经典的demons算法利用参考图像的梯度信息作为配准的驱动力来配准图像,但是只适用于小形变配准,对于大形变配准效果并不理想,且存在收敛速度慢的问题,active demons算法在demons算法基础上加入了浮动图像的灰度梯度信息,加快了图像的收敛速度,并且对于大形变图像配准同样适用,可是拓扑结构容易发生改变.本文在基于变参数active demons算法基础上做改进,通过引入惯性系数b,提出基于变惯性系数active demons算法的DTI多通道配准,将变参数active demons的改进算法应用于DTI多通道图像配准.经典的active demons算法、变参数active demons算法和本文算法的MSE图做比较,可以发现本文算法可以进一步提高配准的收敛速度和精度,也使图像的拓扑结构得到很好的保持.
The dassical Demons algorithm is an important image registration algorithm,which effect. The algorithm uses the gradient information of the reference image as the driving force to register the image. However, the registration effect is not good for large deformation and the convergence rate is slow. The algorithm of active demons uses the gradient information of the reference image and the floating image as the driving force to accelerate the convergence of the image. It is also applicable to registration of large deformation images and the topology is easy to change. In this work, demons algorithm was applied to DTI multi-channel image registration. The algorithm of this paper is improved based on variable parameter active demons. By introducing the coefficient of inertia b, DTI multi-channel- DOI:10.3969/j.issn.1000-1565.2021.04.014基于变惯性系数active demons算法的DTI多通道配准研究郑伟1,2,3,周杨1,2,3,李文华1,2,3,刘帅奇1,2,3,张晓丹4,马泽鹏4(1.河北大学 电子信息工程学院,河北 保定 071002;2.河北省数字医疗工程重点实验室,河北 保定 071002;3.河北省机器视觉中心,河北 保定 071002;4.河北大学附属医院 CT-MRI诊断科,河北 保定 071000)摘 要:经典的demons算法利用参考图像的梯度信息作为配准的驱动力来配准图像,但是只适用于小形变配准,对于大形变配准效果并不理想,且存在收敛速度慢的问题,active demons算法在demons算法基础上加入了浮动图像的灰度梯度信息,加快了图像的收敛速度,并且对于大形变图像配准同样适用,可是拓扑结构容易发生改变.本文在基于变参数active demons算法基础上做改进,通过引入惯性系数b,提出基于变惯性系数active demons算法的DTI多通道配准,将变参数active demons的改进算法应用于DTI多通道图像配准.经典的active demons算法、变参数active demons算法和本文算法的MSE图做比较,可以发现本文算法可以进一步提高配准的收敛速度和精度,也使图像的拓扑结构得到很好的保持.关键词:梯度信息;大形变;驱动力;拓扑结构;非刚性配准 中图分类号:TN911 文献标志码:A 文章编号:1000-1565(2021)04-0436-07Study of DTI multi-channel registration based on the variable inertia coefficient active demons algorithmZHENG Wei1,2,3,ZHOU Yang1,2,3,LI Wenhua1,2,3,LIU Shuaiqi1,2,3,ZHANG Xiaodan4,MA Zepeng4(1. College of Electronic and Information Engineering, Hebei University, Baoding 071002, China; 2. Key Laboratory of Digital Medical Engineering of Hebei Province, Baoding 071002, China;3. Machine Vision Engineering Research Center of Hebei Province, Baoding 071002, China;4. Department of CT-MRI, Affiliated Hospital of Hebei University, Baoding 071000, China)Abstract: The dassical Demons algorithm is an important image registration algorithm,which effect. The algorithm uses the gradient information of the reference image as the driving force to register the image. However, the registration effect is not good for large deformation and the convergence rate is slow. The algorithm of active demons uses the gradient information of the reference image and the floating image as the driving force to accelerate the convergence of the image. It is also applicable to registration of large deformation images and the topology is easy to change. In this work, demons algorithm was applied to DTI multi-channel image registration. The algorithm of this paper is improved based on variable parameter active demons. By introducing the coefficient of inertia b, DTI multi-channel- 收稿日期:2020-03-14 基金项目:国家自然科学基金资助项目(2014000000;61401308;61572063);河北省自然科学基金资助项目(F2020201025; H2020201021) 第一作者:郑伟(1972—),女,黑龙江兰西人,河北大学教授,博士,主要从事图像处理与分析、图像安全通信、图像加密和隐藏研究. E-mail:147685650@qq.com 通信作者:马泽鹏(1989—),男,河北保定人,河北大学附属医院医师,主要从事脑部和心脏疾病的CT、MR、fMRI的诊断及鉴别诊断.E-mail:mzpdan@163.com第4期郑伟等:基于变惯性系数active demons算法的DTI多通道配准研究registration based on variable inertia factor Active demons algorithm was proposed. Comparison was made between the MSE graphs of the three algorithms, namely, the classical Active demons algorithm, the variable parameter Active demons algorithm and the algorithm. It can be found that the algorithm in this paper can further improve the convergence speed and accuracy of registration. Meanwhile, the topological structure of the image remains unchanged.
[1] 王天朴,施海燕.阿尔茨海默病治疗研究进展[J].世界最新医学信息文摘,2019,19(80):72-73. DOI:10.19613/j.cnki.1671-3141.2019.80.032.
[2] 刘学源,方珉.阿尔茨海默病的神经影像学研究[J].老年医学与保健,2012,18(2):73-75. DOI:10.3969/j.issn.1008-8296.2012-02-04.
[3] OLE F O,MADS N. The generic structure of the optic flow field[J].Journal of Mathematical Imaging and Vision, 2006(24):37-53.DOI:10.1007/Sl0851-005-3614-2.
[4] LIU Y,CHENG H D,HUANG J H,et al. An effective non-rigid registration approach for ultrasound image based on Demons algorithm[J]. Journal of Digital Imaging,2013,26(3)521-529. DOI:10.1007/s10278-012-9532-0.
[5] CAZOULAT G, SIMON A, DUMENIL A. Surface-constrained nonrigid registration for dose monitoringin prostate cancer radioth erapy[J]. IEEE Transactions on Medical Imaging, 2014,33(7):1464-1474. DOI:10.1109/TMI.2014.2314574
[6] 李果霖.基于改进Demons算法的图像配准研究[D].哈尔滨:哈尔滨工程大学,2016.
[7] 蒋晨娇.基于改进鸡群算法的AD患者DTI图像配准[D].保定:河北大学,2018.
[8] GUO T, WANG Q,WANG Y. High-Dimensional Spatial Standardization Algorithm for Diffusion Tensor Image Registration[J].Journal of Beijing Institute of Technology,2018,27(4):604-616. DOI:10.15918/j.jbit1004-0579.18033.
[9] 张桂梅,胡强.基于自适应分数阶的医学图像非刚性配准[J].自动化学报,2020,46(9):1941-1951. DOI:10.16383/j.aas.c190027
[10] 许磊,戚睿,袁杰等.基于改进光流场算法在复杂背景变化图像处理中的应用[J].上海航天,2019,36(4):110-114. DOI:10.19328/j.cnki.1006-1630.2019.04.016.
[11] SIMONA M,PANA T L,ANJAN B,et al. Combining sparse and dense features to improve multi-modal registration for Brain DTI Images[J]. Entropy(Basel, Switzerland),2020,22(11):e22111299. DOI:10.3390/E22111299.
[12] WANG H, DONG L,RADHE M, et al.Validation of an accelerated ‘demons’ algorithm for deformable image registration in radiation therapy[J]. Physics in Medicine and Biology, 2005(12):2887-2905. DOI:10.1088/0031-9155/50/12/011.
[13] XUE P,YANG P,CAO Z L,et al.Active demons non-rigid registration algorithm based on balance coefficient[J].Acta Automat-ica Sinica,2016(9):1389-1400.DOI:10.16383/j.aas.2016.c150186.
[14] 王昌,任琼琼等.基于局部联合熵梯度的双向多分辨率Demons算法[J].生物医学工程研究,2018(2):159-162. DOI:10.19529/j.cnki.1672-6278.2018.02.08.
[15] 薛文静,党建武等.基于局部熵的Active Demons多模医学图像配准[J].宁夏大学学报(自然科学版),2019,40(1):6-12. DOI:10.3969/j.issn.0253-2328.2019.01.002.
[16] 徐晓莹,赵杰等.Active demons算法下的多通道弥散张量图像配准[J].变参数中国图像图形学报, 2019,24(1):103-114.DOI:10.11834/jig.180281.
[17] Thirion J P.Image matching as a diffusion process: an analogy with Maxwell's demons[J]. Medical Image Analysis, 1998, 2(3)243-260. DOI:10.1016/S1361-8415(98)80022-4.
[18] PETER R, STANISLAV K. Symmetric image registration[J]. Medical Image Analysis, 2006, 10(3):484-493. DOI:10.1016/j.media.2005.03.003.
[19] 罗国忠.惯性概念建立过程的推断过渡及其解决[J].中学物理教学参,2017,46(17):41-43. DOI:1002-218X(2017)09-0041-03.
[20] PARK H J,KUBICKI M,SHENTON M E,et al.Spatial normalization of diffusion tensor MRI using multiple channels[J].Neuroim-age,2003,20(4):1995-2009. DOI: 10.1016 / j.neuroim-age.2003.08.008. (
