Assessment of image quality and scatter and leakage radiation of an integrated MR-LINAC system.

Purpose: To assess the image quality, scatter, and leakage radiation of an integrated magnetic resonance linear accelerator (MR-LINAC or MRL) system.

Methods: A large American College of Radiology (ACR) magnetic resonance imaging (MRI) accreditation phantom was used to evaluate the MRI capabilities of the integrated MRL system compared with those of other diagnostic MRI systems. Multiple sets of T1 and T2/PD images were acquired with the linear accelerator positioned at various angles and with the radiation beam on and off. Images also were acquired on three different occasions over a period of about 12 months. Scatter and leakage radiation were measured with a large (150 cm ) ion chamber recalibrated for MV energy. For scatter measurements, a 25-cm stack of solid-water materials was placed at the isocenter on the patient couch to simulate a patient. The head leakage was measured at 1 m from the linear accelerator head in directions determined to produce the maximum leakage. All measurements were repeated with the magnetic field turned off to study the effects of the magnetic field.

Results: The geometric distortion, slice thickness accuracy, image uniformity, ghosting ratio, and high-contrast detectability were comparable to other 1.5 T diagnostic MRI scanners. No observable changes in image quality and no appreciable differences were found between radiation beam-on and beam-off images. The measured leakage and scattered radiation changed by less than 5% when the magnetic field was on compared to measurements with the field off. The beam stopper leakage was approximately 0.3 R/1000 MU, and because there was no direct beam imparted on the walls, a vault designed for a modern-day LINAC should have enough required radiation shielding to house the MRL.

Conclusions: The image quality generated by the MRI system of the integrated MRL was similar to that of a diagnostic MRI scanner. Interference from the MV radiation was minimal, and there was no measurable difference in image quality with the beam on and off. Scatter radiation and leakage radiation of the MRL system were within the expected range of a comparable MV-LINAC.