Remodeling after acute myocardial infarction: mapping ventricular dilatation using three dimensional CMR image registration

Declan P. O’Regan, Wenzhe Shi, Ben Ariff, A. John Baksi, Giuliana Durighel, Daniel Rueckert, Stuart A. CookJournal of Cardiovascular Magnetic Resonance 2012 · DOI

Abstract

The severity of ischemic injury has a significant effect on local ventricular wall remodeling with only modest dilatation observed within non-ischemic myocardium. Limitation of chronic remodeling may therefore depend on therapies directed at modulating ischemia-reperfusion injury. CMR co-registration has potential for assessing dynamic changes in ventricular structure in relation to therapeutic interventions.

Introduction

Following an acute ST-elevation myocardial infarction (STEMI), the heart often undergoes a process known as "remodeling." This involves the progressive dilatation and shape change of the left ventricle, which can ultimately lead to heart failure. Traditionally, clinical assessment has relied on global metrics like total end-diastolic volume. However, these measures lack spatial specificity, failing to distinguish between generalized expansion and localized stretching at the site of injury.

Researchers have adapted 3D non-rigid image registration techniques—originally developed for neuroimaging—to track these structural changes with high precision. By mapping baseline and one-year follow-up MRI scans onto a common 3D model, this study demonstrates that ventricular stretching is highly localized to the infarcted zone, particularly in cases of deep tissue scarring or microvascular obstruction.

Methods

The study followed 46 patients (44 male, 2 female; age 33–77) who had suffered an acute STEMI and were treated with primary percutaneous coronary intervention (PPCI). Cardiovascular Magnetic Resonance (CMR) imaging was performed within 7 days of the event and again at one year.

The core methodology involved a multi-step 3D image registration pipeline:

  1. Segmentation: Automated identification of heart walls using a probabilistic cardiac atlas.
  2. Co-registration: Alignment of baseline and follow-up scans using intensity-based similarities to track the deformation of specific myocardial points.
  3. Infarct Mapping: Identification of scar tissue and transmurality using Late Gadolinium Enhancement (LGE).
  4. 3D Mesh Construction: Building a 3D model to quantify "local wall dilatation," representing the percentage of surface area expansion in specific regions.
Overview of the 3D image registration and mesh generation workflow.

Results

The analysis revealed that remodeling is a highly localized phenomenon rather than a uniform global expansion.

Localized vs. Global Dilatation

Dilatation within the infarcted muscle was significantly higher than in healthy, remote muscle (1.6% ± 1.0 vs 0.3% ± 0.9, P = 0.02). In patients exhibiting significant "global" remodeling (>20% increase in total volume), local dilatation at the scar site reached as high as 4.8%.

Predictors of Remodeling: MVO and Transmurality

Microvascular Obstruction (MVO), or "no-reflow," was a critical predictor of adverse remodeling. Patients with MVO experienced 3.8% local dilatation, whereas those without it showed a slight contraction (-1.6%) in the injured zone. Furthermore, a "tipping point" was identified: local stretching became significantly more severe when the infarct involved more than 50% of the wall's thickness.

CMR images illustrating the segmentation and mapping of the left ventricle across different views.

Structural Changes

Over the one-year period, average wall thickness in the acutely infarcted region decreased from 10.1 mm to 8.9 mm. In contrast, healthy regions remained stable at approximately 7.1 mm.

Discussion and Conclusion

This study successfully demonstrates that complex deformation-tracking algorithms can be adapted from neuroscience to the dynamic environment of the heart. The findings emphasize that chronic remodeling is primarily driven by the mechanical failure of the infarcted zone, while non-injured myocardium remains relatively stable.

Limitations

The study cohort was predominantly male, which may limit generalizability. Additionally, extreme wall thinning can reduce the signal-to-noise ratio, potentially affecting registration accuracy. Future work should integrate regional functional measures, such as strain, to complement these structural findings.

Clinical Implications

The ability to map regional dilatation provides a powerful tool for evaluating therapeutic interventions. This 3D mapping technique could serve as a sensitive endpoint in clinical trials, determining whether specific pharmacological treatments effectively prevent "stretching" at the site of injury.

Summary updated

Abstract

The severity of ischemic injury has a significant effect on local ventricular wall remodeling with only modest dilatation observed within non-ischemic myocardium. Limitation of chronic remodeling may therefore depend on therapies directed at modulating ischemia-reperfusion injury. CMR co-registration has potential for assessing dynamic changes in ventricular structure in relation to therapeutic interventions.
BibTeX
@article{oregan2012remodeling,
  author = {Declan P. O’Regan and Wenzhe Shi and Ben Ariff and A. John Baksi and Giuliana Durighel and Daniel Rueckert and Stuart A. Cook},
  title = {{Remodeling after acute myocardial infarction: mapping ventricular dilatation using three dimensional CMR image registration}},
  journal = {Journal of Cardiovascular Magnetic Resonance},
  year = {2012},
  doi = {10.1186/1532-429x-14-41},
}
Figure