Multi-task Deep Learning of Myocardial Blood Flow and Cardiovascular Risk Traits from PET Myocardial Perfusion Imaging

202213 citationsJournal Articlehybrid Open Access

Authors

Ming Wai Yeung · University of Groningen

Jan Walter Benjamins · University of Groningen

Remco J.J. Knol · Medisch Centrum Alkmaar

Friso M. van der Zant · Medisch Centrum Alkmaar

Folkert W. Asselbergs · Utrecht University

Pim van der Harst · University of Groningen

Luis Eduardo Juárez‐Orozco

Abstract

Multi-task DL for the evaluation of quantitative PET myocardial perfusion polar maps is able to identify an impaired MFR as well as cardiovascular risk traits such as sex, smoking and possibly diabetes at the individual-patient level.

Topics & Keywords

Cardiac Imaging and Diagnostics Radiomics and Machine Learning in Medical Imaging Medical Imaging Techniques and Applications

UN Sustainable Development Goals

Good health and well-being

Publication Details

Published in: Journal of Nuclear Cardiology

Volume 29, Issue 6, pp. 3300-3310

DOI: 10.1007/s12350-022-02920-x

Field-Weighted Citation Impact: 1.58