Coregistration and analysis of multi-modal images obtained in different geometries

摘要

A method for coregistration of multi-modal images obtained in different geometries includes acquiring multi-modal image data, wherein the multi-model image data includes image data of a first modality and image data of a second modality, wherein the image data of the respective modalities have different geometries, defining a volume of interest in the multi-modal image data, segmenting the image data of the first modality and incorporating segmentation data of the first modality into a reconstruction of the second modality, and applying a registration of the second modality image data to the first modality image data according to a similarity measure through the volume of interest, wherein an output of the registration comprises superimposed multi-modal image data.

主权项

1. A non-transitory computer readable medium embodying instructions executable by a processor to perform a method for coregistration of multi-modal images, the method steps comprising:
acquiring multi-modal volumetric image data, wherein the multi-modal volumetric image data includes volumetric image data of a first modality and volumetric image data of a second modality, wherein a same tissue captured in the image data of the first and second modalities has different geometries; defining a volume of interest in the multi-modal volumetric image data, wherein defining the volume of interest comprises:
determining three mutually orthogonal signatures for the first and the second modalities; andregistering, iteratively, each of the three mutually orthogonal signatures of one of the first and the second modalities to the three mutually orthogonal signatures of the other modality, wherein registered signatures are combined into a 9-dimensional parameter space corresponding to the volume of interest; segmenting the volumetric image data of the first modality and incorporating priors of segmentation data of the first modality into a reconstruction of the volumetric image data of the second modality; and applying a registration to transform the second modality volumetric image data to align with the first modality volumetric image data according to a similarity measure through the 9-dimensional parameter space corresponding to the volume of interest, wherein an output of the registration comprises superimposed multi-modal volumetric image data wherein the geometries of the same tissue are aligned.