Automated quantification of RETeval™ sensor electrode position from infrared eye images: a preliminary technical report.

IF 2 4区医学 Q2 OPHTHALMOLOGY

Documenta Ophthalmologica Pub Date : 2026-10-20 DOI:10.1007/s10633-026-10153-3

Paul A Constable, Lynne Loh, Hugo Posada-Quintero, Aleksei Zhdanov, Gleb Sokolov, Mikhail Kulyabin

{"title":"Automated quantification of RETeval™ sensor electrode position from infrared eye images: a preliminary technical report.","authors":"Paul A Constable, Lynne Loh, Hugo Posada-Quintero, Aleksei Zhdanov, Gleb Sokolov, Mikhail Kulyabin","doi":"10.1007/s10633-026-10153-3","DOIUrl":null,"url":null,"abstract":"Purpose: To present an automated method for computing the position of the RETeval skin sensor electrode and the iris color ratio (ICratio) from infrared eye images extracted from RETeval recording files.Methods: Infrared eye images were extracted from .RFF binary files recorded in 19 healthy young participants (mean age 10.6 years, range 5.0-15.8; 7 male, 12 female) using the RETeval 6-step Light First Troland ISCEV protocol. A multilabel deep learning segmentation model (SegFormer) was trained to delineate six ocular structures: pupil, iris/limbus, corneal reflex, visible white bulbar area, electrode, and inner lens ring. Electrode position was quantified as the signed displacement vector from the corneal reflex centroid to the electrode arc-contact point (Pₑ), decomposed into horizontal (dx) and vertical (dy) components in millimeters. The iris color ratio (ICratio) was derived from the segmentation masks as the mean gray level of the iris annular ring divided by the median gray level of the dark pupil, with corneal reflex pixels excluded.Results: The segmentation model achieved a mean validation Dice coefficient of 0.971 across the six classes on a held-out recording session. Infrared images (n = 271) were extracted from 19 participants (38 eyes); 67 (24.7%) satisfied all three quality-control criteria, and an electrode measurement was obtained for all 38 eyes. Vertical electrode displacement (dy) was 8.57 ± 1.81 mm (right eye) and 9.14 ± 1.89 mm (left eye; p = 0.13 between eyes); horizontal displacement (dx) was - 3.11 ± 2.42 mm and - 3.70 ± 4.74 mm, respectively (p = 0.67). Euclidean displacement (|d|) differed significantly between eyes (9.45 ± 1.63 vs 10.74 ± 2.59 mm; p = 0.004), while angular displacement did not (p = 0.81). Mask-based ICratio was 1.128 ± 0.036 (right) and 1.121 ± 0.039 (left), with a significant interocular correlation (r = 0.811, p = 0.008); the device-derived ratio was 1.134 ± 0.024 and 1.149 ± 0.038 (interocular r = 0.561, p = 0.12). The two methods differed by - 0.016 on average (95% limits of agreement - 0.108 to + 0.077; p = 0.14).Conclusions: This preliminary study provides an automated method for determining the position of the RETeval sensor electrode from the infrared images recorded by the device. A measure of iris darkness derived from the same segmentation is reported as a secondary technical output.","PeriodicalId":11207,"journal":{"name":"Documenta Ophthalmologica","volume":" ","pages":""},"PeriodicalIF":2.0000,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Documenta Ophthalmologica","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s10633-026-10153-3","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"OPHTHALMOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Purpose: To present an automated method for computing the position of the RETeval skin sensor electrode and the iris color ratio (IC_ratio) from infrared eye images extracted from RETeval recording files.

Methods: Infrared eye images were extracted from .RFF binary files recorded in 19 healthy young participants (mean age 10.6 years, range 5.0-15.8; 7 male, 12 female) using the RETeval 6-step Light First Troland ISCEV protocol. A multilabel deep learning segmentation model (SegFormer) was trained to delineate six ocular structures: pupil, iris/limbus, corneal reflex, visible white bulbar area, electrode, and inner lens ring. Electrode position was quantified as the signed displacement vector from the corneal reflex centroid to the electrode arc-contact point (Pₑ), decomposed into horizontal (dx) and vertical (dy) components in millimeters. The iris color ratio (IC_ratio) was derived from the segmentation masks as the mean gray level of the iris annular ring divided by the median gray level of the dark pupil, with corneal reflex pixels excluded.

Results: The segmentation model achieved a mean validation Dice coefficient of 0.971 across the six classes on a held-out recording session. Infrared images (n = 271) were extracted from 19 participants (38 eyes); 67 (24.7%) satisfied all three quality-control criteria, and an electrode measurement was obtained for all 38 eyes. Vertical electrode displacement (dy) was 8.57 ± 1.81 mm (right eye) and 9.14 ± 1.89 mm (left eye; p = 0.13 between eyes); horizontal displacement (dx) was - 3.11 ± 2.42 mm and - 3.70 ± 4.74 mm, respectively (p = 0.67). Euclidean displacement (|d|) differed significantly between eyes (9.45 ± 1.63 vs 10.74 ± 2.59 mm; p = 0.004), while angular displacement did not (p = 0.81). Mask-based IC_ratio was 1.128 ± 0.036 (right) and 1.121 ± 0.039 (left), with a significant interocular correlation (r = 0.811, p = 0.008); the device-derived ratio was 1.134 ± 0.024 and 1.149 ± 0.038 (interocular r = 0.561, p = 0.12). The two methods differed by - 0.016 on average (95% limits of agreement - 0.108 to + 0.077; p = 0.14).

Conclusions: This preliminary study provides an automated method for determining the position of the RETeval sensor electrode from the infrared images recorded by the device. A measure of iris darkness derived from the same segmentation is reported as a secondary technical output.

查看原文本刊更多论文

红外眼图像中RETeval™传感器电极位置的自动量化：初步技术报告。

目的：建立一种自动计算RETeval皮肤传感器电极位置和虹膜色比（ICratio）的方法。方法：提取人眼红外图像。使用RETeval 6步Light First Troland ISCEV方案记录了19名健康年轻参与者（平均年龄10.6岁，范围5.0-15.8岁；7名男性，12名女性）的RFF二进制文件。采用多标签深度学习分割模型（SegFormer）对瞳孔、虹膜/角膜缘、角膜反射、可见白色球区、电极和晶状体内环等6个眼部结构进行了分割。电极位置被量化为从角膜反射质心到电极弧接触点（Pₑ）的符号位移矢量，分解为以毫米为单位的水平（dx）和垂直（dy）分量。从分割掩模中得到虹膜色比（ICratio），即虹膜环的平均灰度值除以暗瞳孔的灰度中值，剔除角膜反射像素。结果：分割模型在一个保留的记录会话中实现了六个类的平均验证Dice系数为0.971。从19名参与者（38只眼睛）中提取红外图像（n = 271）；67例（24.7%）满足所有三个质量控制标准，所有38只眼睛均进行了电极测量。电极垂直位移（dy）右眼为8.57±1.81 mm，左眼为9.14±1.89 mm，两眼间p = 0.13；水平位移（dx）分别为- 3.11±2.42 mm和- 3.70±4.74 mm （p = 0.67）。眼间欧氏位移（|d|）差异显著（9.45±1.63 vs 10.74±2.59 mm; p = 0.004），而角位移无差异（p = 0.81）。基于眼罩的ICratio分别为1.128±0.036（右）和1.121±0.039（左），眼间相关性显著（r = 0.811, p = 0.008）；器械衍生比值分别为1.134±0.024和1.149±0.038（眼间r = 0.561, p = 0.12）。两种方法的平均差异为- 0.016（95%一致限为- 0.108 ~ + 0.077;p = 0.14）。结论：本初步研究提供了一种从设备记录的红外图像中自动确定RETeval传感器电极位置的方法。虹膜暗的测量从相同的分割被报道为二级技术输出。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Documenta Ophthalmologica 医学-眼科学

CiteScore

3.50

自引率

21.40%

发文量

审稿时长

>12 weeks

期刊介绍： Documenta Ophthalmologica is an official publication of the International Society for Clinical Electrophysiology of Vision. The purpose of the journal is to promote the understanding and application of clinical electrophysiology of vision. Documenta Ophthalmologica will publish reviews, research articles, technical notes, brief reports and case studies which inform the readers about basic and clinical sciences related to visual electrodiagnosis and means to improve diagnosis and clinical management of patients using visual electrophysiology. Studies may involve animals or humans. In either case appropriate care must be taken to follow the Declaration of Helsinki for human subject or appropriate humane standards of animal care (e.g., the ARVO standards on Animal Care and Use).