A Team Approach to Commissioning Online MRI-Guided Adaptive Radiotherapy

Abstract

Magnetic resonance imaging guided adaptive radiation therapy (MRgART) has led to a major paradigm shift resulting in personalized daily patient treatment. Implementing a online MRgART program requires commissioning the adaptive process. This process is performed primarily by medical physicists. In this work, we propose an efficient team approach to the commissioning process. This approach involves dividing different aspects of the verification process among all individuals involved in the adaptive workflow including: qualified medical physicists (QMP), clinical medical dosimetrists (CMD), radiation therapists (RT), and radiation oncologists (RO). Phantom- and volunteer-based commissioning approaches were performed. The phantom based approach is not discussed here. We have implemented a team approach to verifying the ART workflow using volunteers. The commissioning process was divided into the verification of: 1) the treatment planning system (TPS), and 2) the treatment delivery system (TDS). MRI simulations were performed by the RT and images were imported into the TPS by a CMD. Target structures were drawn by a RO. Organ at risk (OAR) structures and structures specific to the ART workflow were drawn by a CMD; including density override and optimization structures. Boolean operations were added to minimize contouring time on the machine. Clinically acceptable plans were then generated by the CMD and reviewed by the QMP; this included verifying contours, Boolean operations, correct electron density override assignment, and appropriate settings. The RO then reviewed the plan quality and approved for delivery. Patient specific QA was then performed by the QMP. The second part was verifying the TDS by simulating a patient treatment. For each volunteer; set-up, image acquisition and registration were performed by the RT and reviewed by the RO. A CMD edited OAR structures followed by RO review. The QMP applied density overrides and Boolean operations who then re-calculated dose on the daily images. The RO reviewed the plan and made a decision on whether to re-optimize. The QMP performed the optimization and a secondary check using a Monte Carlo based algorithm. Volunteers were imaged to commission the adaptive workflow over a week. Daily MR images were acquired and image segmentation was performed. The volunteer approach for commissioning took 2.5 times longer than the phantom based approach. However, the time for commissioning was halved using volunteers with a team based approach. Team members gained more confidence and were better prepared to start the MRgART program. Efficiency was increased when a team approach to commission MRgART program was adapted. This approach served as a proactive training for individuals involved in the adaptive workflow. Patients treated during the initial implementation phase of the adaptive program benefit from this approach due to the confidence and deep understanding of their radiation therapy team.