A5024256820- Radiomics and Machine Learning in Medical Imaging
- COVID-19 diagnosis using AI
- Medical Imaging Techniques and Applications
- Advanced X-ray and CT Imaging
- Cancer Immunotherapy and Biomarkers
- Artificial Intelligence in Healthcare and Education
- COVID-19 Clinical Research Studies
- AI in cancer detection
- Anomaly Detection Techniques and Applications
- Fault Detection and Control Systems
- Machine Learning in Healthcare
Vrije Universiteit Brussel
2020-2025
IMEC
2020-2025
Faculty (United Kingdom)
2020
Engineering (Italy)
2020
Challenges drive the state-of-the-art of automated medical image analysis. The quantity public training data that they provide can limit performance their solutions. Public access to methodology for these solutions remains absent. This study implements Type Three (T3) challenge format, which allows on private and guarantees reusable methodologies. With T3, organizers train a codebase provided by participants sequestered data. T3 was implemented in STOIC2021 challenge, with goal predicting...
In oncology, 18-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) / computed (CT) is widely used to identify and analyse metabolically-active tumours. The combination of the high sensitivity specificity from 18F-FDG PET resolution CT makes accurate assessment disease status treatment response possible. Since cancer a systemic disease, whole-body imaging interest. Moreover, metabolic tumour burden emerging as promising new biomarker predicting outcome for innovative...
Recent research on COVID-19 suggests that CT imaging provides useful information to assess disease progression and assist diagnosis, in addition help understanding the disease. There is an increasing number of studies propose use deep learning provide fast accurate quantification using chest scans. The main tasks interest are automatic segmentation lung lesions scans confirmed or suspected patients. In this study, we compare twelve algorithms a multi-center dataset, including both...
Our motivating application is a real-world problem: COVID-19 classification from CT imaging, for which we present an explainable Deep Learning approach based on semi-supervised pipeline that employs variational autoencoders to extract efficient feature embedding. We have optimized the architecture of two different networks images: (i) novel conditional autoencoder (CVAE) with specific integrates class labels inside encoder layers and uses side information shared attention encoder, make most...
Recent research on COVID-19 suggests that CT imaging provides useful information to assess disease progression and assist diagnosis, in addition help understanding the disease. There is an increasing number of studies propose use deep learning provide fast accurate quantification using chest scans. The main tasks interest are automatic segmentation lung lesions scans confirmed or suspected patients. In this study, we compare twelve algorithms a multi-center dataset, including both...
Background: Antibodies that inhibit the programmed cell death protein 1 (PD-1) receptor offer a significant survival benefit, potentially cure (i.e., durable disease-free following treatment discontinuation), substantial proportion of patients with advanced melanoma. Most however fail to respond such or acquire resistance. Previously, we reported baseline total metabolic tumour volume (TMTV) determined by whole-body [18F]FDG PET/CT was independently correlated and able predict futility...
Automated lesion segmentation is essential to provide fast, reproducible tumor load estimates. Though deep learning methods have achieved unprecedented results in this field, they are often difficult interpret, hampering their potential integration the clinic. An interpretable approach proposed for segmenting melanoma lesions on whole-body fluorine-18 fluorodeoxyglucose ([<sup>18</sup>F]FDG) positron emission tomography (PET) / computed (CT). This consists of an automated PET thresholding...
Our motivating application is a real-world problem: COVID-19 classification from CT imaging, for which we present an explainable Deep Learning approach based on semi-supervised pipeline that employs variational autoencoders to extract efficient feature embedding. We have optimized the architecture of two different networks images: (i) novel conditional autoencoder (CVAE) with specific integrates class labels inside encoder layers and uses side information shared attention encoder, make most...
Challenges drive the state-of-the-art of automated medical image analysis. The quantity public training data that they provide can limit performance their solutions. Public access to methodology for these solutions remains absent. This study implements Type Three (T3) challenge format, which allows on private and guarantees reusable methodologies. With T3, organizers train a codebase provided by participants sequestered data. T3 was implemented in STOIC2021 challenge, with goal predicting...
PET/CT is widely used in oncology. Yet the identification of lesions, as described by PET response criteria solid tumors (PERCIST), still relies on manual a volume interest (VOI), typically liver, for determining optimal threshold. The process requires expert knowledge and prone to errors inter-observer variability. A fully automated procedure application PERCIST whole- body images proposed. method localization liver whole-body CT using dense V-net trained large field-of-view images. Inside...