Related Experiment Video
Updated: Aug 30, 2026
Quantification of Vascular Parameters in Whole Mount Retinas of Mice with Non-Proliferative and Proliferative Retinopathies
Published on: March 12, 2022
Comparison of Choroidal Vascular Index Measurements in Pachychoroid Pigment Epitheliopathy using Manual ImageJ-Based
1Department of Ophthalmology, Sultan Abdülhamid Han Training and Research Hospital, University of Health Sciences, Istanbul, Turkey.
Purpose:
To compare choroidal vascular index (CVI) measurements obtained using manual ImageJ-based Niblack thresholding and protocol-standardized Otsu thresholding in patients with pachychoroid pigment epitheliopathy (PPE), employing a full-length choroidal region of interest (ROI) approach.
Methods:
This retrospective observational study included 30 eyes of 30 patients diagnosed with PPE. Enhanced depth imaging-optical coherence tomography (EDI-OCT) scans were acquired using Heidelberg Spectralis. Both horizontal and vertical fovea-centered B-scans were analyzed. The ROI was manually delineated on the OCT device using the polygon selection tool to include the entire visible choroid from the retinal pigment epithelium/Bruch's membrane complex to the choroidoscleral junction across the full scan width. CVI was calculated as the ratio of luminal area-to-total choroidal area. Manual ImageJ-based Niblack local thresholding was compared with protocol-standardized Otsu global thresholding using identical manually defined ROIs. Agreement was evaluated using paired t-tests, Pearson correlation, intraclass correlation coefficient (ICC), and Bland-Altman analysis.
Results:
A total of 60 OCT B-scans were analyzed. Mean CVI values were 0.615 ± 0.030 for ImageJ-Niblack and 0.626 ± 0.052 for Otsu thresholding. The mean intermethod difference was +0.011 ± 0.028, with no statistically significant difference between methods (P = 0.085). A strong linear association was observed (r = 0.90, P < 0.001), and ICC (2,1) demonstrated good agreement (0.76). Bland-Altman analysis showed a small mean bias (+0.012) with 95% limits of agreement from -0.049 to +0.073. Because no external reference standard or prespecified clinical equivalence margin was available, the findings support methodological consistency rather than interchangeability or accuracy equivalence.
Conclusions:
Protocol-standardized Otsu thresholding yielded CVI measurements that were directionally consistent with manual ImageJ-based Niblack analysis in PPE when a full-length choroidal ROI strategy was applied. This approach may support standardized group-level research analyses, but it should not be interpreted as establishing accuracy equivalence or full clinical interchangeability.
