Date Approved

8-3-2025

Graduate Degree Type

Project

Degree Name

Medical Dosimetry (M.S.)

Degree Program

School of Interdisciplinary Health

First Advisor

George Spencer Arnould

Academic Year

2024/2025

Abstract

Purpose:

This study evaluated whether Philips MRCAT-generated synthetic CT (sCT) can serve as a clinically viable substitute for traditional CT (CT) simulation in proton therapy planning for prostate radiotherapy using a double-scatter technique.

Methods:

Ten prostate cancer patients treated with proton therapy between 2024 and 2025 were retrospectively analyzed. Each patient underwent both CT and MRI in the treatment position. MRCAT sCT images were generated using atlas-based and segmentation algorithms. The original proton treatment plan was transferred onto the MRCAT dataset. Dosimetric parameters, target coverage, organ-at-risk (OAR) doses, and Hounsfield Unit (HU) values were compared. The sign(M) test was used to assess statistical differences.

Results:

Most dose-volume histogram (DVH) comparisons between CT and MRCAT showed no significant differences (p > .05) in key structures, including the prostate, bladder, and rectum. Significant differences (p < .05) were observed in the bladder D15, the bowel region, and the left femoral head, although these did not affect plan quality or target coverage. HU comparisons showed strong agreement, with minor differences in the seminal vesicles and femoral heads that did not translate into clinically relevant dosimetric deviations.

Conclusion:

MRCAT-generated sCT demonstrates comparable dosimetric accuracy and image fidelity to CT, supporting its use in MRI-only workflows for double-scatter proton therapy. These findings validate MRCAT as a practical alternative to CT simulation, with the potential to streamline planning and reduce registration uncertainty.

Download

Included in

Oncology Commons

Share

COinS