Zhejiang Industry and Trade Vocational College, Wenzhou, Zhejiang, China.
School of Ophthalmology and Optometry, Wenzhou Medical University, Wenzhou, Zhejiang, China.
Graefes Arch Clin Exp Ophthalmol. 2021 Sep;259(9):2753-2759. doi: 10.1007/s00417-021-05220-1. Epub 2021 May 11.
To generate a model that evaluates the presence and extent of peripheral anterior synechia (PAS) based on anterior segment optical coherence tomography (AS-OCT).
The extent of PAS involvement in the eyes of patients with angle closure was assessed by indentation gonioscopy, and the part of non-PAS and PAS were assigned into two groups (NPAS and PAS). Anterior chamber angles were then imaged by AS-OCT with light-emitting diode (LED) irradiation directly into the pupils, leading to pupillary constriction and increasing anterior chamber angle width. Parameters including the angle opening distance at 750 μm anterior to the scleral spur (AOD750) and trabecular-iris space area at 750 μm anterior to the scleral spur (TISA750) were then obtained. The differences before and after LED irradiation of AOD750 and TISA750 were calculated and used to generate a PAS model based on binary logistic regression. Validation data were then tested.
A total of 258 AS-OCT images in 14 eyes were assigned to the modeling data, and 120 were assigned to the validation data. There were no differences in AOD750 and TISA750 in the dark between NPAS and PAS (P = 0.258, P = 0.486), whereas after LED light exposure, TISA750 was larger in NPAS than in PAS (P = 0.047). The light-dark differences of both parameters showed significant differences between the two groups (P = 0.019, P < 0.001). The area under the curve of the model performance was 0.841, and the overall correct rate was 80.8% based on the validation data.
The present study demonstrates that the AS-OCT-based PAS model could be useful in the identifying of the presence of synechial angle closure and evaluating the extent of PAS in a single eye.
基于眼前节光学相干断层扫描(AS-OCT)生成一种评估周边前粘连(PAS)存在和程度的模型。
通过压陷式房角镜评估闭角型青光眼患者眼内 PAS 的受累程度,并将非 PAS 和 PAS 部分分为两组(NPAS 和 PAS)。然后,使用直接照射瞳孔的发光二极管(LED)照射对前房角进行 AS-OCT 成像,导致瞳孔收缩并增加前房角宽度。获得巩膜突前 750μm 的房角开口距离(AOD750)和巩膜突前 750μm 的小梁虹膜空间面积(TISA750)等参数。然后计算 LED 照射前后 AOD750 和 TISA750 的差异,并基于二项逻辑回归生成 PAS 模型。然后测试验证数据。
共纳入 14 只眼的 258 张 AS-OCT 图像用于建模数据,120 张用于验证数据。在黑暗中,NPAS 和 PAS 之间 AOD750 和 TISA750 无差异(P=0.258,P=0.486),而在 LED 光暴露后,NPAS 中的 TISA750 大于 PAS(P=0.047)。两组之间这两个参数的光暗差异均有显著差异(P=0.019,P<0.001)。基于验证数据,模型性能的曲线下面积为 0.841,总体正确率为 80.8%。
本研究表明,基于 AS-OCT 的 PAS 模型可用于识别黏连性闭角型青光眼的存在,并评估单眼 PAS 的程度。