Combination of structural and functional methods for the diagnosis and follow‐up of glaucoma:results of a two‐year‐study

Abstract

Abstract Purpose: To determine, whether an appropriate combination of the structural and the functional examination methods can improve the diagnosis and follow‐up of glaucoma Methods: One randomly selected eye of 40 patients with primary open angle glaucoma (POAG) with none or early glaucomatous visual field loss and 40 age‐matched healthy persons were included in the prospective longitudinal study. Structural evaluations were conducted using Heidelberg Retina Tomograph (HRT) and retinal nerve fiber layer (RNFL) loss scoring according to Airaksinen method. Functional parameters were reviewed by means of standard white‐on white (W/W) and blue‐on yellow (B/Y) perimetry. To assess the diagnostic value of different data sets machine learning classifiers , linear discriminant analysis and classification trees were applied. The accuracy of discrimination was described and visualized by the Receiver Operating Characteristic Curves (ROC), and the results of the first and second year of study were compared. Results: There were statistically significant differences between the healthy and glaucomatous group in the scoring of RNFL loss, HRT parameters (CA,CD,RD,RV,CSM) and mean defect of W/W perimetry. Parameters with the highest diagnostic ability obtained by ROC curves were scoring of RNFL loss, HRT analysis of the optic nerve head and W/W perimetry. Conclusions: Combination of different diagnostic methods can enhance precision of early diagnosis and follow‐up of glaucoma. A limitation for a relationship between structure and function is the individual variability of the optic disc morphology and subjective variability of visual field testing.Supported by IGA NR‐8371