人工智能在角膜相关疾病领域的应用研究

狄宇; 李莹

doi:10.12290/xhyxzz.2020-0098

人工智能在角膜相关疾病领域的应用研究

doi: 10.12290/xhyxzz.2020-0098

狄宇,
李莹^,

基金项目:

白求恩·干眼诊疗与研究科研项目 BJ-GY2021015J

详细信息

通讯作者:
李莹电话：010-69152733，E-mail: liyingpumch@126.com

中图分类号: R770.4; TP18
计量
- 文章访问数: 65
- HTML全文浏览量: 146
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2020-12-27
- 录用日期: 2021-01-11
- 网络出版日期: 2021-11-26
- 发布日期: 2021-09-01
- 刊出日期: 2021-09-30

The Application and Research Progress of Artificial Intelligence in Corneal Related Diseases

DI Yu,
LI Ying^,

Funds:

Bethune·Dry eye Diagnosis and Treatment Research Project BJ-GY2021015J

More Information

Corresponding author: LI Ying Tel: 86-10-69152733, E-mail: liyingpumch@126.com

摘要

摘要: 人工智能(artificial intelligence，AI)是计算机领域的前沿科学，近年来在众多领域发展迅猛，其在眼科的研究和应用也日益增多。AI在角膜相关疾病领域的研究主要包括圆锥角膜的早期诊断及分级、角膜屈光手术相关评估、感染性角膜炎的分类及程度判断、角膜移植术后再干预的评估等，主要采用的算法包括神经网络、支持向量机及决策树，模型的灵敏度和特异度均达90%以上。AI可为医生提供客观的临床决策、为患者提供精准的治疗奠定基础。本文将对近年来AI在角膜相关疾病领域的应用研究进行综述。
- 人工智能 /
- 角膜疾病 /
- 圆锥角膜 /
- 屈光手术
Abstract: Artificial intelligence(AI)is the frontier of computer science. In recent years, AI has developed rapidly in many fields, and its research in ophthalmology is also increasing. The research of AI in corneal related diseases mainly includes the early diagnosis and grading of keratoconus, preoperative evaluation of corneal refractive surgery, prediction of surgical parameters, judgment of the classification and degree of infectious keratitis, evaluation of reintervention after corneal transplantation, auxiliary detection of corneal nerve endings in diabetic peripheral neuropathy, and screening of pterygium. Through the neural network, the support vector machine, and the decision tree, the sensitivity and specificity of the model can reach more than 90%. AI can provide objective clinical decision-making for clinicians and precise clinical treatments for patients. This article reviews the research of AI in corneal diseases in recent years.
- artificial intelligence /
- corneal diseases /
- keratoconus /
- refractive surgery

HTML全文

表 1 圆锥角膜相关AI研究

年份(年)	研究者	图像采集仪器	样本量(n)	分组情况	输入参数	AI算法	评价指标
年份(年)	研究者	图像采集仪器	样本量(n)	分组情况	输入参数	AI算法	AUC	灵敏度	特异度
2020	Kuo^[16]	TMS-4	326	圆锥角膜组，正常对照组	/	VGG16	0.931	0.917	0.944
						Inception	0.931	0.917	0.944
						V3	0.958	0.944	0.972
2019	Kamiya^[17]	CASIA	304	Ⅰ~Ⅳ级圆锥角膜组，正常对照组	/	ResNet152	/	1.000	0.984
2019	Lavric^[18]	Pentacam	1350	圆锥角膜组，正常对照组	/	CNN	0.993	/	/
2019	Issarti^[19]	Pentacam	838	中重度圆锥角膜组，可疑圆锥角膜组，正常对照组	/	FNN	0.966	0.956	0.978
2018	Yousefi^[20]	CASIA	3156	Ⅰ~Ⅳ级圆锥角膜组，正常对照组	420个	非监督ML	/	0.977	0.941
2017	Hidalgo^[21]	Pentacam	135	圆锥角膜组，角膜屈光术后组，正常对照组	22个	CNN	0.989	0.991	0.985
2016	Hidalgo^[22]	Pentacam	860	圆锥角膜组，顿挫型圆锥角膜组，正常对照组	25个	SVM	0.989	0.991	/
2016	Kovács^[23]	Pentacam	135	双侧圆锥角膜组，单侧圆锥角膜组，正常对照组	15个	MLPNN	0.99	0.901.00	0.900.95
2013	Smadja^[24]	Gailei	372	圆锥角膜组，顿挫型圆锥角膜组，正常对照组	55个	决策树	/	0.995	1.00
2012	Arbelaez^[25]	Sirius	3502	圆锥角膜组，顿挫型圆锥角膜组，角膜屈光术后组，正常对照组	7个	SVM	0.982	0.95	0.993
2010	Souza^[26]	OrbscanⅡ	318	圆锥角膜组，角膜屈光术后组，正常对照组	/	SVM	0.99	1.00	1.00
						MLPNN	0.99	1.00	1.00
						RBFNN	0.99	0.98	0.98
2005	Twa^[27]	Keratron	244	圆锥角膜组，正常对照组	/	决策树	0.93	0.93	0.92
2002	Accardo^[28]	EyeSys	396	圆锥角膜组，正常对照组	9个	CNN	0.967	0.976	0.941
1997	Smolek^[29]	TMS-1	300	圆锥角膜组，可疑圆锥角膜组	10个	CNN	1.00	1.00	1.00
AI：人工智能；AUC：曲线下面积；CNN：卷积神经网络；FNN：前馈神经网络；ML：机器学习；SVM：支持向量机；MLPNN：多层感知器神经网络；RBFNN：径向基函数神经网络

下载: 导出CSV

参考文献(48)

[1]	Rahimy E. Deep learning application in ophthalmology[J]. Curr Opin Ophthalmol, 2018, 29: 254-260. doi: 10.1097/ICU.0000000000000470
[2]	Lawrence DR, Palacios-González C, Harris J. Artificial intelligence[J]. Camb Q Healthc Ehics, 2016, 25: 250-261. doi: 10.1017/S0963180115000559
[3]	陈有信, 张碧磊, 张弘哲. 眼科人工智能技术的现状与问题[J]. 中华眼底病杂志, 2019, 35: 119-123. Chen YX, Zhang BL, Zhang HZ. Insights and prospectives of ophthalmologic artificial intelligence technology[J]. Zhonghua Yandibing Zazhi, 2019, 35: 119-123.
[4]	Gulshan V, Peng L, Coram M, et al. Development and validation of deep learning algorithm for detection of retinopathy in retinal fundus photographs[J]. JAMA, 2016, 316: 2402-2410. doi: 10.1001/jama.2016.17216
[5]	Burlina PM, Joshi N, Pekala M, et al. Automated grading of age-related macular degeneration from color fundus images using deep convolutional neural network[J]. JAMA Ophthalmol, 2017, 135: 1170-1176. doi: 10.1001/jamaophthalmol.2017.3782
[6]	Wu X, Huang Y, Liu Z, et al. Universal artificial intelligence platform for collaborative management of cataracts[J]. Br J Ophthalmol, 2019, 103: 1553-1560. doi: 10.1136/bjophthalmol-2019-314729
[7]	Asaoka R, Murata H, Iwase A, et al. Detecting preperimetric glaucoma with standard automated perimetry using a deep learning classifier[J]. Ophthalmology, 2016, 123: 1974-1980. doi: 10.1016/j.ophtha.2016.05.029
[8]	Wu XH, Liu L, Zhao L, et al. Application of artificial intelligence in anterior segment ophthalmic disease: diversity and standardization[J]. Ann Transl Med, 2020, 8: 714. doi: 10.21037/atm-20-976
[9]	Mahesh Kumar SV, Gunasundari R. Computer-aided diagnosis of anterior segment eye abnormalities using visible wavelength image analysis based machine learning[J]. J Med Syst, 2018, 42: 128. doi: 10.1007/s10916-018-0980-z
[10]	Long EP, Lin HT, Liu ZZ, et al. An artificial intelligence platform for the multihospital collaborative management of congenital cataracts[J]. Nat Biomed Eng, 2017, 1: 0024. doi: 10.1038/s41551-016-0024
[11]	Fu H, Baskaran M, Xu Y, et al. A deep learning system for automated angle-closure detection in anterior segment optical coherence tomography images[J]. Am J Ophthalmol, 2019, 203: 37-45. doi: 10.1016/j.ajo.2019.02.028
[12]	Aloudat M, Faezipour M, El-Sayed A. High intraocular pressure detection from frontal eye images: a machine lean-ing based approach[J]. Annu Int Conf IEEE Eng Med Biol Soc, 2018: 5406-5409. http://www.ncbi.nlm.nih.gov/pubmed/30441559
[13]	Godefrooij DA, de Wit GA, Uiterwaal CS, et al. Age-Specific incidence and prevalence of keratoconus: a nationwide registration study[J]. Am J Ophthalmol, 2017, 175: 169-172. doi: 10.1016/j.ajo.2016.12.015
[14]	de Sanctis U, Loiacono C, Richiardi L, et al. Sensitivity and specificity of posterior cornea elevation measure by Pentacam in discriminating keratoconus/subclinical keratoconus[J]. Ophthalmology, 2008, 115: 1534-1539. doi: 10.1016/j.ophtha.2008.02.020
[15]	Gordon-Shaag A, Millodot M, Ifrah R, et al. aberrations and tomography in normal, keratoconus-suspect, and keratoconic eyes[J]. Optom Vis Sci, 2012, 89: 411-418. doi: 10.1097/OPX.0b013e318249d727
[16]	Kuo BI, Chang WY, Liao TS, et al. Keratoconus screening based on deep learning approach of corneal topography[J]. Transl Vis Sci Technol, 2020, 25: 53. http://www.researchgate.net/publication/345453287_Keratoconus_Screening_Based_on_Deep_Learning_Approach_of_Corneal_Topography
[17]	Kamiya K, Ayatsuka Y, Kato Y, et al. Keratoconus detection using deep learning of colour-coded maps with anterior segment optical coherence tomography: a diagnostic accuracy study[J]. BMJ Open, 2019, 9: e021313. http://bmjopen.bmj.com/content/9/9/e031313.full
[18]	Lavric A, Valentin P. KeratoDetect: keratoconus detection algorithm using Convolutional neural networks[J]. Comput Intell Neurosci, 2019: 8162567. http://downloads.hindawi.com/journals/cin/2019/8162567.pdf
[19]	Issarti I, Consejo A, Jiménez-García M, et al. Computer aided diagnosis for suspect keratoconus detection[J]. Comput Biol Med, 2019, 109: 33-42. doi: 10.1016/j.compbiomed.2019.04.024
[20]	Yousefi S, Yousefi E, Takahashi H, et al. Keratoconus severity identification using unsupervised machine learning[J]. PLoS One, 2018, 13: e0205998. doi: 10.1371/journal.pone.0205998
[21]	Hidalgo IR, Rozema JJ, Saad A, et al. Validation of an objective keratoconus detection system implemented in a scheimpflug Tomographer and comparison with other methods[J]. Cornea, 2017, 36: 689-695. doi: 10.1097/ICO.0000000000001194
[22]	Hidalgo IR, Rodrigues P, Rozema JJ, et al. Evaluation of a Machine-Learning classifier for keratoconus detection based on scheimpflug tomography[J]. Cornea, 2016, 35: 827-832. doi: 10.1097/ICO.0000000000000834
[23]	Kovács I, Miháltz K, Kránitz K, et al. Accuracy of machine learning classifiers using bilateral data from a scheimpflug camera for identifying eyes with preclinical signs of keratoconus[J]. J Cataract Refract Surg, 2016, 42: 275-283. doi: 10.1016/j.jcrs.2015.09.020
[24]	Smadja D, Touboul D, Cohen A, et al. Detection of subclinical keratoconus using an automated decision tree classification[J]. Am J Ophthalmol, 2013, 156: 237-246. doi: 10.1016/j.ajo.2013.03.034
[25]	Arbelaez MC, Versaci F, Vestri G, et al. Use of a support vector machine for keratoconus and subclinical keratoconus detection by topographic and tomographic data[J]. Ophthalmology, 2012, 119: 2231-2238. doi: 10.1016/j.ophtha.2012.06.005
[26]	Souza MB, Medeiros FM, Souza DB, et al. Evaluation of machine learning classifiers in keratoconus detection from Orbscan Ⅱ examinations[J]. Clnics(Sap Paulo), 2010, 65: 1223-1228. http://www.onacademic.com/detail/journal_1000040504536110_7e67.html
[27]	Twa M, Parthasarathy S, Cynthia R, et al. Automated decision tree classification of corneal shape[J]. Optom Vis Sci, 2005, 82: 1038-1046. doi: 10.1097/01.opx.0000192350.01045.6f
[28]	Accardo PA, Pensiero S. Neural network-based system for early keratoconus detection from corneal topography[J]. J Biomed Inform, 2002, 35: 151-159. doi: 10.1016/S1532-0464(02)00513-0
[29]	Smolek MK, Klyce SD. Current keratoconus detection methods compared with a neural network approach[J]. Invest Ophthalmol Vis Sci, 1997, 38: 2290-2299. http://www.researchgate.net/profile/Stephen_Klyce/publication/13885016_Current_keratoconus_detection_methods_compared_with_a_neural_network_approach/links/53f33d3d0cf2dd48950c9d4e
[30]	王雁, 李晶. 正确应对角膜屈光手术发展中的问题及挑战[J]. 中华眼科杂志, 2018, 54: 3-6. Wang Y, Li J. Problems and challenges in the development of corneal refractive surgery[J]. Zhonghua Yanke Zazhi, 2018, 54: 3-6.
[31]	Sayegh FN. Age and refractive in 46, 000 patients as a potential predictor of refractive stability after refractive surgery[J]. J Refract Surg, 2009, 25: 747-751. doi: 10.3928/1081597X-20090707-10
[32]	Saad A, Gatinel D. Combining Placido and corneal wavefront data for the detection of forme fruste keratoconus[J]. J Refract Surg, 2016, 32: 510-516. doi: 10.3928/1081597X-20160523-01
[33]	Lopes BT, Ramos IC, Salomão MQ, et al. Enhanced tomographic assessment to detect corneal ectasia based on artificial intelligence[J]. Am J Ophthalmol, 2018, 195: 223-232. doi: 10.1016/j.ajo.2018.08.005
[34]	Yoo TK, Ryu IH, Lee GY, et al. Adopting machine learn-ing to automatically identify candidate patients for corneal refractive surgery[J]. NPJ Digit Med, 2019, 2: 59. doi: 10.1038/s41746-019-0135-8
[35]	Cui T, Wang Y, Ji SF, et al. Applying Machine Learning Techniques in Prediction and Analysis for SMILE Treatment[J]. Am J Ophthalmol, 2019, 210: 71-77. http://www.sciencedirect.com/science/article/pii/S0002939419305082
[36]	Ung L, Bispo PJM, Shanbhag SS, et al. The persistent dilemma of microbial keratitis: global burden, diagnosis, and antimicrobial resistance[J]. Surv Ophthalmol, 2019, 64: 255-271. doi: 10.1016/j.survophthal.2018.12.003
[37]	Saini JS, Jain AK, Kumar S, et al. Neural network approach to classify infective keratitis[J]. Curr Eye Res, 2003, 27: 111-116. doi: 10.1076/ceyr.27.2.111.15949
[38]	Wu XL, Qiu QC, Liu Z, et al. Hyphae Detection in Fungal Keratitis Images With Adaptive Robust Binary Pattern[J]. IEEE Access, 2018, 6: 13449-13460. doi: 10.1109/ACCESS.2018.2808941
[39]	Liu Z, Cao YK, Li YJ, et al. Automatic diagnosis of fungal keratitis using data augmentation and image fusion with deep convolutional neural network[J]. Comput Methods Programs Biomed, 2020, 187: 105019. doi: 10.1016/j.cmpb.2019.105019
[40]	刁玉梅, 洪晶. 角膜后弹力层内皮移植术的研究进展[J]. 中华眼科杂志, 2015, 51: 544-547. doi: 10.3760/cma.j.issn.0412-4081.2015.07.018 Diao YM, Hong J. Research advances of Descemet's membrane endothelial keratoplasty[J]. Zhonghua Yanke Zazhi, 2015, 51: 544-547. doi: 10.3760/cma.j.issn.0412-4081.2015.07.018
[41]	Treder M, Lauermann JL, Alnawaiseh M, et al. Using Deep Learning in Automated Detection of Graft Detachment in Descemet Membrane Endothelial Keratoplasty: A Pilot Study[J]. Cornea, 2019, 38: 157-161. doi: 10.1097/ICO.0000000000001776
[42]	Hayashi T, Hitoshi T, Masumoto H, et al. A Deep Learning Approach in Rebubbling After Descemet's Membrane Endothelial Keratoplasty[J]. Eye Contact Lens, 2020, 46: 121-126 doi: 10.1097/ICL.0000000000000634
[43]	Dabbah MA, Graham J, Petropoulos I, et al. Dual-Model Automatic Detection of Nerve-Fibres in Corneal Confocal Microscopy Images[J]. Med Image Comput Comput Assist Inter, 2010, 13: 300-307. http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=PMC3066470&blobtype=pdf
[44]	Chen X, Graham J, Dabbah MA, et al. An Automatic Tool for Quantification of Nerve Fibres in Corneal Confocal Microscopy Images[J]. IEEE Trans Biomed Eng, 2017, 64: 786-794. doi: 10.1109/TBME.2016.2573642
[45]	Li Q, Zhong Y, Zhang T, et al. Quantitative analysis of corneal nerve fibers in type 2 diabetics with and without diabetic peripheral neuropathy: Comparison of manual and automated assessments[J]. Diabetes Res Clin Pract, 2019, 151: 33-38. doi: 10.1016/j.diabres.2019.03.039
[46]	Lopez YP, Aguilera LR. Automatic classification of pterygium-non pterygium images using deep learning[M/OL]. (2019-09-28). [2020-12-27]. https://link.springer.com/chapter/10.1007%2F978-3-030-32040-9_40.
[47]	Zulkifley MA, Abdani SR, Zulkifley NH. Pterygium-Net: a deep learning approach to pterygium detection and localization[J]. Multimed Tools Appl, 2019, 78: 34563-34584. doi: 10.1007/s11042-019-08130-x
[48]	Wan Zaki WMD, Mat Daud M, Abdani SR, et al. Automated pterygium detection method of anterior segment photographed images[J]. Comput Methods Programs Biomed, 2018, 154: 71-78. doi: 10.1016/j.cmpb.2017.10.026