Data-driven automated brachytherapy treatment planning for challenging cancer diagnoses

Brachytherapy is a modality for radiation therapy in which sources of ionizing radiation are inserted close to or within the tumour.

We study mathematical optimization approaches to anatomy-based inverse treatment planning for intensity modulated brachytherapy. Such planning tools are commonly used for external beams (EBRT), but they are less developed for brachytherapy.

Our research aims at effectively and efficiently automate brachytherapy treatment planning, while also incorporating more aspects than is currently possible.

The ultimate goal is to develop data-driven, evidence-based treatment planning methods that can be used in clinical practice, with improved patient survivability and quality of life.Our main tools to achieve better brachytherapy planning are mathematical optimization models and algorithms. Patient anatomy is given by images (e.g.

MRI).

We will also incorporate data from retrospective studies that evaluate treatment outcomes, as these provide decisive evidence for the importance of various treatment goals.

The degrees of freedom considered are dwell times and catheter placements, and the objectives used are directly related to the clinical evaluation of treatment plans and evidence for the outcome of the treatment, with respect to both effect on the tumour and side effects.

We focus on treatment sites for which brachytherapy is known to be effective, but the treatment planning is challenging. These include gynecological cancer (cervix) and breast cancer.