3907例儿童近视早期眼底改变在不同眼轴的分布及其人工智能量化监控指标OACSTPCD

Objective To use artificial intelligence(AI)technology to make qualitative and quantitative analysis of the trace changes of the fundus in the early stage of myopia in children and to provide sensitive criteria for the early de-tection.Methods It was a cross-sectional study.The fundus and related biological parameters were collected and divided into 6 groups according to age.The ocular axial lengths were divided into 8 groups.According to the SHI'standard Chinese eye health archives and diagnostic system,11 items + eye biological parameters were collected for each patient per visit ac-cording to the designed process.SPSS 19.0(SPSS Inc,Chicago,IL,USA)was used for statistical analysis,including x2 analysis,analysis of variance(ANOVA)and multiple stepwise linear regression analysis.Results Among 3 907 children under 18 years of age,55.8%were male and 44.2%were female.Right and left eyes were 50%,respectively.The peak pe-riod of consultation was between 6～11 years old(64%).The axial length between 23.5～25.5 mm was 53.5%,over 25.5 mm was 16.9%.The vertical C/D ratio was 0.4 to 0.5(77%,P<0.05).Tessellation changes less than 0.01 were found in 66.4%,and no crescent changes in 47.68%.The axial length below 24.5 mm showed 0.01 or no tessellation changes and in-creased by more than 0.1 in the 24.5～25.5 mm group.The 2/3 PD cresent was found over 24.5 mm and the 1 PD was over 26.5 mm(P<0.05).Conclusions The 24.5 mm axis of the eye can be used as a key inflection point for monitoring myo-pia progression in school-age children.The expansion of the posterior segment of the eye and the early pathological changes begin to appear and aggravate,like tessellation and cresent,especially when the axis of the eye is more than 26.5 mm,the pathological myopia diagnosed.Therefore,tessellation 0.1 and cresent 1 PD are considered as diagnostic criteria.While tes-sellation 0.04～0.09 and cresent 2/3 PD are considered as the early stage or high-risk stage of pathological changes.