Lung parenchyma and structure visualisation in paediatric chest MRI: a comparison of different short and ultra-short echo time protocols.

AIM: To evaluate image quality acquired at lung imaging using magnetic resonance imaging (MRI) sequences using short and ultra-short (UTE) echo times (TEs) with different acquisition strategies (breath-hold, prospective, and retrospective gating) in paediatric patients and in healthy volunteers. MATERIALS AND METHODS: End-inspiratory and end-expiratory three-dimensional (3D) spoiled gradient (SPGR3D) and 3D zero echo-time (ZTE3D), and 3D UTE free-breathing (UTE3D), prospective projection navigated radial ZTE3D (ZTE3D vnav), and four-dimensional ZTE (ZTE4D) were performed using a 1.5 T MRI system. For quantitative assessment, the contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) values were calculated. To evaluate image quality, qualitative scoring was undertaken on all sequences to evaluate depiction of intrapulmonary vessels, fissures, bronchi, imaging noise, artefacts, and overall acceptability. RESULTS: Eight cystic fibrosis (CF) patients (median age 14 years, range 13-17 years), seven children with history of prematurity with or without bronchopulmonary dysplasia (BPD; median 10 years, range 10-11 years), and 10 healthy volunteers (median 32 years, range 20-52 years) were included in the study. ZTE3D vnav provided the most reliable output in terms of image quality, although scan time was highly dependent on navigator triggering efficiency and respiratory pattern. CONCLUSIONS: Best image quality was achieved with prospective ZTE3D and UTE3D readouts both in children and volunteers. The current implementation of retrospective ZTE3D readout (ZTE4D) did not provide diagnostic image quality but rather introduced artefacts over the entire imaging volume mimicking lung pathology.

Optimal imaging protocol for measuring dynamic expiratory collapse of the central airways.

AIM: To compare measurements of expiratory collapse obtained using multidetector computed tomography (MDCT) of the central airways on routine axial and multiplanar reformatted (MPR) images. MATERIALS AND METHODS: Fifty volunteers with normal pulmonary function and no smoking history were imaged using a 64 MDCT system (40 mAs, 120 kVp, 0.625 mm collimation) with spirometric monitoring at end-inspiration and during forced expiration. Measurements of the trachea, right main (RMB) and left main bronchus (LMB) were obtained on axial and MPR images. Inspiratory and dynamic-expiratory cross-sectional area (CSA) measurements were used to calculate the mean percentage expiratory collapse (%Collapse). A paired t-test was used to assess within-subject differences and a Bland-Altman plot was used to assess agreement between the methods. RESULTS: Among 24 men and 26 women (mean age±standard deviation 50±15 years), CSA values were significantly greater on axial than MPR images (all p<0.001); however, the mean difference in %Collapse values for axial versus MPR were small: trachea ≈1% (55 ±19 versus 56±18, p=0.338); LMB identical (60±20 versus 60±17 p=0.856); and, RMB 4% (62 ±19 versus 66±19 p<0.001). On average, creation of MPR required 12 minutes of additional time per case (range=10-15 min). CONCLUSION: Differences in mean %Collapse for axial versus MPR images were small and unlikely to influence clinical management. This finding suggests that MPR may not be indicated for routine assessment of central airway collapse.

Deep learning for improving ZTE MRI images in free breathing.

INTRODUCTION: Despite a growing interest in lung MRI, its broader use in a clinical setting remains challenging. Several factors limit the image quality of lung MRI, such as the extremely short T2 and T2* relaxation times of the lung parenchyma and cardiac and breathing motion. Zero Echo Time (ZTE) sequences are sensitive to short T2 and T2* species paving the way to improved "CT-like" MR images. To overcome this limitation, a retrospective respiratory gated version of ZTE (ZTE4D) which can obtain images in 16 different respiratory phases during free breathing was developed. Initial performance of ZTE4D have shown motion artifacts. To improve image quality, deep learning with fully convolutional neural networks (FCNNs) has been proposed. CNNs has been widely used for MR imaging, but it has not been used for improving free-breathing lung imaging yet. Our proposed pipeline facilitates the clinical work with patients showing difficulties/uncapable to perform breath-holding, or when the different gating techniques are not efficient due to the irregular respiratory pace. MATERIALS AND METHODS: After signed informed consent and IRB approval, ZTE4D free breathing and breath-hold ZTE3D images were obtained from 10 healthy volunteers on a 1.5 T MRI scanner (GE Healthcare Signa Artist, Waukesha, WI). ZTE4D acquisition captured all 16 phases of the respiratory cycle. For the ZTE breath-hold, the subjects were instructed to hold their breath in 5 different inflation levels ranging from full expiration to full inspiration. The training dataset consisting of ZTE-BH images of 10 volunteers was split into 8 volunteers for training, 1 for validation and 1 for testing. In total 800 ZTE breath-hold images were constructed by adding Gaussian noise and performing image transformations (translations, rotations) to imitate the effect of motion in the respiratory cycle, and blurring from varying diaphragm positions, as it appears for ZTE4D. These sets were used to train a FCNN model to remove the artificially added noise and transformations from the ZTE breath-hold images and reproduce the original quality of the images. Mean squared error (MSE) was used as loss function. The remaining 2 healthy volunteer's ZTE4D images were used to test the model and qualitatively assess the predicted images. RESULTS: Our model obtained a MSE of 0.09% on the training set and 0.135% on the validation set. When tested on unseen data the predicted images from our model improved the contrast of the pulmonary parenchyma against air filled regions (airways or air trapping). The SNR of the lung parenchyma was quantitatively improved by a factor of 1.98 and the CNR lung- blood, which is indicating the visibility of the intrapulmonary vessels, was improved by 4.2%. Our network was able to reduce ghosting artifacts, such as diaphragm movement and blurring, and enhancing image quality. DISCUSSION: Free-breathing 3D and 4D lung imaging with MRI is feasible, however its quality is not yet acceptable for clinical use. This can be improved with deep learning techniques. Our FCNN improves the visual image quality and reduces artifacts of free-breathing ZTE4D. Our main goal was rather to remove ghosting artifacts from the ZTE4D images, to improve diagnostic quality of the images. As main results of the network, diaphragm contour increased with sharper edges by visual inspection and less blurring of the anatomical structures and lung parenchyma. CONCLUSION: With FCNNs, image quality of free breathing ZTE4D lung MRI can be improved and enable better visualization of the lung parenchyma in different respiratory phases.

Chest computed tomography in severe bronchopulmonary dysplasia: Comparing quantitative scoring methods.

PURPOSE: Bronchopulmonary dysplasia (BPD) is the most common complication of extreme preterm birth and structural lung abnormalities are frequently found in children with BPD. To quantify lung damage in BPD, three new Hounsfield units (HU) based chest-CT scoring methods were evaluated in terms of 1) intra- and inter-observer variability, 2) correlation with the validated Perth-Rotterdam-Annotated-Grid-Morphometric-Analysis (PRAGMA)-BPD score, and 3) correlation with clinical data. METHODS: Chest CT scans of children with severe BPD were performed at a median of 7 months corrected age. Hyper- and hypo-attenuated regions were quantified using PRAGMA-BPD and three new HU based scoring methods (automated, semi-automated, and manual). Intra- and inter-observer variability was measured using intraclass correlation coefficients (ICC) and Bland-Altman plots. The correlation between the 4 scoring methods and clinical data was assessed using Spearman rank correlation. RESULTS: Thirty-five patients (median gestational age 26.1 weeks) were included. Intra- and inter-observer variability was excellent for hyper- and hypo-attenuation regions for the manual HU method and PRAGMA-BPD (ICCs range 0.80-0.97). ICC values for the semi-automated HU method were poorer, in particular for the inter-observer variability of hypo- (0.22-0.71) and hyper-attenuation (-0.06-0.89). The manual HU method was highly correlated with PRAGMA-BPD score for both hyper- (ρs0.92, p < 0.001) and hypo-attenuation (ρs0.79, p < 0.001), while automated and semi-automated HU methods showed poor correlation for hypo- (ρs < 0.22) and good correlation for hyper-attenuation (ρs0.72-0.74, p < 0.001). Several scores of hyperattenuation correlated with the use of inhaled bronchodilators in the first year of life; two hypoattenuation scores correlated with birth weight. CONCLUSIONS: PRAGMA-BPD and the manual HU method have the best reproducibility for quantification of CT abnormalities in BPD.

Imaging of respiratory muscles in neuromuscular disease: A review.

Respiratory muscle weakness frequently occurs in patients with neuromuscular disease. Measuring respiratory function with standard pulmonary function tests provides information about the contribution of all respiratory muscles, the lungs and airways. Imaging potentially enables the study of different respiratory muscles, including the diaphragm, separately. In this review, we provide an overview of imaging techniques used to study respiratory muscles in neuromuscular disease. We identified 26 studies which included a total of 573 patients with neuromuscular disease. Imaging of respiratory muscles was divided into static and dynamic techniques. Static techniques comprise chest radiography, B-mode (brightness mode) ultrasound, CT and MRI, and are used to assess the position and thickness of the diaphragm and the other respiratory muscles. Dynamic techniques include fluoroscopy, M-mode (motion mode) ultrasound and MRI, used to assess diaphragm motion in one or more directions. We discuss how these imaging techniques relate with spirometric values and whether these can be used to study the contribution of the different respiratory muscles in patients with neuromuscular disease.

Severe tracheal and bronchial collapse in adults with type II mucopolysaccharidosis.

BACKGROUND: Mucopolysaccharidosis type II (MPSII) patients frequently suffer from dyspnoea caused by restrictive airway disease due to skeletal abnormalities as well as glycosaminoglycans (GAG) accumulation at different levels of the airway, including the trachea. In this study we describe the extent of the tracheal and bronchial narrowing, the changes in airway diameter during respiration and the effects of these obstructions on respiratory function in adult MPSII patients. METHODS: Five adult MPSII patients (mean age 40 years) were included. Pulmonary function tests and in- and expiratory chest CT scans were obtained. Cross-sectional areas of trachea and main bronchi were measured at end-inspiration and -expiration and percentage collapse was calculated. RESULTS: There was diffuse narrowing of the entire intra-thoracic trachea and main bronchi and severe expiratory collapse of the trachea in all patients. At 1 cm above the aortic arch the median % collapse of the trachea was 68 (range 60 to 77%), at the level of the aortic arch 64 (range 21-93%), for the main bronchi this was 58 (range 26-66%) on the left and 44 (range 9-76%) on the right side. The pulmonary function tests showed that this airway collapse results in obstructive airway disease in all patients, which was severe (forced expiratory volume <50% of predicted) in four out of five patients. CONCLUSION: In adult MPS II patients, central airways diameters are strikingly reduced and upon expiration there is extensive collapse of the trachea and main bronchi. This central airways obstruction explains the severe respiratory symptoms in MPSII patients.