Post-operative radiotherapy is standard care for patients with brain metastases undergoing neurosurgery. Current image-guided radiotherapy requires sequential acquisition of CT and MRI scans used for dose planning, and manual delineation of the tumor and surrounding organs at risk.
The novel MRI-linac (MRL), a hybrid system consisting of an MRI and linear accelerator for radiation treatment, developed in our hospital, allows for online MRI-guidance during treatment delivery. In this MR-only workflow, the acquisition of a CT scan is omitted and replaced by a synthetic CT constructed from an MR image. Clinical implementation of such a workflow requires near real-time automatic segmentation of the tumor and of the organs at risk, as well as very fast synthesis of a pseudo-CT scan for real-time dose calculation and recalculation prior to every radiotherapy treatment.
As the successful candidate for this PhD position, you will develop deep learning techniques for segmentation of brain tumors as well as brain and other tissues that should be spared (organs at risk) using MR images. Moreover, you will develop algorithms to synthesize a CT from an MRI scan allowing MR-only radiotherapy treatments. Finally, you will implement real-time automatic dose calculations and recalculations into the workflow.
You are an outstanding candidate with an M.Sc. degree in biomedical engineering, artificial intelligence or comparable training, and a keen interest in medical image analysis and medical physics. You have a strong scientific background, programming experience and are highly motivated and independent. You are confident about working in an multidisciplinary team of medical specialists and engineers.