TOPLINE

A gadolinium-enhanced five-dimensional (5D) MRI technique significantly shortened scan times while maintaining comparable diagnostic quality to standard two- and three-dimensional (2D and 3D) imaging in patients with congenital heart disease (CHD), a retrospective study found.

METHODOLOGY

• Researchers retrospectively evaluated 45 consecutive patients with CHD (mean age, 34.9 years; 21 women) referred for clinically indicated MRI who underwent standard 2D and 3D and the proposed 5D MRI scans.
• The novel 5D sequence used a gadolinium-enhanced fast-interrupted steady-state technique to resolve both cardiac and respiratory motion without breath holding.
• Primary outcomes included scan time comparisons and left and right ventricular end-diastolic volumes (LVEDV and RVEDV), end-systolic volumes (LVESV and RVESV), and ejection fractions (LVEF and RVEF) for 2D vs 5D.
• The researchers recorded the vessel sharpness and length of the right coronary artery (RCA) and the left anterior descending (LAD) CA using prospective 3D images and mid-diastolic 5D images.
• Quantitative assessments by three independent blinded reviewers compared the quality of 3D vs 5D images on a 5-point Likert scale.

TAKEAWAY

• The scan time for 5D MRI (6 minutes 18 seconds) was significantly shorter than that for standard 2D (7 minutes 16 seconds) and 3D (6 minutes 55 seconds) protocols.
• Strong correlations were observed between 5D and 2D measurements of LVEDV (R², 0.94), LVESV (R², 0.83), RVEDV (R², 0.85), and RVESV (R², 0.80), with moderate-to-strong correlations observed for LVEF (R², 0.51) and RVEF (R², 0.66), although significant biases remained.
• 3D imaging demonstrated significantly higher vessel sharpness than 5D imaging for both the RCA (36.8% vs 24.3%) and LAD CA (40.5% vs 31.7%), but the length of the LAD CA measured using 5D imaging was significantly longer than that measured using 3D imaging (6.9 vs 5.3 mm).
• Three blinded observers rated 98% of 5D images vs only 78%-87% of corresponding standard 3D images as diagnostically useful (grade 2 or higher).

IN PRACTICE

"Gd [gadolinium]-enhanced free-running 5D FISS [fast-interrupted steady-state technique] provided dynamic whole heart images that may be used to assess cardiac function and morphology," the authors wrote.

"Despite the simplified and efficient data acquisition, the offline reconstruction times (> 8 h) for 5D imaging remain a significant barrier for clinical translation. However, ongoing developments in deep-learning based reconstruction may bridge this gap and provide efficient reconstruction of 5D data at the scanner," they added.

SOURCE

The study was led by Christopher W. Roy, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland. It was published online on June 24, 2026, in the Journal of Magnetic Resonance Imaging.

LIMITATIONS

The study was limited by offline reconstruction times exceeding 8 hours, the unavailability of the dedicated software for 3D CINE analysis, and an unexplored respiratory data dimension. It was further limited by highly undersampled 5D data, compressed sensing reconstruction, inherent differences between 2D breath-held and free-breathing 5D imaging, and significant biases between 2D and 5D ventricular volume and EF measurements.

DISCLOSURES

The research received funding support through multiple grants from the Swiss National Science Foundation. Open access publishing facilitated by Université de Lausanne. The authors reported having no conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.