Low dose CT reconstruction for online adaptive proton therapy

External radiotherapy is one of the three major treatment options for treating cancer (with surgery and chemotherapy). However, in contrast to photon therapy, even small uncertainties on the range of the protons may have potentially dramatic consequences for the patient. To control the range, it is critical to ensure that the atomic composition and the densities of the tissues are at all times as predicted during treatment planning. Therefore, a 3D or 4D CT-scan should be acquired in each treatment session, such that the treatment plan can be adapted to the exact anatomy and position of the patient. As a result, these patients may receive a high number of CT-scans.

To reduce the radiation burden associated with these CT-scans, the scan dose should be minimized. The aim of this project is to develop dedicated statistical reconstruction algorithms, which can compute sufficiently accurate CT-images from low dose CT-scans. This will be done by optimizing the reconstruction parameters for proton therapy dose calculation, which poses requirements very different from those of diagnostic CT. In addition, we will investigate further dose reduction by using prior CT images of the same patient, as suggested previously by other groups.