Using AI in Keratoconus Diagnosis

Medical Images: Diagnosis, Detection and Classification

Through the analysis of medical images, this project aims to diagnose eye diseases. The project is being led in collaboration with Farabi Hospital. One of the challenges facing the medical system is recognizing some eye features, such as Keratoconus, and categorizing them into normal, suspicious and KC categories. While these diagnoses are critical, they take up much time and resources from the medical system. AI and deep learning methods are being developed in this group for image classification and to assist surgeons with a more comprehensive view of medical images. To this end, a balanced dataset for the diagnosis of keratoconus is being collected, from the patients in Farabi Eye Hospital, and by using synthetic data generation by a variational autoencoder (VAE). The results of this project assist surgeons to decide whether to perform vision refractive surgeries on patients. The following keywords identify some current issues in this project to get a better understanding of the workspace: AI, Computer Vision, Image Classification, Image Analyzing, Image Processing, CNN, VAE, Python.