Dewarping retinal and choroidal thickness maps in widefield optical coherence tomography volumes

Abstract

Purpose : Recent developments in optical coherence tomography (OCT) have demonstrated single acquisition widefield imaging with an optical field of view (FOV) of up to 90°. The purpose of this study is to evaluate the previously developed OCT volume dewarping and normal incidence method, for the measurement of retinal and choroidal thickness maps in human eyes.

Methods : Our previous work (IOVS, Vol. 61, PB00103, 2020) has demonstrated that dewarping and measuring normal to the retinal pigment epithelium layer is a crucial step when analyzing structure and shape of the retina. The mathematical model predicted the results to be dependent on the retinal curvature and to increase as a function of FOV.

In this study we compare retinal and choroidal thickness measurements of 12 healthy right eyes of 12 subjects before and after applying the dewarping and normal incidence method. For each eye we have acquired 12x12 mm OCT volumes with 3 mm scan depth using PLEX® Elite 9000 (ZEISS, Dublin, CA). Axial eye length is needed for the dewarping algorithm and is measured using IOL Master® 500 (ZEISS, Jena, Germany).

Results : The relative difference between thickness measured along an A-scan and thickness measured using the dewarping and normal incidence method ranges between 0% and 4% for 12x12 mm OCT volumes which is consistent with the mathematical prediction. Furthermore, the difference varies in magnitude for every individual eye and increases with FOV. Figure 1 shows the results for all 12 human eyes.

Conclusions : Measurement of human eyes has confirmed that the dewarping and normal incidence method results in thickness maps that are different from the thickness maps calculated along A-scans. The range of differences measured in human eyes matched the previously published mathematical prediction. The results further confirmed that the magnitude of the difference increases with FOV and is different for every individual eye.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Figure 1: Relative difference in retinal and choroidal thickness measured in 12 healthy human eyes (colored lines) in comparison to the prediction of the mathematical model (dashed black lines). The dashed lines represent the limits of a range of modeled retinal radii (r = 11 mm to r = 14 mm). The gap in data between 3.5 mm and 5.5 mm is caused by the optic nerve head where the retinal and choroidal layers are missing.