基于共聚焦显微镜图像拼接方法分析干眼患者角膜神经形态

Analysis of the Corneal Nerve in Dry Eye Patients Using the Mosaic Confocal Microscopy Image Processing Method

目的:探讨使用活体共聚焦显微镜(IVCM)角膜神经图像拼接方法分析干眼患者角膜神经形态特点及其与干眼临床指标的相关性。方法:系列病例研究。收集2021年1─5月于北京大学人民医院眼科就诊的干眼患者16例(16眼)。所有患者均进行无创伤泪河高度(NITMH)、无创伤泪膜破裂时间(NITBUT)、荧光素染色泪膜破裂时间(TBUT)、角膜荧光素染色(FL)评分、睑板腺缺失比例、基础泪液分泌试验(SⅠT)、IVCM等检查。分别使用传统方法和新的拼接图像处理方法分析患者角膜上皮下神经图像面积、神经总长度、神经密度、平均神经长度、最长神经长度、最短神经长度、神经数量、神经数量密度等指标。纳入右眼数据进行分析。2种方法间角膜神经分析的数据比较采用wilcoxon秩和检验。干眼临床指标与角膜神经分析数据的相关性采用Spearman相关性分析。结果:新的拼接图像分析方法在角膜上皮下神经图像面积、神经总长度、神经密度、平均神经长度、最长神经长度、神经数量方面均明显大于传统分析方法(均P<0.05);最短神经长度较传统分析方法短(P<0.001);神经数量密度较传统方法比较差异无统计学意义。使用传统分析方法时,NIKBUT与平均神经长度、神经数量、神经数量密度均有相关性(r=0.52,P=0.037;r=-0.62,P=0.011;r=-0.62,P=0.011),其余干眼指标与角膜神经指标均无相关性。而使用拼接图像分析方法时,NIKBUT与平均神经密度呈负相关(r=-0.56,P=0.025),其余干眼指标与角膜神经指标均无相关性。结论:相比传统分析方法,新的拼接图像分析方法可获得更大角膜神经分析面积。2种分析方法的角膜神经分析结果不同,部分干眼指标与角膜神经分析结果的相关性也不同。新的共聚焦显微镜图像拼接分析方法能够更准确、更可靠地评估干眼患者角膜上皮下神经情况。

Objective:To investigate the mosaic in-vivo confocal microscopy(IVCM)corneal nerve image processing method,and to analyze the morphology of the corneal subbasal nerves and their relationship to clinical parameters in patients with dry eye syndrome.Methods:In this case series study,16 patients(16 eyes)with dry eye syndrome were recruited from the Department of Ophthalmology,Peking University People's Hospital from January 2021 to May 2021.The non-invasive tear meniscus height(NITMH),non-invasive break up time(NITBUT),tear break up time(TBUT),corneal fluorescein staining(FL),the dropout of the meibomian gland,SchirmerⅠtest(SⅠT)and IVCM were examined.Both the traditional method and the new mosaic image processing method were used for corneal nerve morphology analysis.Measurements included area,total nerve length,nerve density,mean length,maximum length,minimum length,nerve number and nerve density,etc.The right eye of each subject was included in this study.A Wilcoxon rank sum test was used to compare the differences in the variables between the two methods,and Spearman correlation analysis was used to determine the correlation between dry eye clinical parameters and corneal nerve variables for the 2 methods.Results:The area,total nerve length,nerve density,mean length,maximum length and nerve number were all larger and longer with the new mosaic image processing method than with the traditional method(all P<0.05).The minimum length was shorter in the new mosaic image processing method(P<0.001).There was no statistically significant difference in nerve density between the two methods.Using the traditional method,NIKBUT was correlated with mean length,nerve number and nerve density(r=0.52,P=0.037;r=-0.62,P=0.011;r=-0.62,P=0.011).Other dry eye parameters were not associated with the corneal nerve analysis parameters.NIKBUT was only correlated with nerve density(r=-0.56,P=0.025)using the mosaic image processing method,while other dry eye parameters were not associated with corneal nerve analysis parameters.Conclusions:Compared to the traditional method,using the new mosaic image processing method provides a corneal assessment over a larger area.The correlations for corneal nerve analysis parameters,and some dry eye clinical parameters and corneal nerve analysis parameters are different between the two methods.The new corneal nerve analysis method can analyze a larger area,is more accurate,and more reliable for corneal subbasal nerve analysis in dry eye patients.