Mapping of regional lung microstructural parameters using hyperpolarized <sup>129</sup>Xe dissolved‐phase MRI in healthy volunteers and patients with chronic obstructive pulmonary disease

Purpose To investigate biases in the measurement of apparent alveolar septal wall thickness (SWT) with hyperpolarized xenon‐129 (HXe) as a function of acquisition parameters. Methods The HXe MRI scans with simultaneous gas‐phase and dissolved‐phase excitation were performed using 1‐dimensional projection scans in mechanically ventilated rabbits. The dissolved‐phase magnetization was periodically saturated, and the dissolved‐phase xenon uptake dynamics were measured at end inspiration and end expiration with temporal resolutions up to 10 ms using a Look‐Locker‐type acquisition. The apparent alveolar septal wall thickness was extracted by fitting the signal to a theoretical model, and the findings were compared with those from the more commonly use chemical shift saturation recovery MRI spectroscopy technique with several different delay time arrangements. Results It was found that repeated application of RF saturation pulses in chemical shift saturation recovery acquisitions caused exchange‐dependent gas‐phase saturation that heavily biased the derived SWT value. When this bias was reduced by our proposed method, the SWT dependence on lung inflation disappeared due to an inherent insensitivity of HXe dissolved‐phase MRI to thin alveolar structures with very short T2∗. Furthermore, perfusion‐based macroscopic gas transport processes were demonstrated to cause increasing apparent SWTs with TE (2.5 μm/ms at end expiration) and a lung periphery‐to‐center SWT gradient. Conclusion The apparent SWT measured with HXe MRI was found to be heavily dependent on the acquisition parameters. A method is proposed that can minimize this measurement bias, add limited spatial resolution, and reduce measurement time to a degree that free‐breathing studies are feasible.

Section: Introductionmentioning

confidence: 93%

Investigating biases in the measurement of apparent alveolar septal wall thickness with hyperpolarized 129Xe MRI

Ruppert

Amzajerdian

Xin

et al. 2020

“…For parameters affected by lung inflation (i.e., SVR), inspiratory volumes calculated as a fraction of TLC may offer more reproducible quantification . Despite these shortcomings, important parameters of δ, d , and HCT are within the ranges reported previously by Chang et al and others investigating alveolar physiology.…”

Section: Discussionmentioning

confidence: 90%

“…Imaging weighted to the temporal uptake of 129 Xe coupled with modeling of gas exchange may provide new insights into the pathophysiological sources of observed gas exchange dysfunction and improve quantification of impaired lung function with parameters that are more easily comparable between different MRI equipment and sites. Multiple techniques to achieve this are currently being explored …”

Section: Introductionmentioning

confidence: 99%

Accelerated interleaved spiral‐IDEAL imaging of hyperpolarized ¹²⁹Xe for parametric gas exchange mapping in humans

Zanette

Santyr

2019

Purpose:To demonstrate the feasibility of mapping gas exchange with single breathhold hyperpolarized (HP) 129 Xe in humans, acquiring parametric maps of lung physiology. The potential benefit of acceleration using parallel imaging for this application is also explored. Methods: Six healthy volunteers were scanned with a modified spiral-IDEAL sequence to acquire gas exchange-weighted images using a single dose of 129 Xe.These images were fit with the model of xenon exchange (MOXE) on a voxel-wise basis calculating parametric maps of lung physiology, specifically: air-capillary barrier thickness (δ), alveolar septal thickness (d), capillary transit time (t x ), pulmonary hematocrit (HCT), and alveolar surface area-to-volume ratio (SVR). An accelerated version of the sequence was also tested in subset of 4 volunteers and compared to the fully sampled (FS) results. Results: Mean image-wide values calculated from MOXE parametric maps derived from FS dissolved 129 Xe spiral-IDEAL images were: δ = 0.89 ± 0.17 μm, d = 7.5 ± 0.5 μm, t x = 1.1 ± 0.2s, HCT = 28.8 ± 2.3%, and SVR = 140 ± 16 cm −1 , in good agreement with previously published values based on whole-lung spectroscopy of healthy human subjects.

“…Moreover, most HP gas MRI methods acquire data during a single breath‐hold, but long breath‐holding time imposes burdens on subjects, especially for those patients with compromised respiratory function. Accordingly, various efforts have focused on accelerating the acquisition of HP gas MRI, including methods based on parallel imaging, radial, spiral, compressed sensing MRI (CS‐MRI), or the combination of these techniques . Among them, CS‐MRI is a good solution to shorten scan time by undersampling measurements in k‐space, and without additional hardware required …”

Section: Introductionmentioning

confidence: 99%

“…Accordingly, various efforts have focused on accelerating the acquisition of HP gas MRI, including methods based on parallel imaging, 4 radial, 5 spiral, 6 compressed sensing MRI (CS-MRI), 7 or the combination of these techniques. 8,9 Among them, CS-MRI is a good solution to shorten scan time by undersampling measurements in k-space, and without additional hardware required. 10 Ajraoui et al first applied CS to reconstruct HP 3 He lung images at an acceleration factor (AF) of 2 for 2D images and 5 for 3D ventilation images.…”

Section: Introductionmentioning

confidence: 99%

Fast and accurate reconstruction of human lung gas MRI with deep learning

Duan

Xiao

et al. 2019

Purpose To fast and accurately reconstruct human lung gas MRI from highly undersampled k‐space using deep learning. Methods The scheme was comprised of coarse‐to‐fine nets (C‐net and F‐net). Zero‐filling images from retrospectively undersampled k‐space at an acceleration factor of 4 were used as input for C‐net, and then output intermediate results which were fed into F‐net. During training, a L2 loss function was adopted in C‐net, while a function that united L2 loss with proton prior knowledge was used in F‐net. The 871 hyperpolarized 129Xe pulmonary ventilation images from 72 volunteers were randomly arranged as training (90%) and testing (10%) data. Ventilation defect percentage comparisons were implemented using a paired 2‐tailed Student's t‐test and correlation analysis. Furthermore, prospective acquisitions were demonstrated in 5 healthy subjects and 5 asymptomatic smokers. Results Each image with size of 96 × 84 could be reconstructed within 31 ms (mean absolute error was 4.35% and structural similarity was 0.7558). Compared with conventional compressed sensing MRI, the mean absolute error decreased by 17.92%, but the structural similarity increased by 6.33%. For ventilation defect percentage, there were no significant differences between the fully sampled and reconstructed images through the proposed algorithm (P = 0.932), but had significant correlations (r = 0.975; P < 0.001). The prospectively undersampled results validated a good agreement with fully sampled images, with no significant differences in ventilation defect percentage but significantly higher signal‐to‐noise ratio values. Conclusion The proposed algorithm outperformed classical undersampling methods, paving the way for future use of deep learning in real‐time and accurate reconstruction of gas MRI.