Parametric Modeling of the Mouse Left Ventricular Myocardial Fiber Structure

Samer S. Merchant, Arnold David Gomez, James L. Morgan, Edward W. Hsu

Research output: Contribution to journalArticlepeer-review

Abstract

Magnetic resonance diffusion tensor imaging (DTI) has greatly facilitated detailed quantifications of myocardial structures. However, structural patterns, such as the distinctive transmural rotation of the fibers, remain incompletely described. To investigate the validity and practicality of pattern-based analysis, 3D DTI was performed on 13 fixed mouse hearts and fiber angles in the left ventricle were transformed and fitted to parametric expressions constructed from elementary functions of the prolate spheroidal spatial variables. It was found that, on average, the myocardial fiber helix angle could be represented to 6.5° accuracy by the equivalence of a product of 10th-order polynomials of the radial and longitudinal variables, and 17th-order Fourier series of the circumferential variable. Similarly, the fiber imbrication angle could be described by 10th-order polynomials and 24th-order Fourier series, to 5.6° accuracy. The representations, while relatively concise, did not adversely affect the information commonly derived from DTI datasets including the whole-ventricle mean fiber helix angle transmural span and atlases constructed for the group. The unique ability of parametric models for predicting the 3D myocardial fiber structure from finite number of 2D slices was also demonstrated. These findings strongly support the principle of parametric modeling for characterizing myocardial structures in the mouse and beyond.

Original languageEnglish (US)
Pages (from-to)2661-2673
Number of pages13
JournalAnnals of biomedical engineering
Volume44
Issue number9
DOIs
StatePublished - Sep 1 2016
Externally publishedYes

Keywords

  • Cardiac
  • Cardiac DTI atlas
  • Cardiac fiber architecture prediction
  • DTI
  • Parametric modelling

ASJC Scopus subject areas

  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'Parametric Modeling of the Mouse Left Ventricular Myocardial Fiber Structure'. Together they form a unique fingerprint.

Cite this