Click here to

Session: Image Registration [Return to Session]

A Novel Edge Gradient Distance Metric for Automated Evaluation of Deformable Image Registration Quality

Y Xu1*, J Ford1, J Williamson3, N Dogan1, (1) University of Miami, Miami, FL, (2) Washington University, Richmond, VA


MO-H345-IePD-F1-3 (Monday, 7/11/2022) 3:45 PM - 4:15 PM [Eastern Time (GMT-4)]

Exhibit Hall | Forum 1

Purpose: To develop and test a new registration quality metric, based on the distance between image edges, for automated evaluation of DIR algorithms.

Methods: A 3D Canny filter is used to create binary gradient images from input images to be compared. A small subregion of one binary image is translated relative to the other. The translational distance maximizing overlap of edges in the subregion is the local edge gradient distance to agreement (EGDTA); repeating over all subregions provides an EGDTA map. The method was tested on phantom images and pelvic CT images, by applying known displacement vector fields (DVFs) as residual DIR error. The method was then applied to test two DIR algorithms (SICLE and Demons) for pelvic CTs from five prostate patients. Three variants were used in the SICLE algorithm: Grayscale-driven (G), Contour-driven (C), and Grayscale+Contour-driven (G+C). C uses manually drawn bladder, prostate and rectum contours on both images to drive the registration. For each patient, a planning CT was registered to three repeat CTs using the above DIR algorithms, resulting in 60 registrations. Mean EGDTA values in concentric ring regions of interest close to and far away from the contoured organs were compared.

Results: Voxel-by-voxel comparison of EGDTA maps with imposed DVF deformations on phantom and CT images showed good agreement. In comparison of the three variants of SICLE: C had lower EGDTA values (better registration quality) close to the pelvic organs, while G showed much better performance in the regions distant from the organs compared to C; and G+C registration exhibited the lowest or comparable EGDTA value among three variant. Demons achieved the lowest EGDTA values.

Conclusion: The EGDTA metric provides the expected dependence on DIR type and ROI location for pelvic CT to CT registration and shows potential as an automated means of comparing DIR algorithms.


Registration, Quality Assurance


IM/TH- Image Registration: CT

Contact Email