Last few years have seen a continuous surge of AI-enabled improvements in the diagnosis of a variety of human disorders, with AI algorithms now rivaling, and in many cases outperforming trained clinicians in terms of diagnostic accuracy. In this tutorial we provide an accessible introduction to the fundamental concepts underlying developing computer-vision-based diagnostic pipelines, present successful examples of such systems implemented to diagnose a wide array of medical disorders, discuss the eccentricities of applying Machine Learning (ML) and Deep Learning (DL) algorithms in different sub-fields of medicine, and describe the challenges that remain in the way of fully integrating this technology into clinical practice.