Prof. dr. André Dekker clinical physicist There is no such thing as a universal tumour. Every person is unique and, similarly, every form of cancer has its specific characteristics, e.g. tumour size, tumour shape, blood perfusion, specific genetic features, risk of metastasis, sensitivity to chemotherapy and radiation therapy, etc. A model is in fact only a mathematical formula based on a gigantic collection of patient and tumour data. These data are collected and stored all over the world, from Maastricht to Sydney. All data together provide a wealth of information. If the data are linked correctly, you can create a smart prediction model. Predictive modelling can indicate, for instance, that radiation therapy will reduce the risk of recurrence to 70 percent, whereas adding chemotherapy will reduce the risk to 40 percent. Therapies may increase the risk of side effects, resulting in a lower quality of life. Of course the model does not indicate which treatment option must be chosen: Patients have a major say in this decision. Also with respect to proton therapy, we want to apply models that can predict the efficacy as accurately as possible. Every year, over 100,000 people in the Netherlands are diagnosed with cancer. However, we can only provide proton therapy to approximately 2,000 of these people. It is therefore essential that we can select the patients in whom a maximum effect of the therapy is expected. Predictive modelling enables us to make this selection. The models will only improve over the next few years, because the more data are included, the more accurate the predictions will be. This is why we collaborate with several knowledge institutes. We do not keep the data to ourselves; the strength of the model is that it allows us to share our knowledge with colleagues all over the world. On the job
Circles_of_innovations_ENG
To see the actual publication please follow the link above