Quantitative assessment of intramyocardial function using Cine DENSE MRI: a validation study

Introduction

Non-invasive assessment of myocardial function plays a critical role in the management of heart disease. Conventional imaging modalities, such as echocardiography and cine MRI, are widely used for quantitative assessment of global cardiac function and qualitative assessment of regional cardiac function. Quantitative assessment of regional wall motion may additionally improve the accuracy of detecting wall motion abnormalities due to heart disease. Recently, fast cine displacement-encoded (DENSE) MRI was developed to produce relatively high quality strain maps within clinically acceptable breath-hold duration of 12 cardiac cycles [1]. While this pulse sequence is promising for clinical applications, it has not been validated. Therefore, the purpose of this study was to validate the relative accuracy of fast cine DENSE MRI in controls and patients with heart disease.

Methods

Pulse sequence

Myocardial tagged MRI was chosen as the reference method [2–5]. Both the conventional tagged MR and fast cine DENSE pulse sequences were implemented on a 3 T whole-body MR scanner (Tim-Trio, Siemens) equipped with a 12-channel phased array receive coil. Relevant imaging parameters for cine DENSE include: effective spatial resolution = 3.3 mm × 3.3 mm, slice thickness = 7 mm, temporal resolution = 35 ms, and breath-hold duration = 12 heartbeats. Relevant imaging parameters for myocardial tagging include: spatial resolution = 1.3 mm × 1.8 mm, slice thickness = 7 mm, temporal resolution = 35 ms, breath-hold duration = 15 heartbeats, and grid tag spacing = 7 mm.

Cardiac imaging

Twelve healthy human subjects (7 males; 5 females; mean age = 34.5 ± 11.0 years), with no history of heart disease and no risk factors for coronary artery disease, and thirteen patients (12 males; 1 female; mean age = 55.3 ± 11.2 years) with prior history of heart disease were imaged in three short-axis (apical, mid-ventricular, basal) planes of the left ventricle (LV), using fast cine DENSE and myocardial tagged MRI pulse sequences.

Image analysis

For cine DENSE, cardiac contours were segmented manually by two observers through the first 15 cardiac phases. For myocardial tagging, contour segmentation and tag tracking were performed by observer 1 through the first 11 cardiac phases. The resulting segmented functional data were further processed to calculate the second principal strain (E₂). The principle direction of E₂ at end systole is primarily circumferential. The three short-axis images were subdivided into 16 segments according to the American Heart Association standardized model.

For cine DENSE, the intra-observer variability was assessed by calculating the mean difference and coefficient of repeatability (CR) of first and second calculations of E₂, which were made in a blinded fashion with a separation of at least two weeks. Inter-observer variability was assessed by calculating the mean and CR between the two observers.

The accuracy of fast cine DENSE was assessed relative to myocardial tagging by performing Bland-Altman and linear correlation analyses on their E₂ calculations. A p < 0.05 was considered statistically significant.

Results

The intra- and inter-observer CR of DENSE for E₂ measurements were 0.03 and 0.04, respectively (n = 5910). Their corresponding 95% limits of agreement were -0.03 to 0.03 and -0.04 to 0.04. Figure 1 shows representative end-systolic E₂ maps that compare the results by DENSE and myocardial tagging. Both the control and patient examples show good agreement between the two methods. Statistically, as shown in Figure 2, the two methods were strongly correlated (slope = 1.06; bias = 0.02; R² = 0.81; n = 4334) and in good agreement (mean = 0.03; 95% limits of agreement = -0.03 and 0.09).

Figure 1

End-systolic E2 maps of (top row) control and (bottom row) patient comparing tagging and DENSE.

Figure 2

Plots showing linear correlation (left) and Bland-Altman (right) on E2 measurements (n = 4334).

Discussion

This study demonstrated that fast cine DENSE MRI is highly reproducible and produces E₂ measurements that are strongly correlated and in good agreement with those produced by myocardial tagging. We conclude that cine DENSE MRI is a validated method for clinical applications.