Predicting Tumor Location by Modeling the Deformation of the Breast
Predicting Tumor Location by Modeling the Deformation of the Breast
复制标题
DOI：
10.1109/tbme.2008.925714
复制
发表时间：
2008-10-01
影响因子：
4.6
通讯作者：
Brady, J. Michael
中科院分区：
文献类型：
--
作者：
Pathmanathan, Pras;Gavaghan, David J.;Brady, J. Michael
乳腺癌是西方世界最大的杀手之一，早期诊断对改善预后至关重要。乳房的形状在磁共振（MR）成像（患者俯卧，乳房在重力下下垂）、X射线乳房摄影（乳房强烈压缩）和超声或活检/手术（患者仰卧）的场景之间变化很大，使得图像融合成为一项极其困难的任务。本文关注的是使用有限元法和非线性弹性，建立一个三维的，病人特定的，解剖学上准确的乳房模型。该模型由MR图像构建，可以变形以模拟乳房形状，并在乳房X线摄影或活检/手术期间预测肿瘤位置。标准弹性问题的两个扩展需要解决：逆弹性问题（由于最初只知道变形，应力状态）和模型压缩的接触问题。将该模型用于头尾向内外斜位乳腺摄影图像的匹配，并进行了数值实验。
Breast cancer is one of the biggest killers in the western world, and early diagnosis is essential for improved prognosis. The shape of the breast varies hugely between the scenarios of magnetic resonance (MR) imaging (patient lies prone, breast hanging down under gravity), X-ray mammography (breast strongly compressed) and ultrasound or biopsy/surgery (patient lies supine), rendering image fusion an extremely difficult task. This paper is concerned with the use of the finite-element method and nonlinear elasticity to build a 3-D, patient-specific, anatomically accurate model of the breast. The model is constructed from MR images and can be deformed to simulate breast shape and predict tumor location during mammography or biopsy/surgery. Two extensions of the standard elasticity problem need to be solved: an inverse elasticity problem (arising from the fact that only a deformed, stressed, state is known initially), and the contact problem of modeling compression. The model is used for craniocaudal mediolateral oblique mammographic image matching, and a number of numerical experiments are performed.