Rapid Fire Abstracts
Pezad Doctor, MD
Pediatric Cardiac MRI/CT fellow
UT Southwestern Medical Center
Munes Fares, MD
Assistant Professor
UT Southwestern
Gerald Greil, MD, PhD
Professor
UT Southwestern Medical Center
Tarique Hussain, MD, PhD
Professor
UT Southwestern
Qing Zou, PhD
CMR Physics
UT Southwestern
Qing Zou, PhD
CMR Physics
UT Southwestern
The novel 3D ferumoxytol-enhanced magnetic resonance angiography (MRA) using the non-Cartesian ultra-short echo time (UTE) sequence [1,2] quickly acquires 3D chest MRA without requiring a respiratory navigator or ECG triggering. In addition, a very short echo time helps reduce the susceptibility artifacts that provide excellent spatial resolution in the presence of vascular stents and coils. In this study, we assess the feasibility of 3D UTE sequence in children with congenital heart disease (CHD) and compare the image quality with standard ECG-triggered and respiratory-navigated 3D balanced steady-state free precession (bSSFP) sequence.
Methods:
Thirty-eight consecutive children who underwent cardiac MR between January and July 2023 at our center underwent 3D UTE MRA of the chest (30 cm coverage along the foot-head direction). This sequence was obtained 15 minutes after the start of ferumoxytol infusion in a free-breathing and ungated fashion without the need for any prepulse (TE/TR = 0.1/5.1, flip angle = 180, isotropic acquired resolution 1.74 mm3, isotropic reconstructed resolution 1.34 mm3). This was followed by respiratory gated and ECG-triggered 3D bSSFP with fat saturation prepulse and compressing sensing factor of 4. Objective image quality grading for these sequences (3D bSSFP and 3D UTE Chest) was conducted independently by two readers and ranged from grade 0 (structure not visualized) to grade 4 (structure visible with sharply defined borders).
Results:
Of the 38 children (mean age: 10.3±6.4, 60% males), 3D UTE chest acquisition time was 54 seconds for isotropic field-of-view 30cm3 vs 5 to 8 minutes for 3D bSSFP (varied depending on respiratory pattern and heart rates). For patients needing to assess abdominal and pelvic vasculature, separate UTE abdomen was obtained in 10 (coverage of 60 cm along the foot-head direction, total running time for the chest and abdomen = 1.7 minutes). Of the ten cardiac and extracardiac structures evaluated, the 3D UTE sequence received significantly higher scores in four pulmonary veins, aorta, and pulmonary arteries (p < 0.005) and lower scores in the ventricle, left atrial appendage, and left atrium compared to 3D bSSFP Figure 1. In 3D UTE, head and neck vasculature was visualized in all scans (100%) vs 28 (74%) in 3D bSSFP. Figure 2 demonstrates cases where UTE showed better image quality.
Conclusion:
Ferumoxytol-enhanced 3D whole-heart MRI using a non-Cartesian UTE sequence is a fast (54s) and effective way of obtaining clinical-acceptable image quality of diagnostic ability robust to motion and susceptibility artifacts in children with complex CHD. Extracardiac structures were better visualized by 3D UTE compared to 3D bSSFP sequence.