Keratoconus Detection Algorithm using Convolutional Neural Networks: Challenges

Over the last few years, we are witnessing a development of image processing algorithms, which, alongside neuronal networks and Artificial Intelligence (A.I.) allowed their application in various medical fields. There is a great potential in having a safer, faster diagnosis, which oftentimes means saving more lives. The development of new mechanisms tailored to diagnosing keratoconus which make use of the latest machine vision technologies of a machine vision type as well as neuronal networks is of utmost necessity. The main contribution of this scientific paper lies in its analysis and study dealing with the importance of using neural networks within the field of ophthalmology, as well as in the representation of the neuronal algorithm when it comes to the detection of keratoconus. The detection algorithm needs to help the ophthalmologist by facilitating the correct diagnosis of early keratoconus, thus helping with the effective long-term management of keratoconus.