Room: Exhibit Hall | Forum 2
Purpose: Magnetic resonance imaging (MRI) presents advantages when used in conjunction with radiation therapy (RT), including excellent soft-tissue contrast and the ability to provide structural and functional information about the tumor and normal structures. With the increasing use of MR-guided RT (MR-gRT), it becomes important to understand and explore imaging quality assurance with more accuracy. The purpose of this work is to evaluate a new 3D distortion and quality control (QC) phantom using diagnostic MR and design a QC protocol for MR-gRT.
Methods: The large field-of-view (FOV) (35x27x21 cm) MagphanRT phantom (The Phantom Laboratory, Inc., Salem, NY), which provides a means for assessing three-dimensional geometric distortion as well as resolution, slice thickness, signal-to-noise ratio (SNR), uniformity, and laser alignment, was evaluated on a 1.5T GE HDxt scanner (GE Healthcare, Waukesha, WI). All imaging was performed with a 3D-FSPGR sequence using the body coil with the following imaging parameters: 40-cm FOV; 1.6-mm in-plane resolution; 3-mm slice thickness; 2.9/7.1-ms TE/TR; 15-degree flip angle. The phantom was repositioned nine times with offsets from 0-2 cm in the lateral and superior-inferior directions and the consistency of various measurement results were evaluated.
Results: Distortion measurements at a set of arbitrarily sampled points within the phantom were robust to phantom translation, with standard deviations of the distortion measurements less than 0.2 mm. The average resolution, defined as the 10-90% transition width of the edge-spread function, was 1.8Â±0.08 mm. The average signal uniformity within a 30-cm diameter (normalized spread) was 0.5Â±0.08 and the average SNR near the phantom center was 24Â±1.1.
Conclusion: A procedure of baselining measurements to establish expected levels of variability leads to a practical QC protocol for monitoring MRI scanners used in MR-gRT planning. Future work includes testing measurement consistency on other imaging protocols as well as other field strengths used in RT.