CE45 - Interfaces: mathématiques, sciences du numérique –biologie, santé

Etiological diagnosis of cardiac diseases based on echocardiographic images and clinical data – ORCHID

Submission summary

The objective of this project is to develop rigorous and explainable artificial intelligence (AI) models for the prediction of etiological diagnosis of cardiac diseases from heterogeneous inputs. For this purpose, we will create a cohort composed of 1000 patients with 4 distinct pathologies that required complementary examinations to establish a specific etiology of the cardiomyopathy. Thanks to this cohort, we will design a deep learning solution to efficiently extract the global and local deformations of the heart muscle. This temporal information is known to have great potential for characterizing cardiac diseases, but the difficulty of measuring it currently limits its use in clinical practice. The key point will be the creation of a large-scale dataset of virtual patients with myocardial motion references from which our algorithms can learn. In parallel, we will develop an AI framework dedicated to the integration of heterogeneous and complex input data for the prediction of etiological diagnosis of cardiac diseases. This formalism will be built on modern attention mechanisms in deep learning in order to model rich and fine interactions between heterogeneous input data. This will allow us to explore the interpretability of our model by clinicians. The developed solutions will be evaluated on a second cohort of 500 patients acquired in another hospital to avoid any bias. The success of this project will contribute to a better management of medical care system and a reduction in hospital costs. This project can also be seen as a pilot study to evaluate the interest and impact of AI algorithms in the management of hospitalized patients.

Project coordination

Olivier Bernard (CENTRE DE RECHERCHE EN ACQUISITION ET TRAITEMENT D'IMAGES POUR LA SANTE)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

CREATIS CENTRE DE RECHERCHE EN ACQUISITION ET TRAITEMENT D'IMAGES POUR LA SANTE
CEDRIC Conservatoire National des Arts et Métiers Paris
ISIR Institut des Systèmes Intelligents et de Robotique
DRCI Centre Hospitalier Universitaire de Caen Normandie

Help of the ANR 340,604 euros
Beginning and duration of the scientific project: November 2022 - 48 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter