1. Agreement of New Automated Matched Alternation Flicker using Undilated Fundus Photography for the Detection of Glaucomatous Structural Change
- Author
-
Jung-Ju Choi, Jaehong Ahn, Marvin Lee, Seran Jang, Seungsoo Rho, and Il Suk Yun
- Subjects
Adult ,Male ,genetic structures ,Fundus Oculi ,Optic Disk ,Glaucoma ,02 engineering and technology ,Altman plot ,Diagnostic Techniques, Ophthalmological ,Fundus (eye) ,Retina ,03 medical and health sciences ,Cellular and Molecular Neuroscience ,0302 clinical medicine ,Normal tension glaucoma ,Photography ,0202 electrical engineering, electronic engineering, information engineering ,medicine ,Humans ,Alternation (formal language theory) ,Low Tension Glaucoma ,Aged ,Retrospective Studies ,medicine.diagnostic_test ,business.industry ,Flicker ,Fundus photography ,Reproducibility of Results ,Middle Aged ,medicine.disease ,eye diseases ,Sensory Systems ,Ophthalmology ,ROC Curve ,Disease Progression ,030221 ophthalmology & optometry ,Optometry ,Female ,020201 artificial intelligence & image processing ,sense organs ,business ,Follow-Up Studies - Abstract
To determine the agreement among glaucoma experts and general ophthalmologists regarding detection of glaucomatous structural changes using a new automated matched alternation flicker (AMAF) method with fundus photographs (FPs) of undilated eyes.Sixty-six pairs of FPs of normal tension glaucoma patients were collected. FPs were taken at intervals of more than 12 months. Alternating flicker images were created using a new AMAF application. In a blinded manner, two glaucoma experts and two general ophthalmologists compared the presence of glaucomatous structural changes using either the AMAF method or the side-by-side comparison method. The interobserver and intraobserver agreements were compared using the Bland and Altman plot analysis.The glaucoma experts detected more glaucoma progression using the AMAF method (average, 50.7%) compared with the side-by-side method (average, 32.5%). General ophthalmologists detected more glaucomatous progression with the AMAF method (average, 40.9%) than with the side-by-side method (average, 25.0%). The AMAF method showed fair to substantial interreader agreement (k = 0.511-0.724) and fair to perfect intrareader agreement (k = 0.631-0.943). Interreader and intrareader agreements using the AMAF method were better for the glaucoma experts compared with the general ophthalmologists.The AMAF method showed more changes in FPs than the classical side-by-side comparison method. Regarding inter- and intrareader agreements, agreement for the glaucoma experts was best using the AMAF method, but for the general ophthalmologists agreement was best using the side-by-side comparison method.
- Published
- 2016