Characterizing regional myofiber damage post acute myocardial infarction using global optimization. (March 2021)

• Record Type: Journal Article
• Authors: Dempsey, Sergio; So, Aaron; Samani, Abbas
• Is Part Of: Computers in biology and medicine. Volume 130(2021)
• Publication Date: 2021-03
• Subjects: Left ventricle mechanics -- Ischemia -- Acute myocardial infarction -- Reconstruction; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2021.104207
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Medical imaging derived cardiac biomechanical models offer a wealth of new information to be used in diagnosis and prognosis of cardiovascular disease. A noteworthy feature of such models is the ability to predict myofiber contraction stresses during acute or chronic ischemic events. Current techniques for heterogeneous contraction models require tissue motion tracking capabilities which are neither available on all imaging modalities, nor currently used in the clinic. Proposed in this article is a proof of concept of a tissue tracking independent technique focused on shape optimization to predict the contraction stresses of in-silico left ventricle models simulating various acute myocardial infarction events. The technique involves three variables defined in the left ventricle muscle. Two of the variables represent the contraction stresses in the healthy and infarct regions while the third is a novel periinfarct variable defining a non-contracting myofiber state allowing finer classification of local myofiber damage. Results indicate that the contraction stress reconstruction errors are overall smaller than 12% when considering standard errors associated with population modelling for the new variable of interest. Highlights: Tissue tracking independent technique is proposed to reconstruct contractility parameters of LV with acute MI. Three variables of contraction stress in healthy/infarct regions, and a myofiber contraction state variable, are considered. Reconstruction problem was formulated based on LV shape at relaxed and contracted states. Shape optimization shape formulation was solved using genetic algorithms. Results show successful parameter reconstruction of a LV in silico model.