Abstract
Purpose: :
To compare the diagnostic accuracy of high definition spectral domain OCT(Cirrus HD-OCT) and scanning laser polarimetry with variable corneal compensator (GDx-VCC) to discriminate between healthy and glaucomatous eyes by means of measurement of retinal nerve fiber layer (RNFL) thickness.
Methods: :
50 eyes from 35 normal subjects (mean age: 65.8) and 70 eyes from 58 glaucomatous patients (mean age: 65.6) were enrolled. All patients underwent a full eye examination, standard achromatic perimetry (SAP) and were imaged with GDx-VCC and Cirrus HD-OCT using RNFL scanning protocol. Glaucoma was defined on the basis of SITA-24-2 visual field loss (PSD and MD p<5% and Glaucoma Hemifield Test outside normal limits) on two consecutive visual fields. Areas under receiver operating characteristic curve were calculated for each GDx-VCC and Cirrus HD-OCT indices (global and sectoral) as measures of diagnostic accuracy and Henley-McNeil method was used to compare the AROC’s of best parameters of each tecniques. Sensitivity at ≥90% fixed specificity was also evaluated.
Results: :
Average thickness (0.927), 11 o’clock segment thickness (0.911), and 12 o’clock segment thickness (0.936) on Cirrus HD-OCT had the greatest AUCs. TSNIT average (0.966) and nerve fiber layer (NFI) (0.969) provided the highest AUCs for the GDx-VCC indices. No statistically significant differences was found between the AUCs of the best parameter of Cirrus HD-OCT (12 o’clock segment thickness) and GDx-VCC (NFI) (z =0.731, p =0.465). At a fixed specificity of 90%, 12 o’clock segment thickness had the highest sensitivity (75.86, cut-off point: ≤109 µ) on Cirrus HD-OCT, lower than best parameter of GDx-VCC, NFI, showing sensitivity of 89.7%(cut-off point: ≥28).
Conclusions: :
Cirrus HD-OCT and GDx-VCC diagnostic accuracies was comparable.12 o'clock segment thickness and NFI had the best diagnostic ability respectively.RNFL thickness measurement provided by both technologies showed good performance helping glaucoma detection.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • optic nerve • retina: proximal (bipolar, amacrine, and ganglion cells)