Correlation between human expert macular fluid height assessment and fluid volume quantification in neovascular...

To investigate the association between manually measured retinal fluid heights and AI-quantified retinal fluid volumes and to explore disease activity indicated by fluid volume distributions in neovascular age-related macular degeneration (nAMD) using an approved AI-based algorithm. This retrospective study analyzed baseline OCT data from a multicenter nAMD cohort. Manually measured maximum macular fluid heights on B-scans in the central millimeter (CMM) for intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED) were compared with vertical fluid heights and three-dimensional volumes obtained by automated quantification using an AI-based tool (RetInSight Fluid Monitor Version 2). Out of 890 eyes, IRF was identified in the CMM in 328 eyes, SRF in 502 eyes, and PED in 705 eyes. The correlation between manual and AI-based height measurements was strong for IRF (r = 0.87) and PED (r = 0.91), and moderate for SRF (r = 0.67). Manual height-AI volume correlations within the CMM were strong for IRF (r = 0.76) and PED (r = 0.87), and moderate for SRF (r = 0.55). These correlations decreased when total fluid volumes within the central 6 mm were compared with manual CMM height, indicating that CMM height does not represent total nAMD disease activity. In conclusion, AI-derived and manual height measurements showed strong agreement for IRF and PED and moderate agreement for SRF. Maximum fluid height did not reliably reflect overall fluid volume or spatial distribution. AI-based retinal fluid assessment enables quantitative, whole-volume evaluation of retinal fluid and provides a measure of nAMD disease activity.