Comparison of Subjective and Objective Methods of Corneoscleral Limbus Identification from Anterior Segment Optical Coherence Tomography Images

SIGNIFICANCE This study evaluates the reliability and validity of an automatic method external internal limbal points identification from anterior segment optical coherence tomography (OCT) images in comparison with manual delineation. PURPOSE The purpose this work was to evaluate repeatability precision a previously proposed compare them delineation by experienced clinicians terms limbus diameter. METHODS Optical tomograms obtained for 12 healthy volunteers without history eye diseases were analyzed. Fifteen OCT captured each patient. For all images, determined using both methods. diameters used as comparative parameter between methods under consideration. statistical analysis included mean, standard deviation, Passing-Bablok regression, Pearson correlation coefficient. RESULTS A strong linear dependence identified. While compared subjective estimates clinicians, technique overestimated diameter (bias equals 0.21 mm optometrist 0.23 ophthalmologist) slightly underestimated 0.13 0.04 ophthalmologist). showed significantly better than case comparably high recognition. CONCLUSIONS Because excellent repeatability, may be successfully estimation dynamic changes geometry eye, where large number needs processed automatically precision.