METHODS: An anonymous online survey was completed by 117 YOs in the Asia-Pacific regarding their RRD surgery experiences in 2021-2022.
RESULTS: To achieve a 90% probability of surgical competency, 91 vitrectomy and 34 scleral buckling (SB) completions during fellowship were needed. In total, 49 (41.9%) YOs had fellowship affected by COVID-19. In the COVID versus pre-COVID era, however, the volume of SB completions per fellowship year decreased significantly (median [IQR] 3.3 [1.5, 9] vs. 13 [6.5, 23]; p
METHODS: In this work, we present a bit-plane slicing (BPS) and local binary pattern (LBP) based novel approach for glaucoma diagnosis. Firstly, our approach separates the red (R), green (G), and blue (B) channels from the input color fundus image and splits the channels into bit planes. Secondly, we extract LBP based statistical features from each of the bit planes of the individual channels. Thirdly, these features from the individual channels are fed separately to three different support vector machines (SVMs) for classification. Finally, the decisions from the individual SVMs are fused at the decision level to classify the input fundus image into normal or glaucoma class.
RESULTS: Our experimental results suggest that the proposed approach is effective in discriminating normal and glaucoma cases with an accuracy of 99.30% using 10-fold cross validation.
CONCLUSIONS: The developed system is ready to be tested on large and diverse databases and can assist the ophthalmologists in their daily screening to confirm their diagnosis, thereby increasing accuracy of diagnosis.