Purpose: To compare fluid detection of autosegmented en face to 12-line radial spectral domain optical coherence tomography scan patterns in neovascular age-related macular degeneration.
Methods: Retrospective observational case series. Sixty-seven patients (94 eyes) with neovascular age-related macular degeneration underwent autosegmented en face optical coherence tomography (with associated 304-line raster scan) and 12-line radial scan patterns. Sensitivity and specificity of fluid detection for en face scan and 12-line radial scans were determined by combining radial and 304-line raster scans as a gold standard.
Results: Two hundred and fifty-eight en face and 12-line radial spectral domain optical coherence tomography scans were interpreted. Seventy-five scans (58.1%) had fluid, whereas 54 scans (41.9%) did not. En face scan pattern fluid detection sensitivity and specificity was 89.3% and 61.1%, respectively. Twelve-line radial scan pattern fluid detection sensitivity and specificity was 97.3% and 100%, respectively. The difference in fluid detection between scan patterns was statistically significant (P = 0.01). Decreased central macular thickness was associated with false-positive (P = 0.035) and false-negative (P = 0.01) fluid detection on en face scans.
Conclusion: En face optical coherence tomography alone is not as sensitive or specific as the 12-line radial scan pattern in detecting fluid in neovascular age-related macular degeneration. En face scans should be corroborated with other optical coherence tomography protocols to guide clinical decision making.