联合显著图强化形变配准网络  

Deformable Registration Network Reinforced by Joint Saliency Map

作　　者：傅泽山 秦斌杰[1] FU Zeshan;QIN Binjie(School of Biomedical Engineering,Shanghai Jiao Tong University,Shanghai,200240)

机构地区：[1]上海交通大学生物医学工程学院

出　　处：《中国医疗器械杂志》2019年第6期397-400,409,共5页Chinese Journal of Medical Instrumentation

基　　金：国家自然科学基金(61271320);上海交通大学转化医学交叉重点项目(ZH2018ZDA19)

摘　　要：配准图像中存在的结构对应性缺失与局部复杂大形变给非刚性图像配准准确寻找一一映射形变变换带来了极大挑战。传统配准方法以及基于深度学习配准方法都不能处理好此类图像配准异常难题。基于全局到局部的递进深度网络策略,该文提出了配准图像联合显著结构上下文信息增强的无监督式深度配准网络。其中,全局到局部的卷积网络通过将待配准图像输入到结果形变场输出的复杂映射分解为两个更易求解的全局映射与局部映射网络,同时结合配准图像联合显著结构上下文信息双向加强网络的学习训练,实现了精确、鲁棒、高效的联合显著图强化形变配准网络,有效地解决了既存在结构对应性缺失又存在局部复杂大形变的图像配准难题。Image outliers such as missing correspondences and large local deformations break the one-to-one pixelwise mapping between target image and moving image to be registered.Both traditional registration methods and deep-learning based deformable image registration methods fail to tackle this problem.This paper proposed an unsupervised globalto-local deformable registration network reinforced by joint saliency map to accurately,robustly and fast address the problem.The global-to-local network divided the overall learning of a complex mapping of image registration into a simpler global mapping learning and local residual mapping.The joint saliency map of the two images to be registered bidirectionally reinforced the whole network’s forward estimation and back-propagation with uncertainty modeling and context-aware intelligence.The experimental results confirm the proposed method’s performance advantages over the state-of-the-arts registration methods in the challenges image registration with missing correspondences and large local deformations.

关 键 词：非刚性图像配准 对应性缺失 局部大形变 深度学习 联合显著图 联合显著结构 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]