Clinical-Evolutionary Staging System of Primary Open-Angle Glaucoma Using Optical Coherence Tomography.

BACKGROUND

Primary open-angle glaucoma (POAG) is considered one of the main causes of blindness. Detection of POAG at early stages and classification into evolutionary stages is crucial to blindness prevention.

METHODS

1001 patients were enrolled, of whom 766 were healthy subjects and 235 were ocular hypertensive or glaucomatous patients in different stages of the disease. Spectral domain optical coherence tomography (SD-OCT) was used to determine Bruch's membrane opening-minimum rim width (BMO-MRW) and the thicknesses of peripapillary retinal nerve fibre layer (RNFL) rings with diameters of 3.0, 4.1 and 4.7 mm centred on the optic nerve. The BMO-MRW rim and RNFL rings were divided into seven sectors (G-T-TS-TI-N-NS-NI). The k-means algorithm and linear discriminant analysis were used to classify patients into disease stages.

RESULTS

We defined four glaucoma stages and provided a new model for classifying eyes into these stages, with an overall accuracy greater than 92% (88% when including healthy eyes). An online application was also implemented to predict the probability of glaucoma stage for any given eye.

CONCLUSIONS

We propose a new objective algorithm for classifying POAG into clinical-evolutionary stages using SD-OCT.