基于互信息的熒光素眼底血管造影圖像序列的自動(dòng)配準(zhǔn)方法

劉小燕; 王皓浩; 孫剛; 張譜; 劉敏; 高玲

doi:10.11999/JEIT170868

基于互信息的熒光素眼底血管造影圖像序列的自動(dòng)配準(zhǔn)方法

doi: 10.11999/JEIT170868

劉小燕^1, ,,
王皓浩¹,
孫剛¹,
張譜²,
劉敏¹,
高玲²

1.
湖南大學(xué)電氣與信息工程學(xué)院 ??長(zhǎng)沙 ??410082
2.
中南大學(xué)湘雅附屬第二醫(yī)院 ??長(zhǎng)沙 ??410011

詳細(xì)信息

作者簡(jiǎn)介:
劉小燕：女，1973年生，教授，博士生導(dǎo)師，研究方向?yàn)閳D像處理技術(shù)及其應(yīng)用、智能建模與控制

王皓浩：男，1994年生，碩士生，研究方向?yàn)獒t(yī)學(xué)圖像處理技術(shù)

孫剛：男，1992年生，博士生，研究方向?yàn)獒t(yī)學(xué)圖像處理技術(shù)

張譜：男，1985年生，博士生，研究方向?yàn)橐暰W(wǎng)膜、脈絡(luò)膜及玻璃體相關(guān)疾病

劉敏：男，1981年生，副教授，博士生導(dǎo)師，研究方向?yàn)橛?jì)算機(jī)視覺(jué)、模式識(shí)別以及機(jī)器學(xué)習(xí)

高玲：女，1968年生，主任醫(yī)師，研究方向?yàn)橐暰W(wǎng)膜、脈絡(luò)膜及玻璃體相關(guān)疾病

通訊作者:
劉小燕 ? xiaoyan.liu@hnu.edu.cn

中圖分類號(hào): TP391; R445
計(jì)量
- 文章訪問(wèn)數(shù): 1971
- HTML全文瀏覽量: 886
- PDF下載量: 70
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2017-09-14
- 修回日期: 2018-05-09
- 網(wǎng)絡(luò)出版日期: 2018-06-07
- 刊出日期: 2018-08-01

A Novel Automatic Registration Method for Fluorescein Fundus Angiography Sequences Based on Mutual Information

1.
College of Electrical and Information Engineering, Hunan University, Changsha 410082, China
2.
Department of Ophthalmology, the Second Xiangya hospital of Central South University, Changsha 410011, China

摘要

摘要: 熒光素眼底血管造影技術(shù)(FFA)是眼底疾病診斷的金標(biāo)準(zhǔn)，但是造影過(guò)程中病人不可避免地轉(zhuǎn)動(dòng)眼球，造成FFA圖像序列中感興趣區(qū)域(例如視網(wǎng)膜血管分支、新生血管)的位置發(fā)生變化，給后續(xù)的圖像定量分析與病情準(zhǔn)確評(píng)估診斷帶來(lái)困難。針對(duì)上述問(wèn)題，該文提出一種基于互信息的FFA圖像序列配準(zhǔn)方法。首先采用多尺度線性濾波方法分割出圖像中的血管，并利用圖像金字塔對(duì)分割后的圖像進(jìn)行下采樣，然后利用互信息計(jì)算待配準(zhǔn)圖像與參考圖像的相似性，通過(guò)進(jìn)化策略對(duì)配準(zhǔn)參數(shù)進(jìn)行優(yōu)化，獲得互信息最大時(shí)圖像的空間變換矩陣，實(shí)現(xiàn)FFA圖像的配準(zhǔn)。采用上述方法，對(duì)4位患者共計(jì)1039幀F(xiàn)FA圖像進(jìn)行測(cè)試，總體配準(zhǔn)率達(dá)到93%，失敗率僅為1%；與常用的配準(zhǔn)方法相比，所提方法的配準(zhǔn)率、配準(zhǔn)速度和魯棒性等綜合性能良好，為FFA影像的定量分析在未來(lái)的臨床應(yīng)用奠定了基礎(chǔ)。
- 熒光素眼底血管造影 /
- 圖像序列 /
- 互信息 /
- 血管分割
Abstract: Fluorescein Fundus Angiography (FFA) is regarded as the golden diagnostic criteria for fundus diseases. However, dislocation or rotation of the interested images on anatomic landmark (like retinal vascular branches, neovascularization), caused by inevitable eyeball movement, brings about difficulties in subsequent quantitative analysis and progress assessment of the diseases. In order to solve the above problems, a novel method based on mutual information is proposed for automatic registration of FFA image sequence. Firstly, the vessels of image sequence are segmented by multi-scale linear filter and down sampled hereafter by image pyramid. Then, the similarity of sampled images is calculated by mutual information and the evolution strategy is adopted to optimize the registration parameters. Finally, the transformation matrix with maximum mutual information is obtained to register the FFA image. Tests with FFA image sequences of 4 patients (total 1039 frames) show that the overall registration rate of the algorithm reaches 93% and the failure rate is only 1%. Compared with the classical registration methods, the proposed method shows better comprehensive performance in terms of registration rate, computing speed as well as robustness. It lays basic foundations for quantitative analysis on FFA images and potential clinical application.
- Fluorescein Fundus Angiography (FFA) /
- Image sequence /
- Mutual information /
- Vascular segmentation

HTML全文

圖 1 基于互信息的FFA圖像序列配準(zhǔn)方法流程圖

下載: 全尺寸圖片幻燈片

圖 2 線性濾波器示意圖

下載: 全尺寸圖片幻燈片

圖 3 FFA序列圖像分割示意圖

下載: 全尺寸圖片幻燈片

圖 4 基于互信息的變換矩陣優(yōu)化過(guò)程

下載: 全尺寸圖片幻燈片

圖 5 患者1#FFA圖像序列配準(zhǔn)過(guò)程的參考圖像、待配準(zhǔn)圖像以及配準(zhǔn)前后的棋盤(pán)圖

下載: 全尺寸圖片幻燈片

圖 6 配準(zhǔn)失敗的圖像示例

下載: 全尺寸圖片幻燈片

圖 7 本文算法、GDB-ICP及Glocker B算法對(duì)完全充盈期FFA圖像的配準(zhǔn)棋盤(pán)圖

下載: 全尺寸圖片幻燈片

圖 8 本文算法、GDB-ICP及Glocker B算法對(duì)靜脈開(kāi)始充盈期FFA圖像的配準(zhǔn)棋盤(pán)圖

下載: 全尺寸圖片幻燈片

表 1 本文算法的配準(zhǔn)精度測(cè)試結(jié)果

編號(hào)	FFA圖像序列的總幀數(shù)n	配準(zhǔn)效果				失敗率(%)	配準(zhǔn)率P(%)
編號(hào)	FFA圖像序列的總幀數(shù)n	Y₁類幀數(shù)	Y₂類幀數(shù)	F₁類幀數(shù)	F₂類幀數(shù)	失敗率(%)	配準(zhǔn)率P(%)
患者1#	218	195	10	12	1	0.4	94
患者2#	219	205	5	9	0	0	96
患者3#	261	232	13	8	8	3.0	94
患者4#	341	299	10	30	2	0.6	91
總計(jì)	1039	931	38	59	11	1.0	93

下載: 導(dǎo)出CSV

表 2 GDB-ICP算法對(duì)4位患者的FFA圖像序列配準(zhǔn)精度

編號(hào)	FFA圖像序列的總幀數(shù)	配準(zhǔn)效果				失敗率(%)	配準(zhǔn)率P(%)
編號(hào)	FFA圖像序列的總幀數(shù)	Y₁類幀數(shù)	Y₂類幀數(shù)	F₁類幀數(shù)	F₂類幀數(shù)	失敗率(%)	配準(zhǔn)率P(%)
患者1#	218	170	2	14	32	15	79
患者2#	219	172	3	6	38	17	80
患者3#	261	244	1	6	10	4	94
患者4#	341	295	13	17	16	6	90
總計(jì)	1039	881	19	43	96	9	87

下載: 導(dǎo)出CSV

表 3 Glocker B算法對(duì)4位患者的FFA圖像序列配準(zhǔn)精度

編號(hào)	FFA圖像序列的總幀數(shù)	配準(zhǔn)效果				失敗率(%)	配準(zhǔn)率P(%)
編號(hào)	FFA圖像序列的總幀數(shù)	Y₁類幀數(shù)	Y₂類幀數(shù)	F₁類幀數(shù)	F₂類幀數(shù)	失敗率(%)	配準(zhǔn)率P(%)
患者1#	218	106	1	1	110	50	49
患者2#	219	82	1	29	107	49	38
患者3#	261	75	3	60	123	47	30
患者4#	341	227	2	18	94	28	67
總計(jì)	1039	490	7	108	434	42	48

下載: 導(dǎo)出CSV

表 4 本文算法、GDB-ICP以及Glocker B算法的運(yùn)行時(shí)間對(duì)比(min)

編號(hào)	FFA圖像序列幀數(shù)	本文算法	GDB-ICP	Glocker B
患者1#	218	32.45	68.17	13.08
患者2#	219	31.75	62.42	13.33
患者3#	261	37.93	101.13	15.32
患者4#	341	50.35	125.77	20.00

下載: 導(dǎo)出CSV

參考文獻(xiàn)(23)

DREO J, NUNES J C, and SIARRY P. Robust rigid registration of retinal angiograms through optimization[J]. Computerized Medical Imaging & Graphics, 2006, 30(8): 453－463. DOI: 10.1016/j.compmedi mag.2006.07.004.

LAAKSONEN L, CLARIDGE E, F?LT P, et al. Comparison of image registration methods for composing spectral retinal images[J]. Biomedical Signal Processing & Control, 2017, 36: 234－245. DOI: 10.1016/j.bspc.2017.03.003.

COPELAND A, MANGOUBI R, DESAI M, et al. Enhancing the surgeons reality: Smart visualization of bolus time of arrival and blood flow anomalies from time lapse series for safety and speed of cerebrovascular surgery[C]. IEEE Applied Imagery Pattern Recognition Workshop, Washington, DC, USA, 2009: 1–4.

GHASSABI Z, SHANBEHZADEH J, and MOHAMMADZADEH A. A structure-based region detector for high-resolution retinal fundus image registration[J]. Biomedical Signal Processing & Control, 2016, 23: 52－61.DOI: 10.1016/j.bspc.2015.08.005

PEREZROVIRA A, CABIDO R, TRUCCO E, et al. RERBEE: Robust efficient registration via bifurcations and elongated elements applied to retinal fluorescein angiogram sequences[J]. IEEE Transactions on Medical Imaging, 2012, 31(1): 140－150. DOI: 10.1109/TMI.2011.2167517.

TSAI C L, LI Chunyi, YANG Gehua, et al. The edge-driven dual-bootstrap iterative closest point algorithm for registration of multimodal fluorescein angiogram sequence[J]. IEEE Transactions on Medical Imaging, 2010, 29(3): 636－649. DOI: 10.1109/TMI.2009.2030324.

汪立, 蔣念平. 基于改進(jìn)Harris角點(diǎn)檢測(cè)的視網(wǎng)膜圖像配準(zhǔn)[J]. 電子科技, 2017, 30(2): 119–122. DOI: 10.16180/j.cnki.issn1007－7820.2017.02.031.

WANG Li and JIANG Nianping. Retinal image registration based on improved Harris corner detection[J]. Electronic Science and Technology, 2017, 30(2): 119–122. DOI: 10.16180/j.cnki.iss n1007－7820.2017.02.031.

YAVUZ Z and K?SE C. Retinal fundus image registration using bifurcation and crossover points[C]. Signal Processing and Communication Application Conference, Zonguldak, Turkey, 2016: 1485–1488.

PALRAJ P and VENNILA I. Retinal fundus image registration via blood vessel extraction using binary particle swarm optimization[J]. Journal of Medical Imaging & Health Informatics, 2016, 6(2): 328－337. DOI: 10.1166/jmihi.2016.1701

SAHA S K, DI X, FROST S, et al. A two-step approach for longitudinal registration of retinal images[J]. Journal of Medical Systems, 2016, 40: 277. DOI: 10.1007/s10916-016-0640-0.

CAN A, STEWART C V, TANENBAUM H L, et al. A Feature-based, robust, hierarchical algorithm for registering pairs of images of the curved human retina[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2002, 24(3): 347－364. DOI: 10.1109/34.990136.

李英杰, 張俊舉, 常本康, 等. 電子與信息學(xué)報(bào), 2016. 38(1): 8－14. DOI: 10.11999/JEIT150479.

LI Yingjie, ZHANG Junju, CHANG Benkang, et al. A multi - band infrared image joint registration and fusion method[J]. Journal of Electronics & Information Technology, 2016. 38(1): 8－14. DOI: 10.11999/JEIT150479.

HENEGHAN C, MAGUIRE P, RYAN N, et al. Retinal image registration using control points[C]. Proceedings IEEE International Symposium on Biomedical Imaging, Washington, DC, USA, 2002: 349–352. doi: 10.1109/ISBI. 2002.1029265.

劉妍, 余淮, 楊文, 等. 利用SAR-FAST角點(diǎn)檢測(cè)的合成孔徑雷達(dá)圖像配準(zhǔn)方法[J]. 電子與信息學(xué)報(bào), 2017, 39(2): 430–436. DOI: 10.11999/JEIT160386.

LIU Yan, YU Huai, YANG Wen, et al. Synthetic aperture radar image registration method using SAR - FAST corner detection [J]. Journal of Electronics & Information Technology, 2017, 39(2): 430–436. DOI: 10.11999/JEIT160386.

LALIBERTé F, GAGNON L, and SHENG Y. Registration and fusion of retinal images--an evaluation study[J]. IEEE Transactions on Med Imaging, 2003, 22(5): 661－673. DOI: 10.1109/TMI.2003.812263.

任克強(qiáng), 胡夢(mèng)云. 基于改進(jìn)SURF算子的彩色圖像配準(zhǔn)算法[J]. 電子測(cè)量與儀器學(xué)報(bào), 2016, 30(5): 748－756. DOI: 10.13382/j.jemi.2016.05.011.

REN Keqiang and HU Mengyun. Color image registration algorithm based on improved SURF operator[J]. Journal of Electronic Measurement and Instrument, 2016, 30(5): 748－756. DOI: 10.13382/j.jemi.2016.05.011.

CHEN Jian, TIAN Jie, LEE N, et al. A partial intensity invariant feature descriptor for multimodal retinal image registration[J]. IEEE Transactions on Biomedical Engineering, 2010, 57(7): 1707－1718. DOI: 10.1109/TBME.2010.2042169.

WANG Guang, WANG Zhicheng, CHEN Yufei, et al. Robust point matching method for multimodal retinal image registration[J]. Biomedical Signal Processing & Control, 2015, 19: 68－76. DOI: 10.1016/j.bspc.2015.03.004.

YANG G, STEWART C V, SOFKA M, et al. Registration of challenging image pairs: initialization, estimation, and decision[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2007, 29(11): 1973－1989. DOI: 10.1109/TPAMI.2007.1116.

NUNES J C and BOUAOUNE Y. A multiscale elastic registration scheme for retinal angiograms[J]. Computer Vision & Image Understanding, 2004, 95(2): 129－149. DOI: 10.1016/j.cviu.2004.03.007.

NGUYEN U T V, BHUIYAN A, PARK L A F, et al. An effective retinal blood vessel segmentation method using multi-scale line detection[J]. Pattern Recognition, 2013, 46(3): 703－715. DOI: 10.1016/j.patcog.2012.08.009.

STYNER M, BRECHBüHLER C, SZCKELY G, et al. Parametric estimate of intensity inhomogeneities applied to MRI[J]. IEEE Transactions on Medical Imaging, 2000, 19(3): 153－165. DOI: 10.1109/42.845174.

GLOCKER B, KOMODAKIS N, PARAGIOS N, et al. Inter and intra-modal deformable registration: Continuous deformations meet efficient optimal linear programming[C]. IPMI'07 Processings of the 20th International Conference on Information Processing in Medical Imaging, Kerkrade, The Netherlands, 2007: 408–420.

相關(guān)文章

施引文獻(xiàn)

資源附件(0)

訪問(wèn)統(tǒng)計(jì)