Application of morphological operators for optic nerve head segmentation in optical coherence tomography images

Optical Coherence Tomography (OCT) is an imaging modality that has several advantages (e.g. high resolution and three dimensional imaging) in comparison with other ophthalmic imaging methods. Retinal structures, such as the optic nerve head (ONH), can be examined in detail. The geometry of the ONH can offer valuable information for the detection of glaucoma, because it is affected by this disease. However, manual examinations are highly subjective. In this study we present an approach using morphological operators for the completely automated segmentation of the ONH region, which is crucial for the further examination of the geometrical parameters that indicate the presence of glaucoma. The method was applied to OCT volumes of 10 individuals (7 glaucoma patients, 3 normal subjects). The automatically extracted results were evaluated both qualitatively and quantitatively, and yielded a sensitivity of 0.89±0.09 and a specificity of 0.98±0.01.