TL;DR
This study compares 2D and 3D MRI feature tracking for assessing heart function in children who have undergone Fontan surgery, a complex procedure for congenital heart defects. They found that 3D strain measurements, particularly global circumferential strain (GCS), are more reliable and correlate better with standard cardiac function metrics, suggesting its potential as a valuable tool for monitoring these patients.
What Problem Does This Paper Address?
Functional single ventricle (FSV) is a severe congenital heart defect often treated with the Fontan procedure. While this surgery improves survival, patients are at risk for ventricular dysfunction, a major predictor of mortality. Traditional methods like echocardiography have limitations in assessing ventricular function in these patients due to the complex, often asymmetrical geometry of the single ventricle. Cardiac Magnetic Resonance (CMR) with feature tracking (FT) has emerged as a more accurate tool, but the comparison between 2D and 3D strain analysis in this specific patient population has been lacking. This paper aims to bridge that gap by quantitatively comparing 2D and 3D strain measurements derived from CMR-FT in Fontan repair patients and healthy children, evaluating their accuracy, reproducibility, and correlation with established cardiac function parameters.
Methods
The researchers retrospectively analyzed CMR scans from 32 Fontan repair patients and 32 age-matched healthy child volunteers. They used a standard 1.5-Tesla MRI scanner and a balanced steady-state free precession (b-SSFP) cine sequence. Ventricular function and strain were assessed using CMR feature tracking (FT) software.
Key methodological aspects include:
- 2D Strain Analysis: Global longitudinal strain (GLS), global circumferential strain (GCS), and global radial strain (GRS) were measured from standard short-axis and long-axis views.
- 3D Strain Analysis: A novel approach was used to generate a 3D deformable model of the myocardium by interpolating endocardial and epicardial boundaries from the 2D analysis. This allowed for the calculation of 3D GLS, GCS, and GRS.
- Reproducibility Assessment: Intra-observer and inter-observer variability for both 2D and 3D strain measurements were evaluated using Bland-Altman analysis, intraclass correlation coefficients (ICC), and coefficients of variation (CoV).
- Correlation Analysis: Pearson's correlation coefficient was used to assess the relationship between strain parameters and conventional cardiac function measures like ejection fraction (EF), end-diastolic volume (EDV), and end-systolic volume (ESV).
Datasets
The study utilized retrospective CMR data from two groups:
- Fontan Repair Patients: 32 children who had undergone the Fontan procedure and had no intervening cardiac surgery or procedures for at least one year prior to the CMR scan.
- Control Group: 32 gender-matched healthy children with no history of cardiac disease and normal left ventricular ejection fraction (LVEF ≥55%) confirmed by echocardiography.
No new datasets were introduced; the study relied on existing clinical data.
Key Insights and Ideas
- 3D Strain Outperforms 2D in Fontan Patients: The complex geometry of the single ventricle in Fontan patients poses a challenge for 2D strain analysis. 3D strain, by creating a more comprehensive model of the ventricle, appears to better capture the true myocardial deformation.
- GCS is a Strong Indicator: Global Circumferential Strain (GCS), particularly in its 3D form, showed the strongest correlations with key cardiac function parameters like EF, EDV, and ESV. This suggests it's a highly sensitive marker for assessing ventricular performance in this population.
- Reproducibility Matters: The study highlights that 3D strain measurements, especially 3D GCS, demonstrated superior inter-observer agreement compared to 2D strain. This improved reproducibility is crucial for reliable clinical monitoring.
- Potential Overestimation by 2D: The authors propose that 2D strain might overestimate myocardial shortening due to the inclusion of out-of-plane motion, a limitation that the 3D incompressible model-based algorithm helps to overcome.
Key Results
- Significant Differences in Strain: Both 2D GLS and all three 3D strain parameters (GRS, GLS, GCS) showed significant differences between Fontan repair patients and healthy volunteers, indicating impaired myocardial function in the Fontan group.
- 3D GCS and EF Correlation: 3D GCS was significantly associated with ejection fraction (EF) (r = -0.491, p = 0.004).
- 3D GCS and Volume Correlation: 3D GCS also correlated strongly with indexed end-diastolic volume (EDV) (r = 0.523, p = 0.002) and indexed end-systolic volume (ESV) (r = 0.602, p < 0.001).
- Superior Reproducibility of 3D: 3D GCS exhibited the best inter-observer agreement (ICC = 0.87, CoV = 5.15), followed by 3D GLS (ICC = 0.84, CoV = 5.36).
Limitations and Future Work
Limitations acknowledged by the authors include:
- Single-Center Design: The study was conducted at a single center, which may limit the generalizability of the findings.
- Retrospective Nature: The retrospective design could introduce selection bias, as patients were included based on clinical referral for CMR.
- Limited Views: Only four-chamber, two-chamber, and short-axis views were collected, excluding the three-chamber view, which might affect strain analysis.
- Spatial and Temporal Resolution: 3D images, in general, may have lower spatial and temporal resolution compared to 2D images, potentially impacting strain analysis accuracy.
- Strain Rate Analysis: Strain rate analysis was not included due to the temporal resolution limitations of the CMR cine images.
Future work suggested includes:
- Further studies are needed to assess Fontan repair patients using 3D strain and 3D speckle-tracking echocardiography.
- Investigating the differences between 2D and 3D regional strain in more detail, particularly in patients with single right ventricles.
- Exploring the clinical utility of 3D GCS for monitoring abnormal myocardial motion in Fontan circulation patients.
Summary updated