Room: AAPM ePoster Library
Purpose: To describe a low-cost method for constructing an MR imaging phantom for geometric distortion measurements. MRI has become an essential part of multi-modality imaging/treatment planning in radiation oncology and many centers do not have access to specialized MR phantoms for verifying the performance of their systems under clinical conditions.
Methods: An MRI distortion phantom was constructed out of 0.68 caliber paintballs (~17 mm dia., gelatin shell /polyethylene glycol-based paint) embedded in polyurethane foam (2-part reacting liquid mix). Phantom exterior consisted of a plastic food canister roughly the size of a human head. Foaming chemicals were mixed in the canister and the paintballs were added during the foaming/curing process resulting in their suspension in a rigid substrate. MR and CT scans were performed using thin slices, and image registration was used to visualize MRI distortion relative to CT. For the purposes of this work, the CT images were assumed to be ground truth. Representative paintballs in CT and MRI were contoured and their center of mass coordinates were compared to determine a quantitative measurement of geometric distortion. For initial experiments, the phantom was imaged using CT before and after MRI to verify that the phantom’s internal structure had not changed.
Results: Phantom provided adequate quality CT and MRI images (T1 and T2 weighted). MRI geometric distortion was measured to be approximately 1-2 mm in magnitude depending on imaging protocols, type of coil, and phantom position relative to isocenter.
Conclusion: The simple, low-cost phantom described in this work successfully provided a method for verification of geometric distortion in an MR imaging system. The mechanical stability of the phantom over time may be of concern due to potential degradation of foam or paintballs. Subsequent CT scans will be used to monitor stability over time.