Detecting Keratoconus and Non-Keratoconus Cases

Bys: Mohammed Saeed, Mustafa and Al-Nima, Raid

The cornea, which provides over two-thirds of the eye's focusing power, is highly sensitive to even minor curvature changes that can significantly affect vision. Keratoconus is a progressive condition characterized by corneal thinning, leading to myopia and irregular astigmatism, making early detection essential for effective management. This book introduces a specialized medical system for early keratoconus detection using a newly developed dataset, the Real Clinical Dataset for Keratoconus (RCDK), which includes multiple corneal maps such as axial/sagittal curvature, thickness, and elevation. Advanced preprocessing techniques are applied to enhance image quality. An Artificial Intelligence approach based on Deep Learning models, is used to analyze these maps. Each model processes a specific map, and a fusion technique combines their outputs to improve diagnostic accuracy. Using 704 images, the system achieved high individual accuracies ranging from 90% to 97.14%, while the fusion method reached 100% accuracy. The system is implemented on Raspberry Pi 4 with Python and a user-friendly interface.

• Format: Paperback
• Language: English
• ISBN: 6209724450
• ISBN13: 9786209724459
• Release Date: April 2026
• Publisher: LAP Lambert Academic Publishing
• Length: 116 Pages
• Weight: 0.36 lbs.
• Dimensions: 0.3" x 6.0" x 9.0"