Evaluation of performance of pelvic CT-MR deformable image registration using two software programs

Abstract We assessed the accuracy of deformable image registration (DIR) between CT and MR images using an open-source software (Elastix, from Utrecht Medical Center) a commercial (Velocity AI Ver. 3.2.0 Varian Systems, Palo Alto, CA, USA) software. Five male patients’ pelvic regions were studied publicly available CT, T1-weighted (T1w) MR, T2-weighted (T2w) images. In cost function Elastix, we used six DIR parameter settings with different regularization weights (Elastix0, Elastix0.01, Elastix0.1, Elastix1, Elastix10, Elastix100). Corrected Deformable algorithm for Velocity AI. The Dice similarity coefficient (DSC) mean distance to agreement (MDA) prostate, bladder, rectum left right femoral heads evaluate accuracy. Except most algorithms produced good DSC MDA results all organs. our study, DSCs bladder ranged 0.75 0.88 (CT-T1w) 0.72 0.76 (CT-T2w). Similarly, ranges 2.4 4.9 mm (CT-T1w), 4.6 5.3 For CT-T1w was outperformed CT-T2w both MDAs at Elastix0, Elastix0.1. case AI, no significant differences in organs observed. This implied that might differ depending on sequence used.