XeMRI to CT lung image registration enhanced with personalized 4DCT-derived motion model

Abstract

This paper presents a novel method for multi-modal lung image registration constrained by a motion model derived from lung 4DCT. The motion model is estimated based on the results of intra-patient image registration using Principal Component Analysis. The approach with a prior motion model is particularly important for regions where there is not enough information to reliably drive the registration process, as in the case of hyperpolarized Xenon MRI and proton density MRI to CT registration. Simultaneously, the method addresses local variations between images in the supervoxel-based motion model parameters optimization step. We compare our results in terms of the plausibility of the estimated deformations and correlation coefficient with 4DCT-based estimated ventilation maps using state-of-the-art multi-modal image registration methods. Our method achieves higher average correlation scores, showing that the application of Principal Component Analysis-based motion model in the deformable registration, helps to drive the registration for the regions of the lungs with insufficient amount of information.