Lung MRI assessment with high-frequency noninvasive ventilation at 3 T

Abstract

PurposeTo investigate three MR pulse sequences under high-frequency noninvasive ventilation (HF-NIV) at 3 T and determine which one is better-suited to visualize the lung parenchyma. MethodsA 3D ultra-short echo time stack-of spirals Volumetric Interpolated Breath-hold Examination (UTE Spiral VIBE), without and with prospective gating, and a 3D double-echo UTE sequence with spiral phyllotaxis trajectory (3D radial UTE) were performed at 3 T in ten healthy volunteers under HF-NIV. Three experienced radiologists evaluated visibility and sharpness of normal anatomical structures, artifacts assessment, and signal and contrast ratio computation. The median of the three readers‘scores was used for comparison, p < .05 was considered statistically significant. Incidental findings were recorded and reported. ResultsThe 3D radial UTE resulted in less artifacts than the non-gated and gated UTE Spiral VIBE in inferior (score 3D radial UTE = 3, slight artifact without blurring vs. score UTE Spiral VIBE non-gated and gated = 2, moderate artifact with blurring of anatomical structure, p = .018 and p = .047, respectively) and superior lung regions (score 3D radial UTE = 3, vs. score UTE Spiral VIBE non-gated = 2.5, p = .48 and score UTE Spiral VIBE gated = 1, severe artifact with no normal structure recognizable, p = .014), and higher signal and contrast ratios (p = .002, p = .093). UTE Spiral VIBE sequences provided higher peripheral vasculature visibility than the 3D radial UTE (94.4% vs 80.6%, respectively, p < .001). The HF-NIV was well tolerated by healthy volunteers who reported on average minor discomfort. In three volunteers, 12 of 18 nodules confirmed with low-dose CT were identified with MRI (average size 2.6 ± 1.2 mm). ConclusionThe 3D radial UTE provided higher image quality than the UTE Spiral VIBE. Nevertheless, a better nodule assessment was noticed with the UTE Spiral VIBE that might be due to better peripheral vasculature visibility, and requires confirmation in a larger cohort.