Matrix与Humphrey视野分析仪II和SITA的比较。

Comparison of Matrix with Humphrey Field Analyzer II with SITA.

作者信息

Fredette Marie-Josée, Giguère Anik, Anderson Douglas R, Budenz Donald L, McSoley John

机构信息

*MD †PhD ‡MD, MPH §OD, FAAO Department of Ophthalmology, Centre Universitaire d'Ophtalmologie, Université Laval, Centre Hospitalier Universitaire de Québec, Québec, le Québec, Canada (M-JF); CEVQ, SP-POS, Centre de Recherche du CHU de Québec, Québec, le Québec, Canada (M-JF, AG); Bascom Palmer Eye Institute, Department of Ophthalmology, Miller School of Medicine, University of Miami, Miami, Florida (M-JF, DRA, JMc); Office of Education and Continuing Professional Development, Department of Family and Emergency Medicine, Université Laval, Québec, le Québec, Canada (AG); and Department of Ophthalmology, University of North Carolina, Chapel Hill, North Carolina (DLB).

出版信息

Optom Vis Sci. 2015 May;92(5):527-36. doi: 10.1097/OPX.0000000000000583.

DOI:10.1097/OPX.0000000000000583

PMID:25875683

Abstract

PURPOSE

To study the performance of the Matrix perimeter compared with the Humphrey Field Analyzer II (HFA) with the Swedish Interactive Thresholding Algorithm over the range of contrast sensitivities each machine could estimate.

METHODS

Fifty stable glaucoma subjects at various stages of disease and three normal subjects had visual fields testing done on five different days within 8 weeks with both perimeters. Intraclass correlation coefficient of mean deviation, pattern standard deviation, and the SD of repeat measurements were evaluated. The repeatability of the sensitivity estimates at individual locations and global indices was quantified, as well as their dependence on disease severity. The relationship between sensitivity determinations with the two instruments was explored (principal curve analysis).

RESULTS

Mean deviation on the HFA ranged from -31 to +2.5 dB. The mean deviation and pattern standard deviation had intraclass correlation coefficients above 0.90 for both instruments. Over most of the useful range (above 20 dB on the HFA), a difference of 1 dB for the Matrix corresponded to a difference of 2 dB for the HFA. The SD of repeat measurements increased with disease severity with HFA, but not with Matrix, except that values of 12 or 34 dB were highly variable on repeat. Variability was reduced for both HFA and Matrix when duplicate sensitivity values were used. A single Matrix test provided only 15 possible sensitivity values, unevenly spaced, but the average of duplicate measurements provided more numerous sensitivity values. A learning effect was detected for Matrix.

CONCLUSIONS

The decibel values reported by the two machines are not equivalent. Variability of sensitivity determinations is affected more by the sensitivity level with HFA than with Matrix. Duplicate measurements for baseline and follow-up evaluation could be important, especially for Matrix. Further information on learning effects is needed, as is commercially available progression software for Matrix.

摘要

目的

研究Matrix视野计与采用瑞典交互式阈值算法的Humphrey视野分析仪II（HFA）在每台机器能够估计的对比度敏感度范围内的性能。

方法

50名处于不同疾病阶段的稳定青光眼患者和3名正常受试者在8周内的5个不同日期使用这两种视野计进行视野测试。评估平均偏差、模式标准差和重复测量标准差的组内相关系数。对各个位置和整体指标的敏感度估计的可重复性及其对疾病严重程度的依赖性进行量化，同时探讨两种仪器的敏感度测定之间的关系（主曲线分析）。

结果

HFA上的平均偏差范围为-31至+2.5 dB。两种仪器的平均偏差和模式标准差的组内相关系数均高于0.90。在大部分有用范围内（HFA上高于20 dB），Matrix上1 dB的差异对应于HFA上2 dB的差异。HFA的重复测量标准差随疾病严重程度增加，而Matrix则不然，只是12或34 dB的值在重复测量时变化很大。当使用重复的敏感度值时，HFA和Matrix的变异性均降低。单次Matrix测试仅提供15个可能的敏感度值，间隔不均匀，但重复测量的平均值提供了更多的敏感度值。检测到Matrix存在学习效应。