An evaluation of image quality and accuracy of eye bank measurement of donor cornea endothelial cell density in the Specular Microscopy Ancillary Study
An evaluation of image quality and accuracy of eye bank measurement of donor cornea endothelial cell density in the Specular Microscopy Ancillary Study
复制标题
DOI：
10.1016/j.ophtha.2004.10.045
复制
发表时间：
2005-03-01
期刊：
影响因子：
13.7
通讯作者：
Cotch, MF
中科院分区：
文献类型：
--
作者：
Lass, JH;Gal, RL;Cotch, MF
目的：镜面显微镜辅助研究旨在检查供体角膜内皮镜面图像质量，比较眼库确定的中央内皮细胞密度与中央镜面显微镜读数中心确定的内皮细胞密度，并评估可能影响镜面图像质量和内皮细胞密度准确性的供体因素。设计：非随机比较试验。参与者：角膜供体中分配的供体角膜的内皮镜面图像研究方法：经过认证的读者评估供体图像质量（从一般到优秀与不可分析的可分析）并确定中央内皮细胞密度。如果阅读器之间的分级质量存在差异或者内皮细胞密度变化大于或等于 5.0%，则进行独立裁决。将平均读数中心确定的内皮细胞密度与每个眼库确定的内皮细胞密度进行比较。主要结果指标：评估图像质量和内皮细胞密度的准确性。结果：在 23 个眼库提交的 688 幅供体内皮图像中，663 幅 (96%) 可分析（优秀，40 幅 [6%]；良好，302 幅 [44%]；一般， 321 [47%]) 和 25 (4%) 无法按照阅读中心标准进行分析。原位检索和更大的上皮暴露与更高的图像质量分级相关。在 663 个可分析图像中，眼库测定的内皮细胞密度中有 434 个 (65%) 与读取中心测定的内皮细胞密度相差在 10% 以内，而 185 个 (28%) 高于 10% 以上，44 个 (7%) 低于 10% 以上。观察到保存时间较短、存在上皮缺损、后弹力层折叠、图像质量较低以及使用固定框架或中心法内皮细胞密度分析，眼库和读取中心之间的内皮细胞密度差异较大。结论：总体而言，供体内皮镜面图像质量和内皮细胞密度测定的准确性良好。然而，数据表明，应该解决可能影响图像质量和导致内皮细胞密度解释变化的因素，因为供体内皮细胞密度是评估长期角膜移植物存活的重要参数。 (C) 2005 年由美国眼科学会颁发。
Purpose: The Specular Microscopy Ancillary Study was designed to examine donor corneal endothelial specular image quality, compare the central endothelial cell density determined by eye banks with the endothelial cell density determined by a central specular microscopy reading center, and evaluate donor factors that may have an impact on specular image quality and endothelial cell density accuracy.Design: Nonrandomized comparative trial.Participants: Endothelial specular images of donor corneas assigned in the Cornea Donor Study.Methods: Certified readers assessed donor image quality (analyzable from fair to excellent vs. unanalyzable) and determined the central endothelial cell density. Independent adjudication was performed if there was a difference in the quality of grading or if the endothelial cell density varied by greater than or equal to5.0% between readers. Average reading center-determined endothelial cell density was compared with the endothelial cell density determined by each eye bank.Main Outcome Measures: Evaluation of image quality and accuracy of endothelial cell density.Results: Of 688 donor endothelial images submitted by 23 eye banks, 663 (96%) were analyzable (excellent, 40 [6%]; good, 302 [44%]; fair, 321 [47%]), and 25 (4%) were unanalyzable by reading center standards. In situ retrieval and greater epithelial exposure correlated with a higher image quality grading. The eye bank-determined endothelial cell density of 434 of the 663 (65%) analyzable images were within 10% of the endothelial cell density determined by the reading center, whereas 185 (28%) were more than 10% higher and 44 (7%) were more than 10% lower. Greater variation in endothelial cell density between the eye banks and the reading center was observed with shorter time of death to preservation, presence of an epithelial defect, folds in Descemet's membrane, lower image quality, and the use of fixed-frame or center method endothelial cell density analysis.Conclusions: Overall, donor endothelial specular image quality and accuracy of endothelial cell density determination were good. However, the data suggest that factors that may affect image quality and contribute to variation in interpretation of the endothelial cell density should be addressed, because the donor endothelial cell density is an important parameter for assessing long-term corneal graft survival. (C) 2005 by the American Academy of Ophthalmology.