COMPARISON AND EVALUATION OF RETROSPECTIVE INTERMODALITY BRAIN IMAGE REGISTRATION TECHNIQUES

Purpose: The primary objective of this study is to perform a blinded e valuation of a group of retrospective image registration techniques us ing as a gold standard a prospective, marker-based registration method . To ensure blindedness, all retrospective registrations were performe d by participants who had no knowledge of the gold standard results un til after their results had been submitted. A secondary goal of the pr oject is to evaluate the importance of correcting geometrical distorti on in MR images by comparing the retrospective registration error in t he rectified images, i.e., those that have had the distortion correcti on applied, with that of the same images before rectification. Method: Image volumes of three modalities (CT, MR, and PET) were obtained fro m patients undergoing neurosurgery at Vanderbilt University Medical Ce nter on whom bone-implanted fiducial markers were mounted. These volum es had all traces of the markers removed and were provided via the Int ernet to project collaborators outside Vanderbilt, who then performed retrospective registrations on the volumes, calculating transformation s from CT to MR and/or from PET to MR. These investigators communicate d their transformations again via the Internet to Vanderbilt, where th e accuracy bf each registration was evaluated. In this evaluation, the accuracy is measured at multiple volumes of interest (VOIs), i.e., ar eas in the brain that would commonly be areas of neurological interest . A VOI is defined in the MR image and its centroid c is determined. T hen, the prospective registration is used to obtain the corresponding point c' in CT or PET. To this point, the retrospective registration i s then applied, producing c '' in MR. Statistics are gathered on the t arget registration error (TRE), which is the distance between the orig inal point c and its corresponding point c ''. Results: This article p resents statistics on the TRE calculated for each registration techniq ue in this study and provides a brief description of each technique an d an estimate of both preparation and execution time needed to perform the registration. Conclusion: Our results indicate that retrospective techniques have the potential to produce satisfactory results much of the time, but that visual inspection is necessary to guard against la rge errors.