Abstract
Purpose: :
To evaluate the repeatability of a Time Domain OCT (STRATUS 3, Carl Zeiss Meditec, CA) and two different Spectral-Domain OCTs (CIRRUS, Carl Zeiss Meditec, CA) and SPECTRALIS HRA+OCT, Heidelberg Engineering, Germany) in diabetic patients affected by central macular edema.
Methods: :
45 eyes from 45 diabetic patients were analysed with the three different OCTs. Patients were enrolled in order to obtain five groups stratified by visual acuity. Group 1: AV >0.8; group 2: 0.6<AV<0.8; group 3 0.4<AV<0.6; group 4: 0.2<AV<0.4, group 5: AV<0.2. Measurements repeatability for each instrument was evaluated through two macular volume analysis acquired during the same visit in two different sessions. The analysis was performed on the thickness values of the five ETDRS grid central sectors.Using average thickness as a covariate, the differences between the two measures were compared with ANOVA in relation to the instrument and the visual acuity.
Results: :
Statistical analysis demonstrated the instrument, the thickness average and the interaction between them to be able to influence the measures repeatability.All the devices showed to be strongly influenced (all p values < 0,001) by retinal thickness with minimal differences among the five sectors (repeatability decreased with increasing of thickness values). SD-OCTs revealed to be more repeatable than Time domain OCTs. In particular Spectralis HRA+OCT showed minimal differences between the two measures and it demonstrated to be less influenced by thickness increasing.
Conclusions: :
These data suggest macular thickness to be the most important source of variation influencing OCT measurements repeatability.Spectral Domain technology revealed to be determinant in increasing repeatability of the instruments, probably due to the more elevated number of scans for area combined with a higher speed acquisition. Eye tracking system (SPECTRALIS HRA+OCT) appeared to be a useful tool in making the device more resistant to the negative influence of thickness increasing.
Keywords: diabetic retinopathy • edema • image processing