Abstract
Purpose:
We compared the discriminative ability between Swept source optical coherence tomography (DRI OCT) and Cirrus HD OCT in 3 different thickness measurement areas. Before verifying discriminative ability, we assessed repeatability and agreement of thickness measurement results in same areas of normal subjects.
Methods:
A cross sectional, prospective study was performed by measuring peripapillary retinal nerve fiber layer thickness (PPRNFL: average PPRNFL, 4 sectors and 12 sectors), macular thickness (MT: 9 sectors and average MT), and ganglion cell-inner plexiform layer thickness (GC-IPL: 6 sectors and average GC-IPL) in DRI OCT and Cirrus HD OCT. The Intraclass Correlation Coefficient (ICC) and the Coefficients of Variability (CV) were calculated for the measurement of repeatability and agreement. To compare the predictive power for glaucoma, the Area of under Receiver Operating Characteristic curve (AUC) was calculated and compared among three groups (C-E; control vs early glaucoma, C-MS: control vs moderate to severe glaucoma, E-MS: early vs moderate to severe glaucoma).
Results:
All ICC for the DRI OCT in all measurement sectors were above 0.9. All ICC for the Cirrus HD OCT in MT and GC-IPL were also above 0.9. Overall, DRI OCT showed high values of ICC and low values of CV. Compared with intrarater ICC, interrater ICC values were variable. Overall, the AUC for the sectors of PPRNFL in DRI OCT were higher than those in Cirrus HD OCT. In the sectors of MT, 4 sectors showed significantly high values of AUC in DRI OCT (Out N in C-E: p=0.006, In T and Out T in C-MS: p=0.015, 0.043, respectively. Out T in E-MS p=0.034). In the sectors of GC-IPL, the AUC of DRI OCT in NI was significantly higher than that of Cirrus HD OCT in C-E (p=0.011). However, the values of AUC in Cirrus HD OCT were higher than those of DRI OCT in I of C-MS (p=0.002) and NI, I, TI of E-MS (p=0.024, p=0.0002, and p=0.002, respectively).
Conclusions:
In conclusion, our study demonstrates similar or better repeatability of DRI OCT in comparison with Cirrus HD OCT. The ability of discrimination normal eyes from glaucomatous eyes or less severe glaucomatous eyes from more severe glaucomatous eyes was not insufficient comparing with Cirrus HD OCT.