Abstract
Purpose :
To assess the effects of intraocular pressure (IOP) and duration of the ex-vivo experiment on the correlation between the parameters of the gamma distribution, used to model corneal optical coherence tomography (OCT) speckle, in porcine intact eyes.
Methods :
Twenty three eyeballs were subjected to IOP from 10 to 40 mmHg with a step of 5 mmHg (Experiment 1), where a computer-controlled syringe pump allowed maintaining a stable IOP level in the anterior chamber. At each IOP level, horizontal OCT B-scans of the central 5 mm of the cornea were acquired using spectral-domain OCT (SOCT REVO). To evaluate the potential influence of duration of the experiment on the OCT speckle statistics, 10 eyeballs were maintained at the constant IOP of 15 mmHg (Experiment 2) and imaged in the same manner as in Experiment 1, to match the overall duration of the experiments (seven time-points). Spatial maps of the gamma distribution parameters, shape (v) and scale (a), were calculated from OCT scans of size 592x1592 pixels by dividing them into 21x21-pixel segments and assigning the estimated parameters to the segments’ central points. For each spatial map, the Pearson’s correlation coefficients (ρv,a) were calculated between the gamma parameters, separately for two regions of corneal stroma: central (2 mm) and non-central (the remaining part of the scan).
Results :
The group mean values of ρv,a (see Figure) differed statistically significantly in Experiment 1 (effect of IOP) for both central and non-central regions (rmANOVA, both p<0.001) and in Experiment 2 (effect of duration) for the central region only (p=0.034). Post-hoc analysis with Bonferroni correction in Experiment 2 showed no statistically significant results for any of the time-points.
Conclusions :
This study shows that variation in IOP causes significant changes in the correlation between the parameters of the gamma distribution, pointing to the potential of OCT imaging technology for visualising IOP-induced changes in corneal stroma, particularly in its central part.
This is a 2021 ARVO Annual Meeting abstract.