Stromal scattering mean-free path, as a quantitative measure of corneal transparency, derived from objective analysis of depth-resolved corneal images
Résumé
Purpose : To address the unmet need for an objective means to quantify corneal transparency, we propose using the scattering mean-free path derived from objective analysis of depth-resolved corneal images, and demonstrate the feasibility of this approach by means of full-field optical coherence tomographic microscopy (FF-OCT or FF-OCM).
Methods : An algorithm and related software were developed in Matlab (Mathworks, Inc., USA) to derive the scattering mean-free path from depth-resolved corneal images. Specifically, mean amplitude depth profiles were generated to assess stromal light backscattering, that is, the mean amplitude value was calculated as a function of stromal depth. The resultant profiles were fitted to an exponential function A(z) ~ e-Bz, which allowed us to extract the scattering mean-free path, ls= 1/B, as a measure of transparency. As a proof-of-concept demonstration, this approach was applied to 3D-image stacks acquired of two corneas (one eye-bank cornea and one pathological cornea with compromised transparency, as per “gold-standard” subjective and qualitative image inspection; see Fig. 1) with ex-vivo FF-OCT (LLTech, France), after automatic 3D-image segmentation, flattening based on the epithelial surface, and extraction of the stroma.
Results : A graphical representation of the results is shown in Fig. 2. Logarithmic amplitude profiles are depicted, representing the grey value averaged over the entire field-of-view of individual en-face stromal slices as a function of depth, and illustrating the robustness of the exponential fitting procedure (linear regression line in log space) and scattering mean-free path extraction, confirmed by the very small standard deviations: ls= 180 ± 1 µm and ls= 125 ± 0 µm for the eye-bank and pathological cornea respectively.
Conclusions : We demonstrated the feasibility and robustness of deriving the scattering mean-free path, as a quantitative measure of corneal transparency, from objective analysis of stromal light backscattering with FF-OCT. This approach has the potential to supply the demand for an objective means to quantify corneal transparency in both the eye-bank and clinical setting, where such means are limited. While in-vivo development of FF-OCT is underway, the proposed analysis may already be implemented into existing depth-resolved corneal imaging methods (e.g., confocal microscopy).