Prospective Clinical Feasibility Study for MRI-Only Brain Radiotherapy.

Area of Science:

• Radiotherapy
• Medical Imaging
• Oncology

Background:

• MRI-only radiotherapy (RT) offers reduced geometric uncertainty and streamlined planning by minimizing image registration between MRI and CT data.
• While synthetic CT (sCT) methods are validated for head regions, clinical implementation for brain tumors remains unreported.
• This study builds upon prior sCT validation to explore MRI-only brain RT feasibility.

Purpose of the Study:

• To investigate the clinical feasibility of MRI-only radiotherapy for brain tumors.
• To evaluate dosimetry and patient setup endpoints in an MRI-only workflow.
• To assess the robustness of synthetic CT (sCT) images generated by deep learning software.

Main Methods:

• Twenty-one glioma patients were enrolled in a prospective clinical feasibility study.
• Synthetic CT (sCT) images were generated from MRI Dixon images using CE-marked deep learning software.
• Treatment plans were created using MRI-delineated structures and sCT for dose calculations, with CT used for sCT quality assurance.

Main Results:

• Ninety-five percent (20/21) of patients were successfully treated with the MRI-only workflow.
• Dose deviations in the target were within ±1% for all patients.
• Gamma pass rates (2%, 2 mm) exceeded 99% in retrospective analysis, and patient positioning was within ±1 mm.

Conclusions:

• The clinical study successfully implemented MRI-only brain radiotherapy using prospective and retrospective analyses.
• CE-marked deep learning-based software generated clinically robust synthetic CT images.
• The MRI-only workflow demonstrated efficacy in dosimetry and patient positioning for brain tumor radiotherapy.