Radiomics & modeling (5 projects)

Radiomics & modeling (5 projects)

The ability to predict the effects of radiation dose in various tissues is essential to optimize the treatment of individual patients. This includes the optimal selection patients for proton therapy. Quantitative prediction is, however, difficult as the effects of radiation are often complex. Knowledge in this field comes from radiobiological experiments and from the clinical observations of patients. From these data, models are developed to predict tumor control probabilities (TCP models) and normal tissue complication probabilities (NTCP models). Also, the relationship with quality of life is assessed.

To improve the prediction performance of these models we have various projects that aim at aiming to improving data collection, enhancing the statistical learning methods, and finding better predictors.

Image biomarkers to improve prediction of side effects

Image biomarkers to improve prediction of treatment outcome

Deep learning and radiomics

Phenomenological modeling

The CITOR project