New paper at MICCAI 2018!

Our paper entitled “Towards a glaucoma risk index based on simulated hemodynamics from fundus images” has been accepted for presentation in MICCAI 2018!

In this article we present one of the first attempts to characterize the hemodynamics of the retinal arterioles on glaucomatous patients. We have used a 0D model for simulating the blood flow in patient specific segmentations of the arteries, and proposed a novel dictionary-learning based strategy for summarizing all the hemodynamic outcomes into a fixed length feature vector. By means of this approach we want to perform clinical studies on large populations to understand the hemodynamics of glaucomatous patients!

We have good news for you, btw! Jointly with the manuscript, we will also release our code… and data! We have built LES-AV, a set of 22 fundus photographs taken from the Leuven Eye Study. The images include not only manual annotations of the retinal vessels, but also their classification into arteries and veins.

This work was part of a collaboration with João Barbosa Breda, Karel van Keer, Matthew B. Blaschko, Pablo J. Blanco and Carlos A. Bulant. This is also part of the work that I made after defending my thesis in Argentina, and before moving to Vienna to join OPTIMA. So, yeah, it is cool to see CONICET again in the MICCAI proceedings :)