Abstract
Purpose: :
To compare the ability to discriminate between healthy and glaucomatous eyes of three optical imaging devices: Heidelberg retina tomography (HRT–II), Optical coherence tomography (Stratus OCT 3000) and Laser polarimetry (GDx VCC).
Methods: :
We included a total of 140 eyes from 140 subjects, classified in two groups (healthy or glaucomatous) according to intraocular pressure and standard automated perimetry. All the subjects underwent complete ophthalmic examination, including HRT–II, OCT and GDx VCC evaluations. Receiver operating characteristic (ROC) curves were plotted, areas under the ROC curves (AUCs) and sensitivities at fixed especificities (85% and 95%) were calculated for the parameters provided by the three techniques.
Results: :
The study sample included 66 healthy subjects and 74 glaucomatous patients. The best parameters discriminating between healthy and glaucomatous eyes were HRT linear discriminant function FSM (AUC = 0.899), OCT retinal nerve fiber layer average thickness (AUC = 0.945) and GDx VCC nerve fiber indicator (AUC = 0.879). No statistically significant differences were found between the AUCs for these parameters. However, sensitivities at fixed especificities were higher in OCT and HRT than GDx VCC parameters.
Conclusions: :
Structural parameters assessed by the optical imaging devices evaluated in this study are useful to discriminate glaucomatous damage, showing no significant difference among the best parameters of HRT, OCT or GDx VCC.
Keywords: optic disc • imaging/image analysis: clinical