Deep learning-based optic disc classification is affected by optic-disc tilt.

We aimed to determine the effect of optic disc tilt on deep learning-based optic disc classification. A total of 2507 fundus photographs were acquired from 2236 eyes of 1809 subjects (mean age of 46 years; 53% men). Among all photographs, 1010 (40.3%) had tilted optic discs. Image annotation was performed to label pathologic changes of the optic disc (normal, glaucomatous optic disc changes, disc swelling, and disc pallor). Deep learning-based classification modeling was implemented to develop optic-disc appearance classification models with the photographs of all subjects and those with and without tilted optic discs. Regardless of deep learning algorithms, the classification models showed better overall performance when developed based on data from subjects with non-tilted discs (AUC, 0.988 ± 0.002, 0.991 ± 0.003, and 0.986 ± 0.003 for VGG16, VGG19, and DenseNet121, respectively) than when developed based on data with tilted discs (AUC, 0.924 ± 0.046, 0.928 ± 0.017, and 0.935 ± 0.008). In classification of each pathologic change, non-tilted disc models had better sensitivity and specificity than the tilted disc models. The optic disc appearance classification models developed based all-subject data demonstrated lower accuracy in patients with the appearance of tilted discs than in those with non-tilted discs. Our findings suggested the need to identify and adjust for the effect of optic disc tilt on the optic disc classification algorithm in future development.

Topographic Relationships among Deep Optic Nerve Head Parameters in Patients with Primary Open-Angle Glaucoma.

PURPOSE: To investigate the topographic relationships among the deep optic nerve head (ONH) parameters representing myopic axial elongation or changes in the lamina cribrosa (LC) in patients with primary open-angle glaucoma (POAG). METHODS: Among patients with POAG who visited the clinic between January 2015 and March 2017, the following deep ONH parameters were measured using spectral-domain optical coherence tomography (SD-OCT): externally oblique border tissue (EOBT) length, ONH tilt angle, optic canal (OC) obliqueness, and anterior LC insertion depth (ALID). In addition, the angular locations of the maximal value of each parameter were measured. We analyzed the correlations between the parameters, correlations with axial length (AL), and the spatial correspondence with glaucomatous ONH damage. RESULTS: A total of 100 eyes with POAG were included in the analysis. The EOBT length, ONH tilt angle, and OC obliqueness were correlated with each other and with AL, whereas ALID showed less correlation with the other parameters and AL. The angular location where the three AL-related parameters had maximum values was also correlated with the predominant region of the glaucomatous ONH damage, while the angular location of the deepest ALID showed less correlation. CONCLUSIONS: Among the deep ONH parameters, the AL-related parameters EOBT length, ONH tilt angle, and OC obliqueness showed strong spatial correspondence with glaucomatous ONH damage, whereas the LC-related parameter ALID was less correlated with both AL and the region with glaucomatous ONH damage. Further studies are needed to determine how these differences affect glaucomatous ONH change.