Clinical Feasibility of Structural and Functional MRI in Free‐Breathing Neonates and Infants

Abstract

Representative Rvent, FVLcc, and QN maps overlaid on the registered GRE image (grayscale) for participant 11 (top) and participant 07 (bottom). Yellow arrows highlight regions of low FVLcc and QN in the left lung. RVent = regional ventilation; FVLcc = flow volume loop cross correlation; QN = normalized perfusion. by Brandon Zanette et al. (1696–1707)

Inter‐ and intravisit repeatability of free‐breathing MRI in pediatric cystic fibrosis lung disease

“…However, in the Kaireit et al study a threshold of 90th percentile × 0.4 was used to determine VDP from RVent distributions, whereas a k-means clustering was used in the present study to stay consistent with previous work in pediatric CF. 11,26,28 VDP parameters derived from PREFUL MRI showed high intravisit repeatability, comparable to VDP Xe-MS and VDP Xe-SS in the pediatric CF group; however, factors affecting the variability of the two methods differ. Variability in Xe-MRI ventilation distributions could be attributed to the differences in the breath-hold performed by the patient (i.e., more or less the gas volume inhaled relative to the prescribed dose), whereas for PREFUL MRI previous studies have shown that changes in breathing frequency and tidal volumes can influence RVent values.…”

“…Similarly, the combined RVent and RFVL defect map showed the highest regional agreement (DSC) with Xe‐MRI defect regions. However, in the Kaireit et al study a threshold of 90th percentile × 0.4 was used to determine VDP from RVent distributions, whereas a k‐means clustering was used in the present study to stay consistent with previous work in pediatric CF 11,26,28 …”

“…24 A combined VDP (VDP combined ) was determined by applying an inclusive OR operation (∪) on VDP RVent and VDP CC . 26,27 For Xe-MRI, images were normalized using an N4ITK bias field correction in 3D Slicer (http://www.slicer. org, 33 ).…”

“…Specifically, FVL cross‐correlation has been shown to be sensitive to chronic lung allograft syndrome 24 and to indacaterol‐glycopyrronium treatment response in COPD 25 . PREFUL MRI has also been demonstrated to be feasible in term neonates and older infants 26 and has been shown to be repeatable in health and COPD 27 …”

“…Postprocessing techniques such as Fourier decomposition, matrix pencil, self-gated non-contrast-enhanced functional lung, and phase-resolved functional lung (PREFUL) can generate ventilation-and perfusion-weighted maps of the lung based on free-breathing MRI data and have been used to assess lung function in a variety of pulmonary diseases. [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] Whereas other postprocessing methods can require extensive pulse sequence modifications, PREFUL MRI can be performed using a conventional spoiled gradient echo sequence, showing promise at both 1.5 and 3 T. PREFUL MRI also sorts images acquired during free breathing into complete respiratory cycles in the time domain, so both static and dynamic ventilation parameters, including regional ventilation (RVent) and flow-volume loops (FVL), can be obtained. Ventilation parameters derived from PREFUL MRI have been shown to correlate to PFTs and Xe-MRI in chronic obstructive pulmonary disease (COPD), CF, and health, 21 and can characterize the severity of coronavirus 2019 infection.…”