Abstract
Purpose :
To compare the diagnostic performance of ROTA for detection of RNFL defects in glaucoma between two swept-source optical coherence tomography (OCT) with different scan resolutions – Triton (512x256 pixels in 12x9mm2) and Maestro2 (512x128 pixels in 12x9mm2) (Topcon, Japan); and to (2) evaluate the inter-device and inter-observer agreements for measurements of the angular width of RNFL defects.
Methods :
47 glaucomatous and 367 healthy eyes from 232 participants with good image quality on both Triton and Maestro2 OCT were included in the study. The sensitivities and specificities of ROTA for detection of RNFL defects were determined with reference from the Cirrus OCT RNFL thickness and RNFL thickness deviation maps. The angular width (along the 3.46mm measurement circle) of the smallest RNFL defect in ROTA were measured using a custom developed software by 2 independent investigators for assessment of the inter-device and inter-observer agreements of the RNFL defect measurements.
Results :
The mean age was 62.1±9.42 years. The sensitivities of Triton and Maestro2 ROTA for detection of RNFL defects in glaucoma were 93.6% and 95.7%, respectively; the specificities were 94.8% and 92.6%, respectively. There were 43 glaucomatous eyes with RNFL defects identified from both Triton and Maestro2 ROTAs for the Bland-Altman analysis. The span of 95% limits of agreement of the angular width measurements on ROTA was 11.7 o between the two OCT devices, 12.0 o between the two observers for Triton, and 14.1o between the two observers for Maestro2. There were no significant inter-device and inter-observer differences in the angular widths of RNFL defects on ROTA (P≥0.652).
Conclusions :
The diagnostic performance for detection of RNFL defects in glaucoma was comparable between Triton and Maestro2 ROTAs. RNFL defect measurements from Triton and Maestro2 ROTA images show good inter-device and inter-observer agreements.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.