Deep learning of image texture to aid diagnostics in cardiac ultrasound

This PhD project aims to apply Convolutional Neural Networks for 3D strain cardiac ultrasound images (echocardiography) and achieve predictions of diagnosis and prognosis with AI-models that can later be embedded in products for actual industrial usage. This work will be done in the context of a multidisciplinary and international project named MARCIUS, where 6 PhD fellows in 3 different locations will work in collaboration to develop a comprehensive in silico simulation platform from data acquired from real patients and the generation of virtual patients.