变态
内丛状层
眼科
视网膜前膜
内核层
医学
神经节细胞层
神经纤维层
接收机工作特性
外丛状层
神经节
视网膜
视力
内科学
解剖
玻璃体切除术
作者
Wenjia Yan,Yuntong Li,Yanqiao Huang,Qiong Wang,Shuya Ke,Miner Yuan,Xiaoyan Ding
标识
DOI:10.1136/bjo-2025-327424
摘要
Aims To investigate the relationship between retinal structure symmetry and metamorphopsia in idiopathic epiretinal membrane (ERM) patients. Methods 181 consecutive ERM patients were divided into four groups based on vertical/horizontal metamorphopsia (MV/MH) presence. Clinical data, including best-corrected visual acuity (BCVA), metamorphopsia scores and optical coherence tomography (OCT)-assessed macular structure, were analysed. The Asymmetry Index (ASI), defined as the thickness difference ratio between corresponding retinal layers, was introduced. Univariate and multivariate linear regression explored relationships among BCVA, metamorphopsia and OCT patterns. Receiver operating characteristic (ROC) curves evaluated the discriminatory ability of OCT parameters for metamorphopsia. Results Vertical ASI (ASI-V) of the Ganglion cell layer and Inner plexiform layer complex (GCL+IPL) layer was significantly higher in MH groups (p<0.0001), while Horizontal ASI (ASI-H) was elevated in MV groups (p=0.0001). Strong correlations were found between MH and ASI-V (p<0.0001, R 2 =0.4536), and MV and ASI-H (p<0.0001, R 2 =0.4870). Multivariate analysis showed a positive association between average metamorphopsia scores and ASI-T of the GCL+IPL layer (standardised beta=0.633, p<0.001). ROC analysis demonstrated good diagnostic accuracy of ASI-V for MH (area under the curve=0.823, Youden’s Index=0.669). Conclusions Vertical and horizontal asymmetric thickening of the GCL+IPL layer correlates with respective metamorphopsia types. The ASI of the GCL+IPL shows diagnostic superiority for metamorphopsia and may serve as a novel biomarker. These findings suggest that asymmetric thickening of the GCL+IPL layer, rather than overall thickening per se, may serve as the pathological basis for the development of metamorphopsia in ERM patients.
科研通智能强力驱动
Strongly Powered by AbleSci AI