Maastricht Radiation Oncology Foundation
Dr. Tanslaan 12 6229ET Maasticht The Netherlands
Prof. dr. ir. A.L.A.J. Dekker
0031 (0)88 4455666
MAASTRO (Maastricht Radiation Oncology, see www.maastro.nl) is an integrated institute combining radiation oncology care with cancer research.
The research division MAASTRO Knowledge Engineering (MAASTRO KE) performs data science research and services in the following themes
- Develop and support for global FAIR data sharing infrastructure in which scientists can learn from routine clinical and imaging data without the need for data to leave the hospital (distributed learning).
- Radiomics research in which pipelines are developed for high throughput, comprehensive quantification of tumour phenotypes from clinical routine images.
- Research into cancer outcome prediction models using machine learning and host these publicly and apply these in clinical routine as decision support systems.
- Biomedical & health
- Medical imaging
- FAIR data
- Machine Learning
- Semantic Web technology
- Population “big data” research
- Lung, rectum and head&neck cancer predictive models
- Federated data sources
- Machine learning and privacy preserving data mining in clinical radiation oncology using routine and clinical trial data
- Data sharing using relational and semantic web technology for clinical radiation oncology and cancer trials
Expertise and Track Record
MAASTRO KE has developed a unique distributed learning infrastructure. In this approach scientist can execute machine learning algorithms on clinical datasets from around the globe, without the need for data to leave the individual hospital. This privacy-by-design infrastructure offers the possibility for orders of magnitude data to be made FAIR and available for learning.
- TraIT: MAASTRO KE is the imaging lead and is responsible for back-end services (http://www.ctmm-trait.nl).
- euroCAT/duCAT: MAATRO KE is the initiator of and leading a global distributed learning infrastructure for cancer (http://www.eurocat.info)
- Health-RI: MAASTRO KE is one of the original initiators of Health-RI, the KNAW roadmap initiative for a research infrastructure for personalized medicine and health research (http://www.health-ri.nl).
- ELIXIR-NL: MAASTOR KE is one of the original initiators of the ELIXIR-NL NWO roadmap initiative for a FAIR data infrastructure for the life sciences.
- Personal Health Train: MAASTRO KE is on of the initiator of the DTL Personal Health train project (http://www.personalhealthtrain.nl).
- Strategy: MAASTRO KE leads the STW Strategy program in which an integrated data and compute infrastructure for medical imaging and analysis is developed.
- FHIR: MAASTRO KE has developed HL7 FHIR based data exchange pipelines to automatically transfer
de-identified clinical and derived imaging data to registries such as DICA and IKNL.
- PredictCancer: MAASTRO KE maintains a website hosting prediction models for cancer (http://www.predictcancer.org)
All the projects mentioned above are cross technology projects including the integration of datasets from the clinical, medical imaging, treatment and follow-up.
- International data-sharing for radiotherapy research: an open-source based infrastructure for multicentric clinical data mining. (https://www.ncbi.nlm.nih.gov/pubmed/24309199)
- BMIA - een nationaal biomedisch imagingarchief voor onderzoek. (http://www.ctmm-trait.nl/news/NVvRMR20.2Dekker.pdf)
- Distributed learning: Developing a predictive model based on data from multiple hospitals without data leaving the hospital - A real life proof of concept. (https://www.ncbi.nlm.nih.gov/pubmed/28029405)
- Informatics methods to enable sharing of quantitative imaging research data. (https://www.ncbi.nlm.nih.gov/pubmed/22770688)
- TraIT (see above) has many international users which MAASTRO KE supports.
- euroCAT/duCAT (see above) is an international distributed learning network for cancer including 25 partners in Asia, Africa, Europe, America and Australia.
- BIONIC: a NWO funded India-Netherlands distributed learning network for medical imaging (Radiomics).
- QIN: a NIH funded USA-China-Netherlands network for medical imaging (Radiomics).
- Postdoctoral researchers (4)
- Software engineers (4)
- PhD students (8)
Software Quality System (based on ISO/IEC 12207 but not certified)