Quantitative Correction for Peripheral Distortion in Ultra-Widefield Fundus Imaging Using a High-Precision 3D-Printed Model Eye

Purpose: To establish a quantitative correction database for peripheral distortion in ultra-widefield (UWF) fundus imaging across varying axial lengths (AXLs), media, and posterior staphyloma variations. Methods: High-precision three-dimensional printed model eyes were imaged under air and balanced salt solution (BSS) conditions using two commercial UWF systems (Optos 200Tx and Clarus 700). The peripheral distortion index (PDI) was quantified at predefined peripheral angles to characterize relative enlargement compared with the central region. Continuous correction models were generated using least-squares regression. Three posterior staphyloma geometries (wide macular, narrow macular, and decentered) were constructed to calculate location-specific intrastaphyloma correction factors. Results: The PDI increased nonlinearly with the peripheral angle, exceeding 1.5 at 70° and approaching 2.0 at 90° across all configurations in which, eccentricity was the dominant and AXL a minor predictor (P < 0.001). Optos 200DTx produced greater distortion than Clarus 700 under both air (1.206 ± 0.202 vs. 1.163 ± 0.171; P < 0.001) and BSS conditions (1.173 ± 0.167 vs. 1.120 ± 0.138; P < 0.001). Distortion under air was slightly but significantly greater than under BSS in both systems (Optos 200DTx, P = 0.016; Clarus 700, P < 0.001). The analytical model accurately reproduced measured distortion (R² = 0.976-0.993). For posterior staphyloma models, final distortion within staphyloma was calculated by multiplying the baseline PDI by a mean correction factor of 0.904 to 0.984. Conclusions: This study provides a comprehensive quantitative correction database across a wide range of AXL values and continuous analytical models, enabling accurate distortion correction in diverse quantitative studies using UWF imaging. Translational Relevance: A validated, ready-to-use correction database enables accurate, reproducible quantification of peripheral retinal structures across devices, axial lengths, and ocular geometries in ultra-widefield imaging.