Cox M J, Wood I C
Department of Physiological Sciences, Medical School, University of Newcastle upon Tyne, England.
Optom Vis Sci. 1993 May;70(5):421-32.
A method is described to allow an IBM PC/AT compatible microcomputer program to compute a score from the visual field data collected from a standard Friedmann Visual Field Analyzer (FVFA) Mk II testing procedure. The method could, in principle, be applied to any visual field data which provide quantitative information on the sensitivity of different retinal locations. The score takes into account the depth of any defects, their position in the visual field, and the clustering of defects into scotomata. Visual fields in 119 normal eyes, 82 ocular hypertensive (OH) eyes, and 75 glaucomatous eyes were assessed to provide the data necessary to implement this method. Analyzing the frequency of missed points against point luminance demonstrates that the most effective cut-off from which to regard points as defective is a miss at 0.8 log units or more above the working threshold. A statistical method of analyzing the distribution of defects following Bayesian principles is described and shows that points in the superior nasal, superior arcuate, and inferior arcuate areas provide the most information for detecting glaucoma. The program, written in Turbo Pascal v5.5, incorporates these findings in producing the score.
本文描述了一种方法，可使IBM PC/AT兼容微型计算机程序根据从标准弗里德曼视野分析仪（FVFA）Mk II测试程序收集的视野数据计算得分。原则上，该方法可应用于任何能提供不同视网膜位置敏感度定量信息的视野数据。该得分考虑了任何缺损的深度、它们在视野中的位置以及缺损聚集成暗点的情况。对119只正常眼、82只高眼压（OH）眼和75只青光眼眼的视野进行评估，以提供实施该方法所需的数据。分析漏点频率与点亮度的关系表明，将点视为有缺陷的最有效截止值是在工作阈值以上0.8对数单位或更高处出现漏点。描述了一种遵循贝叶斯原理分析缺损分布的统计方法，结果表明，鼻上、上弓形和下弓形区域的点为检测青光眼提供了最多信息。用Turbo Pascal v5.5编写的程序将这些结果纳入得分计算中。
