Elsevier

Medical Image Analysis

Volume 9, Issue 4, August 2005, Pages 376-393
Medical Image Analysis

Tagged magnetic resonance imaging of the heart: a survey

https://doi.org/10.1016/j.media.2005.01.003Get rights and content

Abstract

Magnetic resonance imaging (MRI) of the heart with magnetization tagging provides a potentially useful new way to assess cardiac mechanical function, through revealing the local motion of otherwise indistinguishable portions of the heart wall. While still an evolving area, tagged cardiac MRI is already able to provide novel quantitative information on cardiac function. Exploiting this potential requires developing tailored methods for both imaging and image analysis. In this paper, we review some of the progress that has been made in developing such methods for tagged cardiac MRI, as well as some of the ways these methods have been applied to the study of cardiac function.

Introduction

Although heart disease is of great clinical importance, there are many limitations of the conventional methods used to assess cardiac function. Clinical assessment of a patient’s symptoms and their functional or exercise tolerance impairment correlates poorly with conventional imaging-derived measurements of their cardiac function, such as ejection fraction (EF). Furthermore, although there is a good overall statistical correlation between measures such as EF and the subsequent clinical course, in any individual subject they are of poor predictive value. In addition to the limitations of conventional function measures for clinical assessment, they are also very limited as tools to investigate mechanical factors involved in normal and abnormal cardiac function or to aid in the design and evaluation of new therapies for impaired cardiac function.

A major source of the limitations of conventional cardiac function measurements is their inability to follow the motion of individual portions of the heart wall. The availability of tomographic imaging methods, such as conventional magnetic resonance imaging (MRI), has improved the reliability of global cardiac function measurements, such as the ventricular volumes or stroke volumes. However, the paucity of reliably identifiable landmarks in the heart wall largely limits the assessment of regional cardiac function to simply tracking the motion of the endocardial or epicardial boundaries of the heart wall in such images and trying to infer the corresponding distribution of intramural motion. Such contour tracking is subject to potential error due to the inability to reliably correct for the through-plane component of the 3D motion of the heart relative to the fixed image plane. Furthermore, any other components of the heart wall motion, such as torsional components between different “rings” of the heart muscle along the axis of the heart or within-wall components of motion such as shearing or transmural variation of radial thickening, are essentially invisible with conventional tomographic imaging methods. Although qualitative assessment of regional cardiac motion by visual evaluation of dynamic displays of tomographic MRI or ultrasound data can be clinically very useful, it is also very subjective and difficult to compare between different examinations and different subjects.

There is intrinsic sensitivity of MRI to motion, as was recognized even early in the experience with in vivo nuclear magnetic resonance, before the development of MRI per se. This motion sensitivity stems primarily from two sources: (1) There is persistence (within the limits of the relaxation times) of local alterations in the magnetization of material (such as the heart wall), even in the presence of motion; this allows us to deliberately (and non-invasively) produce local perturbations of the magnetization of the heart wall that will be visible in the images and will serve as material tags within the heart wall. (2) The motion of excited (signal-producing) nuclei in the presence of the magnetic field gradients used during MRI data acquisition has the potential to induce a phase shift of their signal due to the motion of the nuclei, e.g., due to heart wall motion, that is related to the velocity of or distance traveled by the nuclei. While these effects can be seen in the course of the performance of conventional MRI of the heart (Van Dijk, 1984), we can also design modified MRI pulse sequences to produce images that specifically exploit these effects, in order to use them to quantitatively assess motion. Although these MRI motion characterization methods have previously been primarily applied to the study of blood flow, they can also be used to study the motion of the heart wall (as discussed in Section 5). We will here focus on a review of some aspects of the use of magnetization tagging in MRI of the heart to evaluate regional cardiac function. This will include consideration of some aspects of the imaging process itself, some approaches to the process of analyzing the tagged images and extracting quantitative information about cardiac function from them, and some initial results of using these tagged MRI methods to study cardiac function in health and disease or models of disease. We will present a survey of some current approaches to tagged MRI, rather than original results, and will compare tagged MRI to some other conventional methods for cardiac motion assessment.

Section snippets

Tagged MRI methods for imaging local cardiac function

Myocardial tagging with tag analysis is an MRI technique that can be used for quantitative assessment of intramyocardial contractile function (Zerhouni et al., 1988, Axel and Dougherty, 1989a). The principle of myocardial tagging is based on producing a spatial pattern of saturated magnetization within the heart wall, e.g., at end diastole, and then imaging the resulting deformation of the pattern as the heart contracts through the cardiac cycle. Tagged MR imaging can be divided into two

Tagged image analysis

The analysis of tagged cardiac cine MR images yields measures of global and regional cardiac function. The analysis of tagged MR images may involve several steps, including: (1) image preparation, (2) boundary surface extraction, (3) tag tracking, (4) 3D motion reconstruction, and (5) intra/intersubject comparison, statistical model formation and computer-aided diagnosis.

Image preparation facilitates subsequent analysis through steps such as: (1) removing image artifacts such as RF

Applications of tagged cardiac MRI

While the methods of performing and analyzing tagged cardiac MRI described above are still under development, some interesting initial results of their use are already available. We will briefly review some representative examples below.

An important initial application of tagged cardiac MRI is the mapping of the normal patterns of heart wall motion. Previous imaging methods used to study heart wall motion have been limited largely to following the motion of the surface of the heart wall, due to

Comparison with some alternative cardiac function imaging methods

While other imaging methods, including new multi-slice X-ray computed tomography, can provide good quality cross-sectional images of the moving heart, only MRI and ultrasound can provide noninvasive data on the motion patterns within the heart wall.

In addition to the ability to track material points through magnetization tagging, MRI offers the possibility of tracking their motion through motion-induced phase shifts in their MR signal. The origin of these shifts is the dependence of the

Summary

While it still a relatively new area and is continuing to undergo active development, tagged MRI of cardiac function promises to provide new insight into normal and abnormal cardiac function. Achieving a greater practical impact of tagged MRI will be dependent on the further development of more efficient methods of acquiring and analyzing the tagged image data. Challenges for tagged MRI include increasing spatial and temporal resolution while decreasing imaging time. While these are all

Acknowledgements

Dimitris Metaxas and a series of several students and programmers have contributed significantly to the development of tagged MRI methods in our laboratory. Grant support for this work was provided by the NIH through Grant RO1-HL-43014 and Grant RO1-LM-06638-01.

References (133)

  • A. Amini et al.

    Coupled B-snake grids and constrained thin-plate splines for analysis of 2D tissue deformations from tagged MRI

    IEEE Transactions on Medical Imaging

    (1998)
  • A. Amini et al.

    Measurement of cardiac deformations from MRI: Physical and mathematical models

    (2001)
  • K. Augenstein et al.

    Ch 3: Finite element modeling for three-dimensional motion reconstruction and analysis

  • L. Axel et al.

    Intensity correction in surface-coil MR imaging

    American Journal of Radiology

    (1987)
  • L. Axel et al.

    MR imaging of motion with spatial modulation of magnetization

    Radiology

    (1989)
  • L. Axel et al.

    Heart wall motion: improved method of spatial modulation of magnetization for MR imaging

    Radiology

    (1989)
  • L. Axel et al.

    Regional heart wall motion: two-dimensional analysis and functional imaging with MR imaging

    Radiology

    (1992)
  • F. Behloul et al.

    Neuro-fuzzy systems for computer-aided myocardial viability assessment

    IEEE Transactions on Medical Imaging

    (2001)
  • P.D. Bergey et al.

    Focal hypertrophic cardiomyopathy simulating a mass: MR tagging for correct diagnosis

    American Journal of Radiology

    (2000)
  • J. Bogaert et al.

    Functional recovery of subepicardial myocardial tissue in transmural myocardial infarction after successful reperfusion

    Circulation

    (1999)
  • J. Bogaert et al.

    Regional nonuniformity of normal adult human left ventricle

    American Journal of Physiology-Heart and Circulatory Physiology

    (2001)
  • B.D. Bolster et al.

    Myocardial tagging in polar coordinates with use of striped tags

    Radiology

    (1990)
  • S. Bouton et al.

    Differentiation of tumor from viable myocardium using cardiac tagging with MR imaging

    Journal of Computer Assisted Tomography

    (1991)
  • Chandrashekara, R., Mohiaddin, R.H., Rueckert, D., 2002. Analysis of myocardial motion in tagged MR images using...
  • Chen, Y., Amini, A., 2001. A MAP framework for tag line detection in SPAMM data using Markov random fields on the...
  • P. Croisille et al.

    Differentiation of viable and nonviable myocardium by the use of three-dimensional tagged MRI in 2-day-old reperfused canine infarcts

    Circulation

    (1999)
  • C.W. Curry et al.

    Mechanical dyssynchrony in dilated cardiomyopathy with intraventricular conduction delay as depicted by3D tagged magnetic resonance imaging

    Circulation

    (2000)
  • X. Deng et al.

    Rapid 3D LV strain reconstruction from tagged cardiac MR images

    Proceedings of ISMRM

    (2003)
  • T. Denney

    Identification of myocardial tags in tagged MR images without prior knowledge of myocardial contours

  • T. Denney et al.

    Optimal brightness functions for optical flow estimation of deformable motion

    IEEE Transactions on Image Processing

    (1994)
  • S.-J. Dong et al.

    Regional left ventricular systolic function in relation to the cavity geometry in patients with chronic right ventricular pressure overload: a three-dimensional tagged magnetic resonance imaging study

    Circulation

    (1995)
  • S.-J. Dong et al.

    Independent effects of preload, afterload, and contractility on left ventricular torsion

    American Journal of Physiology-Heart and Circulatory Physiology

    (1999)
  • S.-J. Dong et al.

    MRI assessment of LV relaxation by untwisting rate: a new isovolumetric phase measure of t

    American Journal of Physiology-Heart and Circulatory Physiology

    (2001)
  • M.T. Donofrio et al.

    Regional wall motion and strain of transplanted hearts in pediatric patients using magnetic resonance tagging

    American Journal of Physiology

    (1999)
  • Dornier, C., Ivancevic, M., Lecoq, G., Osman, N., Foxall, D., Righetti, A., Vallée, J., 2002. Assessment of the left...
  • L. Dougherty et al.

    Validation of an optical flow method for tag displacement estimation

    IEEE Transactions on Medical Imaging

    (1999)
  • R.R. Edelman et al.

    Coronary arteries: breath-hold MR angiography [see comments]

    Radiology

    (1991)
  • R.L. Ehman et al.

    Adaptive technique for high-definition MR imaging of moving structures

    Radiology

    (1989)
  • R.L. Ehman et al.

    Magnetic resonance imaging with respiratory gating: techniques and advantages

    American Journal of Roentgenology

    (1984)
  • F.H. Epstein et al.

    Segmented k-space fast cardiac imaging using an echo-train readout

    Magnetic Resonance in Medicine

    (1999)
  • Z.A. Fayad et al.

    Right ventricular regional function using MR tagging: normals versus chronic pulmonary hypertension

    Magnetic Resonance in Medicine

    (2001)
  • S.E. Fischer et al.

    Improved myocardial tagging contrast

    Magnetic Resonance in Medicine

    (1993)
  • S.E. Fischer et al.

    True myocardial motion tracking

    Magnetic Resonance in Medicine

    (1994)
  • M.A. Fogel et al.

    A study in ventricular–ventricular interaction

    Circulation

    (1995)
  • M.A. Fogel et al.

    Mechanics of the single ventricle. A study in ventricular–ventricular intersction II

    Circulation

    (1998)
  • M.A. Fogel et al.

    Diastolic biomechanics in normal infants utilizing MRI tissue tagging

    Circulation

    (2000)
  • C.G. Fonseca et al.

    Aging alters patterns of regional nonuniformity in LV strain relaxation: a 3-D MR tissue tagging study

    American Journal of Physiology-Heart and Circulatory Physiology

    (2003)
  • J. Garot et al.

    Fast determination of regional myocardial strain fields from tagged cardiac images using harmonic phase (HARP) magnetic resonance imaging

    Circulation

    (2000)
  • G. Geskin et al.

    Quantitative assessment of myocardial viability after infarction by dobutamine magnetic resonance tagging

    Circulation

    (1998)
  • M.J. Goette et al.

    Quantification of regional contractile function after infarction: strain analysis superior to wall thickening analysis in discriminating infarct from remote myocardium [comment]

    Journal of the American College of Cardiology

    (2001)
  • Cited by (102)

    • A comprehensive comparison between shortest-path HARP refinement, SinMod, and DENSEanalysis processing tools applied to CSPAMM and DENSE images

      2021, Magnetic Resonance Imaging
      Citation Excerpt :

      Several non-invasive Magnetic Resonance (MR) imaging techniques have been used to estimate myocardial strain. Among them, Tagging MR imaging has been intensely used for the evaluation of strain [5,6], considering the conventional tag analysis (i.e., following the intersections of the tag lines) the current gold-standard MR method for the estimation of heart deformation [7]. One of the most used Tagging modalities is Complementary Spatial Modulation of Magnetization (CSPAMM) [8], which uses two complementary SPAMM acquisitions to generate a new image with better relaxation properties.

    • Nonlinear theory of elasticity: Applications in biomechanics (revised edition)

      2023, Nonlinear Theory Of Elasticity: Applications In Biomechanics (Revised Edition)
    View all citing articles on Scopus
    View full text