values extracted from medical images. Therefore, it is utmost importance that calculated by different institutes follow the same definitions. For this purpose, imaging biomarker standardization initiative (IBSI) provides detailed mathematical descriptions, as well (mathematical) test phantoms and corresponding reference values. We present here easy use...
Abstract Purpose Biomarkers that can accurately predict outcome in DLBCL patients are urgently needed. Radiomics features extracted from baseline [ 18 F]-FDG PET/CT scans have shown promising results. This study aims to investigate which lesion- and feature-selection approaches/methods resulted the best prediction of progression after 2 years. Methods A total 296 were included. 485 radiomics ( n = 5 conventional PET, 22 morphology, 50 intensity, 408 texture) for all individual lesions at...
This study investigated whether radiomic features extracted from pretreatment [<sup>18</sup>F]FDG PET could improve the prediction of both histopathologic tumor response and survival in patients with locally advanced cervical cancer (LACC) treated neoadjuvant chemoradiotherapy followed by surgery compared conventional parameters features. <b>Methods:</b> The medical records all consecutive LACC referred between July 2010 2016 were reviewed. PET/CT was performed before chemoradiotherapy....
In the acquisition of Magnetic Resonance (MR) images shorter scan times lead to higher image noise. Therefore, automatic denoising using deep learning methods is high interest. this work, we concentrate on MR containing line-like structures such as roots or vessels. particular, investigate if special characteristics these datasets (connectivity, sparsity) benefit from use loss functions for network training. We hereby translate Perceptual Loss 3D data by comparing feature maps untrained...
Radiomics is aimed at image-based tumor phenotyping, enabling application within clinical-decision-support-systems to improve diagnostic accuracy and allow for personalized treatment. The purpose was identify predictive 18-fluor-fluoro-2-deoxyglucose (18F-FDG) positron-emission tomography (PET) radiomic features predict recurrence, distant metastasis, overall survival in patients with head neck squamous cell carcinoma treated chemoradiotherapy.Between 2012 2018, 103 retrospectively (training...
Introduction: Radiomics features may predict outcome in diffuse large B-cell lymphoma (DLBCL). Currently, multiple segmentation methods are used to calculate metabolic tumor volume (MTV). We assessed the influence of method on discriminative power radiomics DLBCL for patient level and largest lesion. <b>Methods:</b> 50 baseline <sup>18</sup>F-fluorodeoxyglucose positron emission tomography computed (PET/CT) scans patients who progressed or relapsed within 2 years after diagnosis were matched...
In longitudinal oncological and brain PET/CT studies, it is important to understand the repeatability of quantitative PET metrics in order assess change tracer uptake. The present studies were performed precision as function system, reconstruction protocol, analysis method, scan duration (or image noise), repositioning field view.Multiple (repeated) scans have been using a NEMA quality (IQ) phantom 3D Hoffman filled with 18 F solutions on two systems. Studies without randomly (< 2 cm) all...
The aim of this paper is to describe a public, open-access, computed tomography (CT) phantom image set acquired at three centers and collected especially for radiomics reproducibility research. dataset useful test radiomic features with respect various parameters, such as acquisition settings, scanners, reconstruction algorithms.Three phantoms were scanned in independent institutions. Images the following acquired: Catphan 700 COPDGene Phantom II (Phantom Laboratory, Greenwich, NY, USA),...
Abstract Background Positron emission tomography (PET) is routinely used for cancer staging and treatment follow-up. Metabolic active tumor volume (MATV) as well total MATV (TMATV—including primary tumor, lymph nodes metastasis) and/or lesion glycolysis derived from PET images have been identified prognostic factor or the evaluation of efficacy in patients. To this end, a segmentation approach with high precision repeatability important. However, implementation repeatable accurate algorithm...
Background PET-based tumor delineation is an error prone and labor intensive part of image analysis. Especially for patients with advanced disease showing bulky FDG load, segmentations are challenging. Reducing the amount user-interaction in segmentation might help to facilitate tasks especially when labeling complex tumors. Therefore, this study reports on workflows/strategies that may reduce inter-observer variability large tumors shapes different levels user-interaction. Methods Twenty...
Positron-emission tomography (PET) simulators are frequently used for development and performance evaluation of segmentation methods or quantitative uptake metrics. To date, most PET simulation tools based on Monte Carlo simulations, which computationally demanding. Other analytical lack the implementation time flight (TOF) resolution modelling (RM). In this study, a fast easy-to-use simulation-reconstruction package, SiMulAtion ReconsTruction (SMART)-PET, is developed validated, includes...
Background Radiomics refers to the extraction of a large number image biomarker describing tumor phenotype displayed in medical image. Extracted from positron emission tomography (PET) images, radiomics showed diagnostic and prognostic value for several cancer types. However, radiomic features are nonreproducible or highly correlated with conventional PET metrics. Moreover, used clinic should yield relevant information about texture. In this study, we propose framework identify technical...
Abstract Background Machine learning studies require a large number of images often obtained on different PET scanners. When merging these images, the use harmonized following EARL-standards is essential. However, when including retrospective EARL accreditation might not have been in place. The aim this study was to develop convolutional neural network (CNN) that can identify retrospectively if an image compliant and it meeting older or newer EARL-standards. Materials methods 96 acquired...
Low photon count in 89Zr-Immuno-PET results images with a low signal-to-noise ratio (SNR). Since PET radiomics are sensitive to noise, this study focuses on the impact of noise radiomic features from clinical images. We hypothesise that derived have: (1) noise-induced variability affecting their precision and (2) bias accuracy. This aims identify those not or only minimally affected by terms Count-split patient scans previous studies three different 89Zr-labelled monoclonal antibodies were...
Objective Magnetic resonance imaging is the standard modality to assess articular cartilage. As surrogate of degenerative joint disease, cartilage thickness commonly quantified after tissue segmentation. In lack a method, this study systematically compared five methods for automatic measurements across knee and as function region sub-region: 3D mesh normals (3D-MN), nearest neighbors (3D-NN), ray tracing (3D-RT), 2D centerline (2D-CN), surface (2D-SN). Design Based on manually segmented...