基于水平集的三维放射治疗CT图像弹性配准

石翔翔唐 涛庞皓文*

基于水平集的三维放射治疗CT图像弹性配准

石翔翔①唐 涛①庞皓文①*

目的:将基于水平集的图像弹性配准应用于放射治疗CT图像中,为精确评估患者肿瘤及危及器官变化过程及其累加剂量提供技术支持.方法:根据水平集进化理论,通过matlab软件编程,分别对宫颈癌患者与鼻咽癌患者放射治疗分次间的两组CT图像实施三维全自动的弹性配准.结果:宫颈癌患者配准前后对比显示,最小均方误差(MSE)减少55.1%,相关系数(CC)提高了5.3%;鼻咽癌患者配准前后对比显示,MSE减少32.1%,CC提高4.6%.结论:从配准前后差分图以及评价参数可见,该理论取得了较好的试验效果,但要真正应用到临床放射治疗中,需进一步探寻更精确的数学算法以模拟人体器官的运动.

水平集;弹性配准;放射治疗

石翔翔,男,(1985- ),硕士研究生,技师.西南医科大学附属医院肿瘤科,从事肿瘤放射治疗工作.

[First-author's address] Department of Oncology, Affiliated Hospital of Southwest Medical University, Luzhou 646000, China.

弹性配准研究是放射治疗中的热点和前沿课题,放射治疗的整个实施过程中肿瘤的退缩以及周围正常组织的变化是影响整个放射治疗剂量精确的关键,精确评估每例患者肿瘤及危及器官变化过程及其累加剂量是放射治疗热点与难点[1-5].弹性配准算法为统计这些变化过程提供了技术支持,也是放射治疗四维模型建立的基础[6-7].目前,许多弹性配准方法已经被用于放射治疗图像中,如基于样条函数的方法、光流法以及危及器官中植入标记点的方法等[8-14].本研究尝试应用Vemuri等[15]提出的水平集理论,通过matlab软件编程对放射治疗分次间CT图像实施三维全自动弹性配准.

1 资料与方法

1.1 一般资料

随机选取西南医科大学附属医院肿瘤科10例宫颈癌放射治疗患者不同分次时的两组CT图像,并与10例鼻咽癌放射治疗患者不同分次时的两组CT图像进行仿真试验验证.

1.2 方法

根据Vemuri等[15]在水平集进化理论的基础上提出的水平集移动理论,假定有两组图像为M(ν)和S(ν),设ν=(x1,x2,x3)为空间3个方向,M(ν)是待配准图像灰度值,S(ν)是参考图像灰度值,给定图像弹性形变的位移场为u(ν),M(ν)与S(ν)之间就存在一个相应的灰度映射关系,该关系可以通过两组灰度图像之间的水平集的映射来实现,可以用一非线性的双曲偏微分方程(partial differential equations PDE)表达.因此,两组图像的弹性配准就可以直观的认为是要将待配准图像的灰度函数的水平集变换成目标参考图像的灰度函数的水平集,即可以认为将M(ν)的水平集沿着法线方向移动,直到其与参考图像S(ν)保持一致.则两组图像的弹性形变的位移场为u(ν)可表达为公式1～3:

则为公式5:

式中σ=2.

1.3 弹性配准步骤

⑴首先初始化位移场,迭代次数n=0,u(ν)0=0.

⑵设n=n+1,根据公式5更新位移场u(ν)n+1,更新浮动图像为Mn(ν+u(ν)n+1).

⑶判断迭代次数大于设定值,则转步骤(4),否则执行步骤(2).

⑷迭代终止,Mn(ν+u(ν)n+1)为最终的配准图像,配准过程中采用多分辨策略.

2 结果

通过Matlab7.6软件,根据上述水平集算法原理,编写三维全自动弹性配准程序,并用此程序对10例宫颈癌放射治疗患者不同分次时的两组CT图像与10例鼻咽癌放射治疗患者不同分次时的两组CT图像进行仿真试验验证.试验平台为Intel Core i5-3210M CPU,2.5 GHz,4 G内存.宫颈癌患者CT图像断层大小均为512X512,层距为2.5 mm,共采集64层;鼻咽癌患者CT图像断层大小均为512X512,层距为2.5 mm,共采集145层.试验参数见表1.

表1 试验评价参数

试验结果如图1、图2所示;宫颈癌患者配准前后对比,均方误差(mean squared error,MSE)减少55.1%,平均相关系数(correlation coefficient,CC)提高了5.3%;鼻咽癌患者配准前后对比,平均MSE减少57.0%,平均CC提高4.6%.

3 讨论

图1 1例基于水平集的宫颈癌放射治疗患者CT图像弹性配准冠状面差分图

图2 1例基于水平集的鼻咽癌放射治疗患者CT图像弹性配准冠状面差分图

在本研究中,基于根据Vemuri等在水平集进化理论的基础上提出的水平集移动理论,该理论是一种迭代近似估计法,通过水平集移动理论计算待配准图形的位移场,通过matlab软件对放射治疗CT图像进行三维全自动弹性配准仿真实验,从配准前后差分图的比较及表1两种算法的评价参数可见,该理论取得了较好的试验效果,但要真正应用到临床放射治疗中,仍显精度不够.

4 结语

为了更精确评估每例患者肿瘤及危及器官变化过程及其累加剂量,本研究期望弹性配准的整个形变过程能更加准确地模拟肿瘤器官的退缩与周围正常组织的运动.通过不同的数学方法模拟人体组织的运动是非常困难的事情,而如何获得整个放射治疗期间器官真实的累加剂量,可能需要更进一步地改进算法,并充分考虑效率问题,为个体化的自适应放射治疗提供更加精确的支持.

[1]Yan D,Jaffray DA,Wong JW.A model to accumulate fractionated dose in a deforming organ[J].Int J Radiation Oncol Biol Phys,1999,44(3):89-105.

[2]Noel CE,Santanam L,Olsen JR,et al.An automated method for adaptive radiation therapy for prostate cancer patients using continuous fiducial-based tracking[J].Phys Med Biol,2010,55(1):65-82.

[3]Yang D,Chaudhari SR,Goddu SM,et al. Deformation registration of abdominal kilovoltage treatment planning CT and tomotherapy megavoltage CT for treatment adaptation[J].Med Phys,2009,36(2):329-338.

[4]Mackie TR,Kapatoes J,Ruchala K,et al.Image guidance for precise conformal radiotherapy[J].Int J Radiat Oncol Biol Phys,2003,56(1):89-105.

[5]Brock KK,Lee M,Eccles CL,et al.Deformableregistration and dose accumulation to investigate marginal liver cancer recurrences[J].Int J Radiat Oncol Biol Phys,2008,72(1):S538.

[6]Brock KM,Balter JM,Dawson LA,et al. Automated generation of a four-dimensional model of the liver using warping and mutual information.[J].Med Phys,2003,30(6):1128-1133.

[7]Zhang T,Jeraj R,Keller H,et al.Treatment plan optimization incorporating respiratory motion[J]. Med Phys,2004,31(6):1576-1586.

[8]Zhang TZ,Jeraj R,Keller H,et al.Treatment plan optimization incorporating respiratory motion[J]. Med Phys,2004,31(6):1576-1586.

[9]Lucas B,Kanade T.An iterative image registration technique with an application to stereo vision[C]//Proceedings of the 7th International Joint Conference on Artificial Intelligence,1981:674-679.

[10]Horn B,Schunck B.Determining optical flow[J]. Artif Intell,1981,17(2):185-203.

[11]Jin P,van der Horst A,de Jong R,et al. Marker-based quantification of interfractional tumor position variation and the use of markers for setup verification in radiation therapy for esophageal cancer[J].Radiother Oncol,2015,117(3):412-418.

[12]Bolton WD,Richey J,Ben-Or S,et al. Electromagnetic Navigational Bronchoscopy:A Safe and Effective Method for Fiducial Marker Placement in Lung Cancer Patients[J].Am Sur,2015,81(7):659-662.

[13]Garcia MM,Gottschalk AR,Brajtbord J,et al. Endoscopic gold fiducial marker placement into the bladder wall to optimize radiotherapy targeting for bladder-preserving management of muscle-invasive bladder cancer:feasibility and initial outcomes[J].Plos One,2014,9(3):e89754.

[14]Langerak T,Mens JW,Quint S,et al.Cervix motion in 50 cervical cancer patients assessed by daily cone beam computed tomographic imaging of a new type of marker[J].Int J Radiat Oncol Biol Phys,2015,93(3):532-539.

[15]Takao S,Miyamoto N,Matsuura T,et al. Intrafractional Baseline Shift or Drift of Lung Tumor Motion During Gated Radiation Therapy With a Real-Time Tumor-Tracking System[J]. IntJ Radiat Oncol Biol Phys,2016,94(4):172-180.

[16]Vemuri BC,Ye J,Chen Y,et al.Image registration via level-set motion:applications to atlas-based segmentation[J].Med Image Anal,2003,7(1):1-20.

Level set motion-based elastic registration of 3D CT images in radiotherapy

SHI Xiangxiang, TANG Tao, PANG Hao-wen

Objective: To apply level set motion-based elastic registration method to radiotherapy CT image, then to provide technical support for the accurate evaluation of tumor and changing process of patients with endanger organ and its accumulative dose. Methods: Based on Vemuri's level set motion method, we wrote the algorithm program using Matlab software and applied on two sets of 3D CT images from patients with cervical cancer and NPC for fully automatic elastic registration. Results: Comparing CT images before and after registration, for the cervical cancer patient, the minimum mean square error (MSE) decreased by 55.1% and correlation coefficient (CC) increased by 5.3%. For the NPC patient, MSE decreased by 32.1% and CC increased by 4.6%. Conclusion: From the image difference and evaluation parameters, the efficacy of level set motion-based elastic registration method was preliminarily demonstrated. In order to apply this method to clinical radiotherapy, dit needs to find a more accurate mathematical algorithm further in order to compute human anatomy deformation through image motion.

Level set motion; Elastic registration; Radiotherapy

1672-8270(2016)067-0020-03 [中图分类号] R814.42

A

10.3969/J.ISSN.1672-8270.2016.07.007

①西南医科大学附属医院肿瘤科 四川 泸州 64600

279165416@qq.com

2016-01-08