Geometric and concavity characteristics as diagnostic markers for differentiating nAMD-related MNV from pachychoroid neovascularisation.

PURPOSE: To investigate morphometric differences in macular neovascularisation (MNV) outer contours between neovascular age-related macular degeneration (nAMD) and pachychoroid neovasculopathy (PNV) using quantitative optical coherence tomography angiography (OCTA) analysis. METHODS: This retrospective cross-sectional study enrolled treatment-naïve patients with newly diagnosed MNV, classified into nAMD and PNV groups. The nAMD group was further subdivided into Type 1 MNV, Type 2 MNV, and polypoidal choroidal vasculopathy (PCV). A semi-automated pipeline integrating FIJI-based contour extraction and custom Python algorithms quantitatively analysed four MNV morphometric domains: geometric parameters, shape descriptors, concavity features, and orientation. RESULTS: Eighty-nine eyes (89 patients) were included (64 nAMD, 25 PNV). PNV patients were significantly younger (P < 0.001) with better visual acuity (P = 0.026), longer symptom duration (P < 0.001), and greater subfoveal choroidal thickness (P < 0.001). nAMD-related MNV exhibited significantly larger geometric dimensions (all P < 0.05) and deeper marginal indentations (maximum: 0.41 ± 0.33 vs. 0.21 ± 0.14 mm, P = 0.009; mean: 0.17 ± 0.12 vs. 0.10 ± 0.06 mm, P = 0.002) compared to PNV. These differences were more pronounced between PCV and PNV (all P < 0.001). ROC analysis demonstrated moderate discriminative performance for nAMD versus PNV (AUC = 0.711 and 0.723) and good performance for PCV versus PNV (AUC = 0.778 and 0.840) using equivalent diameter and mean concavity depth. CONCLUSION: Outer contour morphometry effectively differentiates MNV aetiologies, with nAMD (particularly PCV) showing larger, more lobulated patterns than PNV's compact configurations, likely reflecting distinct pathophysiological mechanisms.