The objective is to develop and test a diagnostic tool for retinal dystrophy based on a database of fundus autoflourescence (FAF) images to arrive at specific disease diagnosis. The core functionality of this diagnosis tool lies in machine learning image classification techniques.
The primary inputs to the diagnostic tool are pairs of images, corresponding to patients left and right eyes, and a set of statistical-based image classifier seed objects. Image-matching is the proposed methodology though the algorithm should explore other methodologies. The output of such a tool will aid the ophthalmologists with disease diagnosis by presenting the highest likely matched images to disease-specific features.