Free breathing 2D multi-slice real-time gradient-echo cardiovascular magnetic resonance imaging: impact on left ventricular function measurements compared with standard multi-breath hold 2D steady-state free precession imaging

Objective Multi-breath hold cine-imaging (standard-2D-SSFP) is the standard technique for ventricular function assessment. However, image acquisition is time-consuming and breath holding is required. Ultrafast 2D real-time imaging (2D real time-SSFP) does not require breath holding, but spatial resolution is sacrificed. The accuracy of 2D real time-SSFP free-breathing ventricular function assessment by using an automated contour detection programme has not been systematically studied. Methods and results Twenty-eight subjects (14 with impaired LV function) were investigated by 1.5 Tesla magnetic resonance imaging. Left ventricular short-axis images were acquired with multi-breath hold standard 2D-SSFP and free-breathing 2D real time-SSFP. LV-volumes (EDV, ESV), EF, and mass were determined using a semi-automated contour detection programme. EDV, ESV, EF, and mass were not significantly different between real time- and standard 2D-SSFP in subjects (absolute differences: EDV 3.2 +/- 0.7 mL, ESV 3.0 +/- 0.3 mL, EF 1.9 +/- 0.4 %, mass 0.8 +/- 0.4 g; P >= 0.78) and patients (absolute differences: EDV = 3.0 +/- 0.8 mL, ESV 3.3 +/- 1.0 mL, EF 0.9 +/- 0.5 %, mass 0.9 +/- 0.5 g; P >= 0.73). Automated contour detection required extensive manual correction for real-time imaging (<= 86%). Conclusions Differences in LV function measurements between real-time and standard 2D-SSFP are small, and not significant. Real-time SSFP may be used for rapid LV function assessment when examination time is limited.