Image Distortions on a Plastic Interstitial Computed Tomography/Magnetic Resonance Brachytherapy Applicator at 3 Tesla Magnetic Resonance Imaging and Their Dosimetric Impact

Abstract

To quantify magnetic resonance imaging (MRI) distortions on a plastic intracavitary/interstitial applicator with plastic needles at a field strength of 3T and to determine the dosimetric impact, using patient data. For 11 cervical cancer patients, our clinical MRI protocol was extended with 3 scans. From the first scan, a multi-echo acquisition, a map of the magnetic field (B0) was calculated and used to quantify the field inhomogeneity. The expected displacements of the applicator were quantified for the clinical sequence using the measured field inhomogeneity and the clinical sequence's bandwidth. The second and third scan were our routine clinical sequence (duration: <5minutes each), acquired consecutively using opposing readout directions. The displacement of the applicator between these scans is approximately twice the displacement due to B0 inhomogeneity. The impact of the displacement on the dose was determined by reconstructing the applicator on both scans. The applicator was then shifted and rotated the same distance as the observed displacement to create a worst-case scenario (ie, twicethe actual displacement due to B0 inhomogeneity). Next, the dose to 98%/90% (D98/D90) of the clinical target volume at high risk, as well as the dose to the most irradiated 2cm3 for bladder and rectum, were calculated for the original plan as well as the shifted plan. For a volume of interest containing the intrauterine device and the ovoids the 95th percentile of the absolute displacement ranged between 0.2 and 0.75mm, over all patients. For all patients, the difference in D98/D90 in the opposing readout scans with the original plan was at most 4.7%/4.3%. For the dose to the most irradiated 2cm3 of bladder/rectum, the difference was at most 6.0%/6.3%. The dosimetric impact of distortions on this plastic applicator with plastic needles is limited. Applicator reconstruction for brachytherapy planning purposes is feasible at 3T MRI.