Automatic measurement of cup-disc ratio is a challenging problem. This problem could be efficiently solved with a neural network based learning technique. The eye signature renders the measurement of this ratio a tedious task. The physicians prepared maps for grading of this ratio. These maps classify the development of disease in a human eye in nine stages as shown in Figure 1. Neural networks and their solution time is very strongly dependent upon the initial random values of the synaptic connections and the order of pattern presentation to the input layer. Therefore it was necessary to study this phenomena and experiment with different distributions and show their effect on the speed of convergence. A multilayer neural network is designed and trained with the back propagation technique to assign an eye signature to its corresponding ratio. This helps the physician in the diagnosis of glaucoma. The automatic grading of cup-disc ratio offers the advantage of objective grading over thevisual inspection.
Tolba, A., & Sitteen, S. (2021). Neural Networks Applied to Grading of Cup-Disk Ratio in Glaucomatous Cases.. MEJ- Mansoura Engineering Journal, 18(4), 39-48. doi: 10.21608/bfemu.2021.166333
MLA
A. Tolba; S. A. Sitteen. "Neural Networks Applied to Grading of Cup-Disk Ratio in Glaucomatous Cases.". MEJ- Mansoura Engineering Journal, 18, 4, 2021, 39-48. doi: 10.21608/bfemu.2021.166333
HARVARD
Tolba, A., Sitteen, S. (2021). 'Neural Networks Applied to Grading of Cup-Disk Ratio in Glaucomatous Cases.', MEJ- Mansoura Engineering Journal, 18(4), pp. 39-48. doi: 10.21608/bfemu.2021.166333
VANCOUVER
Tolba, A., Sitteen, S. Neural Networks Applied to Grading of Cup-Disk Ratio in Glaucomatous Cases.. MEJ- Mansoura Engineering Journal, 2021; 18(4): 39-48. doi: 10.21608/bfemu.2021.166333
)
View on SCiNiTO