Skip to Main content Skip to Navigation
Conference papers

Automatic regional estimation of myocardial strain using deep learning

Abstract : Although the ejection fraction remains essential in terms of diagnosis and prognosticity, it still leads to many errors and low reproducibility. Thus, quantifying local deformities of the myocardium is a crucial step of major interest for the diagnosis and follow-up of such a heart condition. The study of this quantitative parameter, in magnetic resonance imaging (MRI), has resulted in several methods. The objective of this paper is to review some relevant methods of calculating regional deformations based on deep learning such as tagged MRI, coded displacement imaging (DENSE) and strain-encoded imaging (SENC). Discussion of the results of some research has shown that monitoring segment deformation automatically makes the task more accurate and allows for better treatment planning by offering the ability to distinguish akinetic myocardial segments that have permanently lost their contractility from those likely to recover their contractile function after revascularization.
Document type :
Conference papers
Complete list of metadata
Contributor : wafa baccouch Connect in order to contact the contributor
Submitted on : Monday, February 7, 2022 - 10:20:47 AM
Last modification on : Friday, August 5, 2022 - 2:54:52 PM



Wafa Baccouch, Sameh Oueslati, Basel Solaiman, Salam Labidi. Automatic regional estimation of myocardial strain using deep learning. ICICT2021: International Congress on Information and Communication Technology, Feb 2021, Londres, United Kingdom. pp.41-48, ⟨10.1007/978-981-16-1781-2_5⟩. ⟨hal-03559690⟩



Record views