CT/PET Lung Registration Using Adaptive Thin Plate Spline-Based Total Variation Regularization

doi:10.3778/j.issn.1002-8331.1711-0288

Computer Engineering and Applications ›› 2019, Vol. 55 ›› Issue (3): 202-208.DOI: 10.3778/j.issn.1002-8331.1711-0288

Previous Articles Next Articles

CT/PET Lung Registration Using Adaptive Thin Plate Spline-Based Total Variation Regularization

DU Xueying1，2, GONG Lun1，2, LIU Zhaobang1, ZHANG Cheng1, LIU Hanqiu3, DING Min4, ZHENG Jian1

1.Department of Medical Imaging, Suzhou Institute of Biomedical Engineering Technology, Chinese Academy of Sciences, Suzhou, Jiangsu 215163, China
2.College of Materials Science and Opto-Electronic Technology, University of Chinese Academy of Sciences, Beijing 101400, China
3.Huashan Hospital, Fudan University, Shanghai 200040, China
4.School of Science, Nanjing University of Science and Technology, Nanjing 210000, China

Online:2019-02-01 Published:2019-01-24

基于自适应薄板样条全变分的肺CT/PET图像配准

杜雪莹1，2，龚伦1，2，刘兆邦1，章程1，刘含秋3，丁敏4，郑健1

1.中国科学院苏州生物医学工程技术研究所医学影像室，江苏苏州 215163
2.中国科学院大学材料科学与光电技术学院，北京 101400
3.复旦大学附属华山医院，上海 200040
4.南京理工大学理学院，南京 210000

Abstract

Abstract: The discontinuities displacement field at the boundaries is preserved through Total Variation（TV） regularization when registering organ images with sliding motion. But TV assumes a global regularization which will lead poor correspondences at local areas of images. This paper creates adaptive Thin Plate Spline-based Total Variation（TPS-TV） regularization combining Thin Plate Spline（TPS） and TV operator according to the distance of pixels to boundaries. Then this paper chooses Correlation Ratio-based Mutual Information（CRMI） similarity measure function and Limited-memory Broyden Fletcher Goldfarb Shanno（L-BFGS） optimization to establish a non-rigid registration frame. Compared with TV and TPS regularization on the public DIR-Lab dataset and clinical CT/PET dataset, the proposed method has been demonstrated a more dependable displacement field and higher registration accuracy.

Key words: adaptive Thin Plate Spline-based Total Variation（TPS-TV）, sliding motion, non-rigid registration, lung

摘要： 全变分正则项虽然能够在具有滑移运动的肺等胸腹部器官图像配准时校正边界不连续位移场，但仍然无法保留图像的局部特征，损失配准精度。针对肺图像CT单模配准和CT/PET双模配准，通过像素点空间位置权重将薄板样条能量算子与全变分算子进行空间加权建立自适应薄板样条全变分正则项。然后，将正则项与CRMI相似性测度以及L-BFGS优化方法结合建立非刚性配准算法。通过DIR-Lab 4D-CT公共数据集和CT/PET临床数据集上的实验结果表明，提出的方法能够在保证边界不连续运动的同时保证图像内部的平滑性，具有更高的配准精度。

关键词: 自适应薄板样条全变分, 滑移运动, 非刚性配准, 肺

DU Xueying1，2, GONG Lun1，2, LIU Zhaobang1, ZHANG Cheng1, LIU Hanqiu3, DING Min4, ZHENG Jian1. CT/PET Lung Registration Using Adaptive Thin Plate Spline-Based Total Variation Regularization[J]. Computer Engineering and Applications, 2019, 55(3): 202-208.

杜雪莹1，2，龚伦1，2，刘兆邦1，章程1，刘含秋3，丁敏4，郑健1. 基于自适应薄板样条全变分的肺CT/PET图像配准[J]. 计算机工程与应用, 2019, 55(3): 202-208.

[1]	XU Qihao, LI Bo. Method for Detecting Pulmonary Nodules Based on NRU Network [J]. Computer Engineering and Applications, 2021, 57(4): 83-90.
[2]	WANG Liuyi, SONG Wen’ai, LIN Xinshan, YUE Ning, YANG Jijiang, WANG Qing, LEI Yi. Review of Computer-Aided Diagnosis of Bronchiectasis with CT Images [J]. Computer Engineering and Applications, 2021, 57(11): 11-20.
[3]	TANG Ning, WEI Zeliang, ZHANG Rui, YI Dong, WU Yazhou. Comparative Study on Classification of Pulmonary Nodules Based on Multi-Scale and Multi-Mode Image [J]. Computer Engineering and Applications, 2020, 56(3): 165-175.
[4]	ZHONG Sihua, GUO Xingming, ZHENG Yineng. Improved U-Net Network for Lung Nodule Segmentation [J]. Computer Engineering and Applications, 2020, 56(17): 203-209.
[5]	WU Baorong, QIANG Yan, WANG Sanhu, TANG Xiaoxian, LIU Xijing. Fusing Multi-Dimensional Convolution Neural Network for Lung Nodules Classification [J]. Computer Engineering and Applications, 2019, 55(24): 171-177.
[6]	YAN Xiaofei1, QIANG Yan1, YUAN Jie2. Diagnosis of benign and malignant lung nodules based on hybrid restricted Boltzmann machine [J]. Computer Engineering and Applications, 2017, 53(23): 153-158.
[7]	SHI Lukui1，2, LIU Wenhao1, LI Zhanru1. Feature extraction method of lung sound based on LDA and wavelet decomposition [J]. Computer Engineering and Applications, 2017, 53(22): 116-120.
[8]	CHEN Jin1，2, ZHANG Yu1，2. Multi-phase similarity based super-resolution reconstruction for lung 4D-CT [J]. Computer Engineering and Applications, 2017, 53(15): 191-195.
[9]	WU Panpan, XIA Kewen, LIN Yongliang, BAI Jianchuan. Improved RVM based research and application on lung nodule detection [J]. Computer Engineering and Applications, 2016, 52(19): 201-207.
[10]	SHI Rui, HUANG Xiangjuan. Lung segmentation based on improved Snake model [J]. Computer Engineering and Applications, 2014, 50(15): 164-167.
[11]	ZHANG Xin1, WANG Bing2, YANG Ying3, WANG Miao2, WU Chen2, ZHANG Jinxing2. Automatic repair of lung boundary concave in chest CT images [J]. Computer Engineering and Applications, 2013, 49(24): 191-194.
[12]	FAN Linan, LI Daojing, SUN Shenshen, CHANG Chaohai. Applied research on automatic detection of lung nodules ROI based on Top-hat and Gabor filter [J]. Computer Engineering and Applications, 2012, 48(4): 190-192.
[13]	HAN Yanyan1, FENG Jun1, CUI Xin1, WANG Qiuping2. Lung nodule detection based on Dynamic Multiple Classifiers Selection ensemble algorithm [J]. Computer Engineering and Applications, 2012, 48(2): 218-221.
[14]	ZHANG Jing1，LI Bin1，TIAN Lianfang1，CHEN Ping2. Lung nodule recognition system using rules [J]. Computer Engineering and Applications, 2011, 47(17): 200-202.
[15]	XIANG Shi-jie¹，LI Bin¹，TIAN Lian-fang¹，WANG Li-fei²，CHEN Ping³. Suspected lung nodule segmentation method of edge-preserving smoothing and improved FCM [J]. Computer Engineering and Applications, 2010, 46(32): 202-204.

CT/PET Lung Registration Using Adaptive Thin Plate Spline-Based Total Variation Regularization

基于自适应薄板样条全变分的肺CT/PET图像配准

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics