Inaccuracy of 2D Cine MRI in Cardiac Function Analysis: Is 3D Whole Heart MRI a Better Alternative?

Background: Morphological imaging is method of choice for cardiac function analysis. Initially it was based only on registration of cardiac wall movement of a small cardiac segment. Due to inaccuracies of 2D-echocardiography, 2D cine MRI (2Dcine) became gold standard for cardiac function measurement in congenital heart defects. Therefor 2D sections are used to reconstruct end-diastolic and end-systolic volume (EDV, ESV) and to calculate dynamic parameters like stroke volume (SV) and ejection fraction (EF). Nevertheless, vessel and valve morphologies are not taken into account using this method. This results in measurement inaccuracies, especially in patients with congenital heart defects. 3D whole heart MRI (3Dwh) represents the heart in its entirety, but is currently only used for morphological representation. A functional analysis is conceivable by acquiring two 3Dwh representations in ES and ED. In direct comparison, we want to determine whether measurement in whole heart technology is equal or superior to functional analysis using 2D cine technology.

Method: Cardiac MR (CMR) was performed in 19 healthy control subjects. All measurements were performed in one session on a 3-T scanner: (1) Navigator controlled 3Dwh in ES and ED, (2) Prospective 2D short-axis cine in ES and ED, and (3) phase contrast (PC) MRI of aortic root. For the assessment of left ventricle (LV) volumes in 2Dcine a simplified contouring was used. 3Dwh volume quantification was performed using threshold-based segmentation. Difference between 2Dcine and 3Dwh volume was calculated in ES and ED (diffVol).

Results: Compared with 3Dwh, 2Dcine shows significantly decreased volume both in ES (50.5 ± 9.7; 61.1 ± 11.3, p < 0.01) and in ED (134.9 ± 16, 9; 140.1 ± 6.2, p < 0.01). DiffVol was significantly lower in ES than in ED (−6.2 mL ± 2.6), which results in an underestimation of ES volume in 2Dcine (50.5 vs. 61.2, p < 0.01). This correlates with a significant decrease in EF and SV in the 3Dwh (56.4 ± 5.9 vs. 61.8 ± 7.7, p < 0.01). In contrast to 2Dcine a strong correlation between 3Dwh-measured ejection fraction and PC could be found (r ² = 0.98).

Conclusion: 3Dwh is superior to 2Dcine in determining function parameters (EDV, ESV, EF, and SV). This is particularly due to an underestimation of ES volume in 2Dcine as a result of incomplete volume acquisition. For susceptible measurements, especially for patients with congenital heart defects, 3Dwh measurements are significantly less susceptible to errors.

Publication History

Article published online:
03 February 2022