Hypertension and diabetes are amongst the main emerging pathology in the
developed world. In the USA alone there are 50 millions of hypertensive and
15 millions of diabetic patients. Both pathologies affect various anatomical
districts, and one of the target organs is the retina: in the USA the
leading cause of blindness in working age is diabetic retinopathy. So far,
the grading of retinopathy has been based on a subjective evaluation of
images of the fundus oculi. This situation leads to a serious difficulty
when comparing evaluations of different ophthalmologists, and to the
impossibility of assessing the pathology progression in time, and therefore
the follow-up of the patients during and after a therapy.
The talk will address the development of a tool able to provide an automatic
evaluation of images of the retina, with a sufficient sensitivity and
specificity to detect even minor changes in the state of the retina.