Diagnotic capabilitie of frequency-doubling technology, canning laer polarimetry, and nerve fiber layer photograph to ditinguih glaucomatou damage

PURPOE: To invetigate the ability of three diagnotic tet: frequency-doubling technology (FDT), canning laer polarimetry (GDx), and nerve fiber layer (NFL) photograph to ditinguih normal from glaucomatou eye. METHOD: Data were obtained in a cro-ectional, hopital clinic-baed tudy, including one eye from each of 253 peron older than 40 year (68 normal, 94 glaucoma upect and 91 glaucoma patient). We performed a comprehenive ocular examination, a well a tatic automated perimetry (Humphrey 24-2), creening FDT, GDx, optic nerve tereocopic photograph and high-contrat NFL photograph. REULT: The following were ignificantly different for glaucomatou patient compared with upect and normal: mean value of mean deviation (MD, Humphrey 24-2) and corrected pattern tandard deviation (CPD), 11 GDx indice, mean FDT teting time and mied point, and NFL graded defect (ANOVA, Mantel-Haenzel tet; p = 0.0001). Uing Humphrey 24-2 tet reult and clinical aement a the defining feature of glaucoma, we found that the optimal mix of enitivity and pecificity value were 84% and 100% for FDT (preence of any defect); 62% and 96% for GDx (The Number, cut-off value of 27); and, 95% and 82% for NFL photograph (preence of any abnormality). FDT teting took the leat time to be adminitered. CONCLUION: The FDT had the bet diagnotic performance. Neural network analyi of GDx data outperformed other element of it oftware.