The purpose of this paper is to describe a practical approach to commissioning and quality assurance (QA) of a dedicated wide-bore 3 Tesla (3T) magnetic resonance imaging (MRI) scanner for radiotherapy planning. Methods: A comprehensive commissioning protocol focusing on radiotherapy (RT) specific requirements was developed and performed. RT specific tests included: uniformity characteristics of radio-frequency (RF) coil, couch top attenuation, geometric distortion, laser and couch movement and an end-to-end radiotherapy treatment planning test. General tests for overall system performance and safety measurements were also performed. Results: The use of pre-scan based intensity correction increased the uniformity from 61.7% to 97% (body flexible coil), from 50% to 90% (large flexible coil) and from 51% to 98% (small flexible coil). RT flat top couch decreased signal-to-noise ratio (SNR) by an average of 42%. The mean and maximum geometric distortion was found to be 1.25 mm and 4.08 mm for three dimensional (3D) corrected image acquisition, 2.07 mm and 7.88 mm for two dimensional (2D) corrected image acquisition over 500 mm × 375 mm × 252 mm field of view (FOV). The accuracy of the laser and couch movement was less than ±1 mm. The standard deviation of registration parameters for the end-to-end test was less than 0.41 mm. An on-going QA program was developed to monitor the system's performance. Conclusion: A number of RT specific tests have been described for commissioning and subsequent performance monitoring of a dedicated MRI simulator (MRI-Sim). These tests have been important in establishing and maintaining its operation for RT planning.

The exit detector tool has been demonstrated to be faster for performing the DQA with equivalent sensitivity for detecting MLC LOT errors relative to a conventional phantom-based QA method. In addition, comprehensive MLC performance evaluation and features of reconstructed dose provide additional insight into understanding DQA failures and the clinical relevance of DQA results.

The developed software has been shown to accurately autosegment the seed positions in kV planar images except for two 20° arcs where seeds are obscured by anatomical structures. The isocenter trajectories determined by the system, based on the monoscopic images, provide useful information for monitoring the prostate position. The developed system has potential application for monitoring prostate position during treatment delivery in linear accelerator based SBRT.

No abstract