Pulmonary MRI in Newborns and Children

Abstract

Lung MRI is an important tool in the assessment and monitoring of pediatric and neonatal lung disorders. MRI can provide both similar and complementary image contrast to computed tomography for imaging the lung macrostructure, and beyond this, a number of techniques have been developed for imaging the key functions of the lungs, namely ventilation, perfusion, and gas exchange, through the use of free-breathing proton and hyperpolarized gas MRI. Here, we review the state-of-the-art in MRI methods that have found utility in pediatric and neonatal lung imaging, the structural and physiological information that can be gleaned from such images, and strategies that have been developed to deal with respiratory (and cardiac) motion, and other technological challenges. The application of lung MRI in neonatal and pediatric lung conditions, in particular bronchopulmonary dysplasia, cystic fibrosis, and asthma, is reviewed, highlighting our collective experiences in the clinical translation of these methods and technology, and the key current and future potential avenues for clinical utility of this methodology. EVIDENCE LEVEL: 2 TECHNICAL EFFICACY: Stage 2.

Keywords: MRI; free‐breathing; hyperpolarized 129Xe; lung; neonatal; pediatric.

FIGURE 1

Structural T₁‐weighted gradient echo and T₂‐weighted (HASTE) imaging alongside gadolinium contrast enhanced perfusion imaging in infants and children 0–5 years old with cystic fibrosis. Reprinted with permission of the American Thoracic Society. Copyright © 2024 American Thoracic Society. All rights reserved. The American Journal of Respiratory and Critical Care Medicine is an official journal of the American Thoracic Society. White arrows: airway structural abnormalities including wall thickening and/or bronchiectasis, white arrowheads: mucus plugging, black arrowheads: perfusion abnormalities. HASTE = half‐Fourier single‐shot turbo spin echo.

FIGURE 2

Comparison of UTE structural lung images acquired with 3D radial and FLORET trajectories in a neonate with bronchopulmonary dysplasia. Both images were acquired with 180 mm³ FOV and 0.7 mm isotropic resolution. Scan time was 16 minutes 44 sec for 3D radial UTE and 4 minutes 41 sec for FLORET UTE. (Images provided courtesy of Professor Sean Fain and Dr Andrew Hahn at University of Iowa.) FLORET = fermat‐looped, orthogonally encoded trajectories; FOV = field of view; UTE = ultrashort echo time.

FIGURE 3

(a) Illustration of improved motion compensation and signal‐to‐noise ratio of iterative motion compensation (iMoCo) reconstruction compared with non‐gated, XD‐GRASP‐like motion‐resolved, soft‐gated reconstructions, and non‐iterative motion compensated (MoCo) reconstructions in a 10‐week‐old infant. Reprinted with permission from Reference . Red arrows indicate vessels and green arrows the airways visualized on the iMoCo images. (b) Image‐based motion navigator, iMoCo images without (c) and with (d) bulk motion rejection showed reduced motion blurring after bulk motion compensation. (Data highlighted by the red‐dash rectangle were removed prior to reconstruction as they were acquired prior to a bulk movement; red arrow.) XD‐GRASP = extra‐dimensional golden‐angle radial sparse parallel.

FIGURE 4

PREFUL MRI results obtained from an 8‐month‐old infant using a 2D gradient echo sequence. Reprinted with permission from Reference . (a) Ventilation‐, (b) perfusion‐weighted maps. R _vent = regional ventilation; FVL_cc = flow‐volume loop cross correlation; Q _N  = normalized perfusion percentage maps; PREFUL = phase‐resolved functional lung.

FIGURE 5

Examples of HASTE images showing the key features of the UNiforme Scoring of the disEAsed Lung in bronchopulmonary dysplasia (UNSEAL BPD) scoring system. Reprinted with permission from Reference . Each feature is scored from 1 to 5 where 1 is absence and 5 is global high prevalence/severe disease. HASTE = half‐Fourier single‐shot turbo spin echo.

FIGURE 6

Inhaled hyperpolarized gas images of lung ventilation in a 2‐month‐old infant using ³He (top)—reprinted with permission from Reference —and a 3‐month‐old infant using ¹²⁹Xe (bottom)—reprinted with permission from Reference . Both infants were born prematurely, the latter with confirmed severe bronchopulmonary dysplasia. Arrows indicate major ventilation abnormalities.

FIGURE 7

¹²⁹Xe diffusion‐weighted imaging in the lungs of school‐aged children born prematurely with and without bronchopulmonary dysplasia (BPD). Reprinted with permission of the American Thoracic Society. Copyright © 2024 American Thoracic Society. All rights reserved. The American Journal of Respiratory and Critical Care Medicine is an official journal of the American Thoracic Society. Left: apparent diffusion coefficient (ADC) and lung microstructural dimension (LmD) maps acquired from children born prematurely with and without BPD compared with term born children (global ADC and LmD are quoted). Right: distribution of ¹²⁹Xe ADC values in the three groups showing elevated ADC in children born pre‐term with BPD compared with those born prematurely without BPD and term born children.

FIGURE 8

Rate of pulmonary exacerbations (a) alongside the time‐course of global (b), morphological (c), perfusion (d), wall‐thickness/bronchiectasis (e) and mucus (f) MRI scores derived from the scoring system in Table 3 in infants and preschool children with cystic fibrosis over the first few years of life, stratified according to diagnosis method; new‐born screening (NBS), early clinical diagnosis (ECD), late clinical diagnosis (LCD). Reprinted with permission of the American Thoracic Society. Copyright © 2024 American Thoracic Society. All rights reserved. The American Journal of Respiratory and Critical Care Medicine is an official journal of the American Thoracic Society.

FIGURE 9

Comparison of pulmonary MRI before and after triple combination therapy in an 11‐year‐old patient with cystic fibrosis. From left to right: morphological (PROPELLER) images, fractional ventilation (FV), perfusion (Q) and ventilation and perfusion defect masks derived from dynamic 2D bSSFP‐based matrix pencil imaging, respectively before (a) and after (b) the therapy showed partial resolution of atelectasis in the right upper lobe and reduced bronchial wall thickening on morphological imaging, and corresponding improvement of ventilation and reduced ventilation‐perfusion mismatch. Reprinted with permission from Reference . bSSFP = balanced steady‐state free precession; PROPELLER = Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction.

FIGURE 10

(a) Worsening of ventilation heterogeneity (VDP change 16.6% to 27.3%) over 16 months while FEV₁ z‐score and LCI did not change significantly in a pediatric patient with cystic fibrosis (CF) visualized by hyperpolarized ³He MRI. (b) Dramatic improvement in ventilation in response to a 4 week course of ivacaftor treatment in an adult with CF visualized by hyperpolarized ³He MRI. Reprinted with permission from Reference . LCI = lung clearance index; FEV₁ = forced expiratory volume in 1 second; VDP = ventilation defect percentage.

FIGURE 11

Comparison of breath‐hold ¹²⁹Xe gas‐ and free‐breathing ¹H PREFUL‐based ventilation maps in pediatric patients with CF, showing a variable level of agreement. Reprinted with permission from Reference . (a) Patient with mild CF (normal FEV₁) and (b) patient with moderate CF (abnormal FEV₁). PREFUL = phase‐resolved functional lung; FEV₁ = forced expiratory volume in 1 second.

FIGURE 12

³He images of lung ventilation heterogeneity (left) alongside 3D SPGR images of two patients (B* and J*) and computed tomography images (B** and J**) in school‐aged children with primary ciliary dyskinesia (PCD). Reprinted with permission of the American Thoracic Society. Copyright © 2024 American Thoracic Society. All rights reserved. Annals of the American Thoracic Society is an official journal of the American Thoracic Society. Right: correlation of ventilation defect percentage (VDP) with the lung clearance index (LCI) metric derived from the multi‐breath washout pulmonary function test. SPGR = spoiled gradient recalled echo.