Frontiers in Biomedical Technologies (Dec 2016)
An Automated Non-Rigid Registration Method for Accurate Quantification of Dynamic Contrast Enhanced MR Imaging (DCE-MRI) in Complex Adnexal Masses Employing Residual Complexity Framework
Abstract
Purpose: Quantification of dynamic contrast enhanced (DCE-) MRI of ovarian masses is susceptible to errors caused by motion artifacts and intensity inhomogeneity induced by bias fields. Motion artifacts and bias fields introduce signal intensity variations in the images that must be resolved from intensity changes caused by the passage of contrast agent. Thus, registration of DCE-MRI image sequence is a challenging issue. In this work, we proposed a solution to the misregistration problem of DCE-MR images. Methods: We acquired pre-operative DCE-MR images of 16 patients diagnosed with solid or solid/cystic complex ovarian masses on ultrasound examination (with post-operative histopathological assessment showing 8 benign and 8 malignant cases). Residual complexity (RC) similarity measure was exploited in a non-rigid registration framework, to account for complex intensity variations. The performance of the proposed method was evaluated by computed semi-quantitative parameters, determined in the regions of interest (ROIs) selected on the solid portion of the tumor and the psoas muscle. The results were compared with unregistered data and registered images using mutual information (MI) similarity measure. Results: The registered data using RC similarity measure indicated lower variations in the signal intensity over the time course of contrast agent passage. The derived quantitative parameters showed enhanced separation of benign and malignant tumors using RC registration in comparison with unregistered and MI-registered data. Conclusion: RC registration is a useful tool for correcting the misalignment of DCE-MR image series in the presence of bias field artifact, while it conserves the quantitative information of the contrast enhancement.
