摘 要：糖尿病性视网膜病变、青光眼病变等视网膜疾病是目前高致盲眼病，其病理特征表现为层状组织结构的异常，因此具有高精度、高鲁棒性的视网膜层分割技术是视网膜疾病筛查的重要依据。通过对四种经典分层模型实现原理和发展历程的详细阐述，对比分析出各种分层模型的固有特性，介绍了自动分层技术的最新研究进展以及眼科领域应用。直观地展示了自动分层技术与人工智能相结合的发展趋势，为视网膜层状结构分割技术的深入研究和实用化提供参考。
关键词：视网膜层分割技术；经典分层模型；眼科应用；人工智能
中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2020）24-0023-07
A Review of Automatic Slicing Techniques for Retinal Optical Coherence Tomography
FANG Yang1，HU Jianming1，CHEN Ge1，2
（1.College of Physics and Electronic Engineering，Chongqing Normal University，Chongqing 401331，China；2.School of Physics，University of Electronic Science and Technology of China，Chengdu 610054，China）
Abstract：Diabetic retinopathy，glaucoma and other retinal diseases are currently high blinding diseases，whose pathological characteristics are abnormal layered tissue structure. Therefore，retinal layer segmentation technology with high accuracy and high robustness is an important basis for retinal disease screening. Based on the elaboration of the realization principle and development history of four classical hierarchical models，the inherent characteristics of each stratification model are compared and analyzed，and the latest research progress of automatic layering technology and its application in ophthalmology are introduced. The development trend of the combination of automatic layering technology and artificial intelligence is an intuitive demonstration，which provides a reference for the further research and practical application of retinal layered structure segmentation technology.
Keywords：retinal layer segmentation technology；classical hierarchical model；ophthalmic application；artificial intelligence
参考文献：
[1] 蔡怀宇，张玮茜，陈晓冬，等. 眼科光学相干层析成像的图像处理方法 [J]. 中国光学，2019，12（4）：731-740.
[2] HUANG D，SWANSON E A，LIN C，et al. Optical coherence tomography [J].Science，1991，254（5035）：1178-1181.
[3] SRINIVASAN P P，HEFLIN S J，IZATT J A，et al.Automatic segmentation of up to ten layer boundaries in SD-OCT imagesof the mouse retina with and without missing layers due to pathology [J].Biomedical Optics Express，2014，5（2）：348-365.
[4] HEE M R，IZATT J A，SWANSON E A，et al. OpticalCoherence Tomography of the Human Retina [J].Archives ofOphthalmology，1995，113（3）：325-332.
[5] LU S J，CHEUNG C Y L，LIU J，et al. Automated Layer Segmentation of Optical Coherence omography Images [J].IEEE Transactions on bio-medical engineering，2010，57（10），2605-2608.
[6]T I A N J，VA R G A B，S O M FA I G M，e t a l . R e a l -Time Automatic Segmentation of Optical Coherence TomographyVolume Data of the Macular Region [J/OL].Plos One，2015，10（8）：e0133908[2020-07-05].https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0133908.DOI：10.1371/journal.pone.0133908.
[7] 刘云. 基于深度学习的视网膜OCT 图像分层与疾病筛查研究 [D]. 济南：山东大学，2018.
[8] SHAH A，ZHOU L X，MICHAEL A，et al. Multiple surfacesegmentation using convolution neural nets：application to retinal layersegmentation in OCT images [J].Biomedical Optics Express，2018，9（9）：4509-4526.
[9] 陈明惠，秦显富，贾文宇，等. 一种眼底黄斑水肿OCT图像分割方法 [J]. 光学技术，2019，45（6）：730-736.
[10] CHEN A，ONG C Z L，LUO W W，et al. Comparisonof training strategies for the segmentation of retina layers in opticalcoherence tomography images of rodent eyesusing convolutional neuralnetworks [C]//Image Processing.Houston：SPIE Medical Imaging，2020：821-825
[11] 中山大学. 基于深度学习的视网膜层和积液区域的层分割方法及系统：CN202010313682.7 [P].2020-08-25.
[12] 贺琪欲，李中梁，王向朝，等. 基于光学相干层析成像的视网膜图像自动分层方法 [J]. 光学学报，2016，36（10）：69-78.
[13] FABRITIUS T，MAKITA S，MIURA M，et al. AutomatedSegmentation of the Macula by Optical Coherence Tomography [J].Optics Express，2009，17（18）：15659-15669.
[14] ISHIKAWA H，STEIN D M，WOLLSTEIN G，et al.Macular Segmentation with Optical Coherence Tomography [J].InvestOphthalmol Vis Sci，2005，46（6）：2012–2017.
[15] MONEMIAN M，RABBANI H. Analysis of A NovelSegmentation Algorithm for Optical Coherence Tomography ImagesBased on Pixels Intensity Correlations [J].IEEE Transactions onInstrumentation and Measurement，1（70），2020.
[16] FERNÁNDEZ D C，VILLATE N，PULIAFITO C A，et al. Comparing total macular volume changes measured by OpticalCoherence Tomography with retinal lesion volume estimated by activecontours [J].Investigative Ophthalmology & Visual Science，2004，45：3072.
[17] YAZDANPANAH A，HAMARNEH G，SMITH B R，et al.Segmentation of intra-retinal layers from optical coherence tomographyimages using an active contour approach [J].IEEE Transactions onMedical Imaging，2011，30（2）：484-496.
[18] DODO B，LI Y M，Liu X H，et al. Level set segmentationof retinal oct images [C]//Proceedings of the 12th International JointConference on Biomedical Engineering Systems and Technologies.Prague：SCITE PRESS，2019：49-56.
[19] YANG Q，REISMAN C A，CHAN K P，et al. Automated Segmentation of outer Retinal Layers in Macular OCT Images of Patients with Retinitis Pigmentosa [J].Biomedical Optical Express，2011，2（9），2493-2503．
[20] M I S H R A Z，GANEGODA A，S E L I C H A J，e t a l .Automated Retinal Layer Segmentation Using Graph-based AlgorithmIncorporating Deep-learning-derived Information [J].Scientific Reports，2020（10）：9541.
[21] NGO L，CHA J，HAN J H，Deep Neural NetworkRegression for Automated Retinal Layer Segmentation in OpticalCoherence Tomography Images [J].IEEE Transactions on ImageProcessing，2019，29：303-312.
[22] MASOOD S，FANG R G，LI P，et al. Automatic Choroid Layer Segmentation from Optical Coherence Tomography Images Using Deep Learning [J].Scientific Reports，2019，9（1）：3058.
[23] HAEKER M，ABRÀMOFF M，KARDON R，et al.Segmentation of the surfaces of the retinal layer from OCT images [C]//Medical Image Computing and Computer-Assisted Intervention – MICCAI 2006.Gopenhagen：Springer，2006：800–807.
[24] GARVIN M K，ABRAMOFF M D，KARDON R，et al. Intraretinal Layer Segmentation of Macular Optical CoherenceTomography Images Using Optimal 3–D Graph Search [J].IEEETransaction on Medical Imaging，2008，27（10）：1495–1505.
[25] QUELLEC G，LEE K，DOLEJSI M，et al. Threedimensionalanalysis of retinal layer texture: identification of fluid-filled regions in SD-OCT of the macula [J].IEEE Transaction on Medical Imaging，2010，29（6）：1321–1330.
[26] TIAN J，VARGA B，SOMFAI G M，et al. Real-Time Automatic Segmentation of Optical Coherence Tomography Volume Data of the Macular Region [J/OL].Plos One，2015，10（8）：e0133908[2020-07-25].https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0133908.DOI:10.1371/journal.pone.0133908.
[27] FANG L Y，CUNEFARE D，WANG C，et al. Automatic segmentation of nine retinal layer boundaries in OCT images of nonexudative AMD patients using deep learning and graph search [J]. Biomedical Optics Express，2017，8（5）：2732-2744.
[28] MENDES O L C，LUCENA A R，LUCENA D R，et al. Automatic Segmentation of Macular Holes in Optical Coherence Tomography Images: A review [J].Journal of Artificial Intelligence and Systems，2019（1）：163-185.
[29] 国家卫生和计划生育委员会.“十三五”全国眼健康规划（2016—2020 年） [J]. 中华眼科杂志，2017，53（7）：484-486.
[30] LEE R，WONG T Y，SABANAYAGAM C. Epidemiology of Diabetic Retinopathy，Diabetic Macular Edema and Related Vision Loss [J].Eye and vision（London，England），2015，2（17）.
[31] HELMY Y M，ALLAH H R A，et al. Optical Coherence Tomography Classification of Diabetic Cystoid Macular Edema [J].Clinical Ophthalmology，2013（7）：1731–1737.
[32] LIEW G，MICHAELIDES M，BUNCE C. A comparison ofthe Causes of Blindness Certifications in England and Wales in WorkingAge Adult（16-64years），1999-2000 with 2009-2010 [J/OL].BMJOpen，2014（4）：e004015[2020-07-29].https://bmjopen.bmj.com/ content/4/2/e004015.DOI：10.1136/bmjopen-2013-004015.
[33] THOMAS R L，HALIM S，GURUDAS S，et al. IDF Diabetes Atlas: a Review of Studies Utilising Retinal Photography on the Global Prevalence of Diabetes Related Retinopathy between 2015and 2018 [J/OL].Diabetes Research and Clinical Practice，2019，157：107840[2020-08-05].https://www.sciencedirect.com/science/article/pii/S0168822719312574.DOI:10.1016/j.diabres.2019.107840.
[34] WONG T Y，BRESSLER N M. Artificial intelligence with deep learning technology looks into diabetic retinopathy screening [J].The Journal of the American Medical Association，2016，316（22）：2366-2367.
[35] WANG Z H，ZHANG W P，SUN Y N，et al. Detection of Diabetic Macular Edema in Optical CoherenceTomography Image Using an Improved Level Set Algorithm [J].BioMed Research International，2020（1）：1-7．
[36] 中华医学会眼科学分会青光眼学组，中国医师协会眼科医师分会青光眼学组. 中国青光眼指南（2020） [J]. 中国眼科杂志，2020，56（8）：573-586.
[37] CHEN T C，ZENG A，SUN W，et al. Spectral Domain Optical Coherence Tomography in Glaucoma [J].International Ophthalmology Clinics，2008，48（4）：29-45.
[38] GREWAL D S，TANNA A P. Diagnosis of Glaucoma and Detection of Glaucoma Progression using Spectral Domain Optical Coherence Tomography [J].Current opinion in ophthalmology，2013，24（2）：150-161.
[39] KHALIL T，AKRAM M U，RAJA H，et al. Detection of Glaucoma using Cup to Disc Ratio from Spectral Domain Optical Coherence Tomography Images [J].IEEE Access，2018，6：4560-4576.
[40] RAJA H，AKRAM M U，SHAUKAT A，et al. Extraction of Retinal Layers Through Convolution Neural Network （CNN） inan OCT Image for Glaucoma Diagnosis [J].Journal of Digital Imaging，2020（33）：1428–1442.
[41] WANG Y，ZHANG Y N，YAO Z M，et al. Machine Learning Based Detection of Age-related Macular Degeneration（AMD）and Diabetic Macular Edema（DME） from Optical CoherenceTomography（OCT）images [J].Biomed Optics Express，2016，7（12）：4928–4940.
[42] 黄洁，吴星伟. 中国中医眼科杂志 [J]. 中医药通报，2009（5）：48.
[43] TREDER M，LAUERMANN J L，ETER N. Automated Detection of Exudative Age-related Macular Degeneration in Spectral Domain Optical Coherence Tomography using deep Learning [J].Graefe's Archive for Clinical and Experimental Ophthalmology，2018，256（2）：259–265.
[44] RIM T H，LEE A Y，TING D S，et al. Detection of Features Associated with Neovascular Age-related Macular Degeneration in Ethnically Distinct Data Sets by an Optical Coherence Tomography: Trained Deep Learning Algorithm [J/OL].British Journal of Ophthalmology，2020：[2020-09-20].https://bjo.bmj.com/content/early/2020/09/08/bjophthalmol-2020-316984.DOI:10.1136/bjophthalmol-2020-316984.
[45] CHEN Z L，LI D B，SHEN H L，et al. Automated Segmentation of Fluid Regions in Optical Coherence TomographyB-scan Images of Age-related Macular Degeneration [J/OL].Optics& Laser Technology，2020，122：105830[2020-10-01].https://www.sciencedirect.com/science/article/abs/pii/S0030399219301963.DOI:10.1016/j.optlastec.2019.105830.
通讯作者：
方杨（1996—），女，汉族，重庆人，硕士研究生在读，研究方向：光学成像与图像处理；
胡建明（1974—），男，汉族，重庆人，教授，博士研究生，研究方向：OCT 系统设计与光谱分析；
陈葛（1996—），男，汉族，四川眉山人，硕士研究生在读，研究方向：光学成像。