Abstract
Purpose: :
To compare agreement of a novel automated alternation flicker technology (AF) and stereophotograph (SP) inspection for detection of progressive glaucomatous optic neuropathy.
Methods: :
Glaucoma patients meeting eligibility criteria were retrospectively recruited. All subjects had a minimum of 36 months follow-up, standard automated perimetry (24-2) every 4 months and annual stereoscopic disc photography. Four experienced graders (RR, DSG, NR, TK) assessed SP for optic nerve head (ONH) progression, disc hemorrhage (DH), or blood vessel (BV) movement using predefined criteria. At a separate session, the graders evaluated the photos using AF applying the same criteria. The order of eyes and visits were randomized. Visual field progression was assessed using regression analysis and defined as a slope < or = -1 dB/year for inner points (-2 dB/year for outer points) significant at P< 5%.
Results: :
Thirty-four eyes of 17 patients were included. Using SP, the overall agreement (kappa±SE) among graders for ONH change was 0.17±0.07, for DH was 0.77±0.07 and for BV movement was -0.23±0.05. Using AF, the overall agreement among graders for ONH change was 0.22±0.07, for DH was 0.48±0.07 and for BV movement was 0.15±0.07. The agreement among graders was not significantly different using SP or AF (p=0.6) for ONH change but was significantly (p=0.003) better using SP for DH and significantly (p<0.001) better using AF for BV movement. The overall agreement between visual field progression and disc progression using SP (0.10±0.05) and AF analysis (0.14±0.05 ) was similar (p=0.57). Using AF progression detection was significantly associated with BV movement for 3 graders (p=0.008 to p<0.001).
Keywords: optic nerve • imaging/image analysis: clinical