Towards a glaucoma risk index based on simulated hemodynamics from fundus images

JOSÉ IGNACIO ORLANDO; BARBOSA BREDA, JOÃO; VAN KEER, KAREL; BLASCHKO, MATTHEW B.; PABLO BLANCO; CARLOS BULANT

Granada

Conferencia; 21st INTERNATIONAL CONFERENCE ON MEDICAL IMAGE COMPUTING & COMPUTER ASSISTED INTERVENTION; 2018

MICCAI Society

Abstract. Glaucoma is the leading cause of irreversible but preventableblindness in the world. Its major treatable risk factor is the intra-ocularpressure, although other biomarkers are being explored to improve theunderstanding of the pathophysiology of the disease. It has been recentlyobserved that glaucoma induces changes in the ocular hemodynamics.However, its effects on the functional behavior of the retinal arterioleshave not been studied yet. In this paper we propose a first approach forcharacterizing those changes using computational hemodynamics. Theretinal blood flow is simulated using a 0D model for a steady, incompressiblenon Newtonian fluid in rigid domains. The simulation is performedon patient-specific arterial trees extracted from fundus images. We alsopropose a novel feature representation technique to comprise the outcomesof the simulation stage into a fixed length feature vector that canbe used for classification studies. Our experiments on a new database offundus images show that our approach is able to capture representativechanges in the hemodynamics of glaucomatous patients. Code and dataare publicly available in https://ignaciorlando.github.io