Nonlinear Registration of Brain Magnetic Resonance Images with Cross Constraints of Intensity and Structure

作     者：Han Zhou HongtaoXu Xinyue Chang Wei Zhang Heng Dong 

作者机构：Digital Fujian Research Institute of Big Data forAgriculture and ForestryFujian Agriculture and Forestry UniversityFuzhou350002China College of Computer and Information ScienceFujian Agriculture and Forestry UniversityFuzhou350002China 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2024年第79卷第5期

页      面：2295-2313页

核心收录：

学科分类：08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

基　　金：National Natural Science Foundation of China(Grant Nos.62171130,62172197,61972093) the Natural Science Foundation of Fujian Province(Grant Nos.2020J01573,2022J01131257,2022J01607) Fujian University Industry University Research Joint Innovation Project(No.2022H6006) in part by the Fund of Cloud Computing and BigData for SmartAgriculture(GrantNo.117-612014063) NationalNatural Science Foundation of China(Grant No.62301160) Nature Science Foundation of Fujian Province(Grant No.2022J01607) 

主　　题：Medical image registration cross constraint semantic consistency directional consistency dual-channel 

摘      要：Many deep learning-based registration methods rely on a single-stream encoder-decoder network for computing deformation fields between 3D ***,these methods often lack constraint information and overlook semantic consistency,limiting their *** address these issues,we present a novel approach for medical image registration called theDual-VoxelMorph,featuring a dual-channel cross-constraint *** innovative network utilizes both intensity and segmentation images,which share identical semantic information and feature *** encoder-decoder structures calculate deformation fields for intensity and segmentation images,as generated by the dual-channel cross-constraint *** design facilitates bidirectional communication between grayscale and segmentation information,enabling the model to better learn the corresponding grayscale and segmentation details of the same anatomical *** ensure semantic and directional consistency,we introduce constraints and apply the cosine similarity function to enhance semantic *** on four public datasets demonstrates superior performance compared to the baselinemethod,achieving Dice scores of 79.9%,64.5%,69.9%,and 63.5%for OASIS-1,OASIS-3,LPBA40,and ADNI,respectively.