Room: Exhibit Hall | Forum 6
Purpose: To measure the image distortion, signal to noise, uniformity, modulation transfer function (MTF) and isocenter of a Viewray MR-Linac, and determine stability over time.
Methods: A large, 35x27x21 cm oblate cylinder shaped phantom (MagPhan from The Phantom Laboratory, Salem NY) was scanned approximately weekly then monthly on our MR-Linac system (19 scans). These images were analyzed to quantify image distortion (maximum distortion magnitude within 200 and 300 mm diameter spheres), signal to noise ratio (SNR), uniformity (mean signal in a 200 and 300 mm diameter spheres) and MTF (edge spread and point spread widths). By aligning the center of the phantom well with the external lasers, the offset of these lasers and couch translation with the imaging isocenter can be estimated. Future scans will be included.
Results: All imaging tests observed an improvement in quality in the last two months due to removal of a radiofrequency noise source. Before this, maximum distortion magnitude remained steady at an average value (standard deviation) of 0.59(0.16) and 1.12(0.14) mm for the 200 and 300 mm spheres. SNR decreased slowly from 49.6 to 42.5. The mean signal in the inner and outer spheres remained constant around 297(9) and 422(18) respectively until the last two months. The average edge and point spread widths were 2.0(0.12) and 2.3(0.11) mm respectively. The offset combined phantom setup, couch translation and laser errors. It is difficult to extract errors in the lasers from this data which oscillated between Â±1.5 mm in the anterior and lateral directions. In the longitudinal, the offset was within 0-1.4 mm.
Conclusion: MR-Linac image quality was found to be stable. Changes in quality were attributable to fixes implemented. There was no deterioration of quality yet observed. The phantom allowed a comprehensive analysis of the image characteristics.