Axial length correction in retinal thickness measurements in diabetic retinopathy

To use the RS-1 Glauvas (NIDEK) system to evaluate the effect of axial length (AL) correction on retinal thickness (RT) with optical coherence tomography (OCT).

Observational, cross-sectional study.

Between November 2023 and April 2024, 165 eyes from patients with diabetes underwent OCT imaging. RT was measured with and without AL-based magnification correction with the Early Treatment Diabetic Retinopathy Study (ETDRS) grid, which comprises three concentric circles (diameters; 1, 3, and 6 mm), with the inner and outer rings divided into superior, temporal, nasal, and inferior sectors. Measurements were calculated using two settings: a fixed AL of 24 mm (uncorrected) and the individually measured AL for each eye (corrected). The difference between uncorrected and corrected RT measurements was analyzed using Bland-Altman analysis. Linear regression analysis was performed to assess the association between AL and the measurement difference.

The mean AL was 23.9 ± 1.2 mm. Bland-Altman analysis for central subfield thickness (CST) showed a small mean difference between corrected and uncorrected measurements (bias -0.34 µm; 95% limits of agreement -6.32 to 5.63 µm). Linear regression demonstrated a significant positive association between AL and the difference in CST measurements (R² = 0.70, p < 0.0001), indicating that uncorrected measurements tended to underestimate and overestimate RT in eyes with shorter and longer AL, respectively.

AL differences affect the effective OCT scan area and may influence RT measurements. Applying AL correction minimizes measurement bias caused by transverse magnification and improves the accuracy of RT assessment.