MRI-Derived Myocardial Strain Measures in Normal Subjects

Area of Science:

• Cardiology
• Medical Imaging
• Biophysics

Background:

• Cardiac magnetic resonance imaging (MRI) feature tracking (FT) quantifies myocardial deformation using cine images.
• Previous studies reported non-overlapping confidence intervals for MRI-FT parameters, indicating significant variability.
• Understanding normal ranges and sources of variation is crucial for accurate clinical application.

Purpose of the Study:

• To systematically review and meta-analyze normal ranges of MRI-based feature tracking (FT) myocardial strain.
• To identify factors contributing to variations in these strain measurements.
• To compare normal ranges across different MRI-derived strain techniques.

Main Methods:

• A systematic search of four databases (EMBASE, SCOPUS, PUBMED, Web of Science) was conducted for MRI strain studies.
• Key terms included "feature tracking," "cardiac MRI," and "strain" for left (LV) and right ventricular (RV) analysis.
• A random-effects model pooled data for LV global longitudinal strain (GLS), global circumferential strain (GCS), global radial strain (GRS), and RVGLS, with meta-regression analyzing sources of variation.

Main Results:

• Data from 18 papers included 659 healthy subjects.
• Pooled means: LVGLS -20.1%, LVGCS -23%, LVGRS 34.1%, RVGLS -21.8%.
• Meta-regression revealed field strength as a significant source of variation for LVGCS (p=0.041); other factors did not significantly influence variations in LVGLS, LVGRS, or RVGLS.

Conclusions:

• This meta-analysis establishes pooled normal ranges for 4 MRI-derived myocardial strain methods.
• Left ventricular circumferential strain (LVGCS) demonstrated variability associated with magnetic field strength.
• Normal ranges for LVGLS and LVGCS were largely concordant, while LVGRS and RVGLS showed substantial variation.