Rao Harsha L, Kadambi Sujatha V, Mehta Pooja, Dasari Srilakshmi, Puttaiah Narendra K, Pradhan Zia S, Rao Dhanraj A S, Shetty Rohit
a Department of Glaucoma, Narayana Nethralaya, Rajajinagar , Bangalore , Karnataka , India.
Curr Eye Res. 2017 May;42(5):743-747. doi: 10.1080/02713683.2016.1238944. Epub 2016 Nov 29.
To evaluate the diagnostic ability of automated pupillography measurements in glaucoma and study the effect of inter-eye asymmetry in glaucomatous damage on the diagnostic ability.
In an observational, cross-sectional study, 47 glaucoma patients and 42 control subjects underwent automated pupillography using a commercially available device. Diagnostic abilities of the pupillary response measurements were evaluated using area under receiver operating characteristic (ROC) curves (AUC) and sensitivities at fixed specificities. Influence of inter-eye asymmetry in glaucoma [inter-eye mean deviation (MD) difference on visual fields (VF)] on the diagnostic ability of pupillography parameters was evaluated by ROC regression approach.
The AUCs of automated pupillography parameters ranged from 0.60 (amplitude score with peripheral blue stimulus) to 0.82 (amplitude score with full field white stimulus, Amp-FF-W). Sensitivity at 95% specificity ranged between 5% (amplitude score with full field blue stimulus) and 45% (amplitude score with full field green stimulus). Inter-eye MD difference significantly affected the diagnostic performance of automated pupillography parameters (p < 0.05). AUCs of Amp-FF-W at inter-eye MD difference of 0 dB, 5 dB, 10 dB and 15 dB were 0.71, 0.80, 0.87 and 0.93, respectively, according to the regression model. The corresponding sensitivities at 95% specificity were 20%, 34%, 50% and 66%, respectively.
The diagnostic abilities of even the best automated pupillography parameters were only moderate in glaucoma. The performance of these pupillography measurements in detecting glaucoma significantly increased with greater inter-eye asymmetry in the glaucomatous damage.
评估自动瞳孔测量法在青光眼诊断中的能力，并研究青光眼性损害中双眼不对称性对诊断能力的影响。
在一项观察性横断面研究中，47例青光眼患者和42例对照者使用市售设备进行自动瞳孔测量。使用受试者操作特征（ROC）曲线下面积（AUC）和固定特异性下的敏感性评估瞳孔反应测量的诊断能力。通过ROC回归方法评估青光眼双眼不对称性[视野（VF）上的双眼平均偏差（MD）差异]对瞳孔测量参数诊断能力的影响。
自动瞳孔测量参数的AUC范围为0.60（周边蓝色刺激下的幅度评分）至0.82（全视野白色刺激下的幅度评分，Amp-FF-W）。95%特异性下的敏感性介于5%（全视野蓝色刺激下的幅度评分）和45%（全视野绿色刺激下的幅度评分）之间。双眼MD差异显著影响自动瞳孔测量参数的诊断性能（p<0.05）。根据回归模型，双眼MD差异为0 dB、5 dB、10 dB和15 dB时，Amp-FF-W的AUC分别为0.71、0.80、0.87和0.93。95%特异性下相应的敏感性分别为20%、34%、50%和66%。
即使是最佳的自动瞳孔测量参数在青光眼诊断中的能力也仅为中等。随着青光眼性损害中双眼不对称性增加，这些瞳孔测量法在检测青光眼方面的性能显著提高。
