先天性眼球震颤眼动波形的层次聚类分析研究  

Hierarchical clustering analysis of congenital nystagmus waveforms

作　　者：刘彦孜 史学锋[1] Liu Yanzi;Shi Xuefeng(Clinical College of Ophthalmology of Tianjin Medical University,Tianjin Eye Hospital and Institute,Tianjin Key Laboratory of Ophthalmology and Visual Science,Tianjin 300020,China;Department of Ophthalmology,Third People's Hospital of Dalian,Dalian 116091,China)

机构地区：[1]天津医科大学眼科临床学院天津市眼科医院天津市眼科研究所天津市眼科学与视觉科学重点实验室,天津300020 [2]大连市第三人民医院眼科,大连116091

出　　处：《中华眼科杂志》2022年第3期194-199,共6页Chinese Journal of Ophthalmology

基　　金：国家自然科学基金面上项目(81770956,81371049);天津市杰出青年科学基金(17JCJQJC46000);天津市131创新型人才团队项目(201936);天津市卫生计生行业高层次人才选拔培养工程津门医学英才项目。

摘　　要：目的探讨机器学习层次聚类算法用于先天性眼球震颤(CN)眼动波形自动分类和波形诊断的可行性。方法回顾性病例系列研究。收集2018年12月至2019年9月就诊于天津市眼科医院的90例(90只眼)CN患者资料,其中男性67例,女性23例,年龄(12±9)岁。所有患者采用高速视频眼动仪记录眼动波形。对标准化后的眼动波形进行无监督机器学习层次聚类分析,获得可视化分类结果并予以波形命名,统计每种波形的发生比例,分析波形成分与CN患者视功能的相关性。统计学方法为独立样本t检验和Pearson相关性分析。结果在90例(90只眼)CN患者的46620个有效眼动波形中,通过机器学习层次聚类算法自动分出7种波形,分别命名为波形Ⅰ(14259个,30.59%)、波形Ⅱ(11498个,24.66%)、波形Ⅲ(4083个,8.76%)、波形Ⅳ_(1)(5430个,11.65%)、波形Ⅳ_(2)(3451个,7.40%)、波形Ⅳ_(3)(3015个,6.47%)及波形Ⅳ_(4)(2663个,5.71%);有2221个(4.76%)波形未分类。波形Ⅰ、Ⅱ、Ⅲ分别与3种CN基本眼动波形即速度递增型冲动型、速度递减型冲动型及钟摆型波形相符,波形Ⅳ1~4为复杂波形。波形Ⅰ、Ⅱ、Ⅲ及Ⅳ_(1~4)在90例患者中的发生比例依次为78.89%(71例)、41.11%(37例)、17.78%(16例)、20.00%(18例)、7.78%(7例)、15.56%(14例)和11.11%(10例)。38例(42.22%)患者仅表现为1种眼动波形,其余52例(57.78%)同时存在2种或2种以上眼动波形,其中23例(25.56%)存在3种或3种以上眼动波形,5例(5.56%)存在4种眼动波形。患者眼动波形中波形Ⅰ所占比例与最佳矫正视力(最小分辨角对数视力)有显著相关性(r=-0.39;P<0.01),波形Ⅱ所占比例与最佳矫正视力无相关性(P>0.05)。以波形Ⅰ为主导的CN患者的最佳矫正视力(0.19±0.14)优于以波形Ⅱ为主导的CN患者(0.45±0.37),差异有统计学意义(t=2.77;P<0.05)。结论机器学习层次聚类算法可实现CN眼动波形的自动分类和波形诊断,为CN的精准诊断与评估提供Objective To explore the feasibility of applying machine-learning hierarchical clustering algorithm to waveform-type automatic classification and diagnosis in congenital nystagmus(CN).Methods A retrospective case series study.A total of 90 patients(90 eyes)diagnosed with CN at Tianjin Eye Hospital from December 2018 to September 2019 were included in the study,including 67 males and 23 females,aged(12±9)years old.Eye movement waveforms were recorded with the video eye tracker in all patients.Analyses with unsupervised machine-learning hierarchical clustering algorithm were performed on the normalized eye movement waveforms.The visualized clustering results were obtained for further waveform naming.The occurrence rate of each waveform type was calculated,and the correlation between the proportion of each waveform type and the visual function of CN patients was analyzed.Independent sample t-test and Pearson correlation analysis were used for statistical analysis.Results The 46620 cycles of validated waveforms from the 90 CN patients were categorized into 7 types of waveforms through machine-learning hierarchical clustering algorithm,named typeⅠ,typeⅡ,typeⅢ,and typesⅣ_(1-4),respectively.In the 46620 cycles of eye movement waveforms from the 90 patients with CN,there were 14259 cycles of typeⅠ(30.59%),11498 cycles of typeⅡ(24.66%),4083 cycles of typeⅢ(8.76%),5430 cycles of typeⅣ_(1)(11.65%),3451 cycles of typeⅣ_(2)(7.40%),3015 cycles of typeⅣ_(3)(6.47%),2663 cycles of typeⅣ_(4)(5.71%)and 2221 cycles of unclassified waveforms(4.76%).The waveforms of typesⅠ,ⅡandⅢcorresponded to the 3 basic CN eye movement waveforms(velocity-increasing jerk waveform,velocity-decreasing jerk waveform and pendular waveform)described in the textbooks,and the waveforms of typesⅣ1-4 were complex waveforms.The proportions of patients with the 7 types of waveforms were 78.89%(71 cases),41.11%(37 cases),17.78%(16 cases),20.00%(18 cases),7.78%(7 cases),15.56%(14 cases)and 11.11%(10 cases),respectively.According to

关 键 词：眼震 先天性 机器学习 聚类分析 眼震电图描记术 眼球运动 

分 类 号：R777.46[医药卫生—眼科]