Optimization of steady-state free precession MRI for lung ventilation imaging with 19 F C3 F8 at 1.5T and 3T

Feasibility of Dynamic Inhaled Gas MRI-Based Measurements Using Acceleration Combined with the Stretched Exponential Model

Dynamic inhaled gas (³He/¹²⁹Xe/¹⁹F) MRI permits the acquisition of regional fractional-ventilation which is useful for detecting gas-trapping in lung-diseases such as lung fibrosis and COPD. Deninger's approach used for analyzing the wash-out data can be substituted with the stretched-exponential-model (SEM) because signal-intensity is attenuated as a function of wash-out-breath in ¹⁹F lung imaging. Thirteen normal-rats were studied using ³He/¹²⁹Xe and ¹⁹F MRI and the ventilation measurements were performed using two 3T clinical-scanners. Two Cartesian-sampling-schemes (Fast-Gradient-Recalled-Echo/X-Centric) were used to test the proposed method. The fully sampled dynamic wash-out images were retrospectively under-sampled (acceleration-factors (AF) of 10/14) using a varying-sampling-pattern in the wash-out direction. Mean fractional-ventilation maps using Deninger's and SEM-based approaches were generated. The mean fractional-ventilation-values generated for the fully sampled k-space case using the Deninger method were not significantly different from other fractional-ventilation-values generated for the non-accelerated/accelerated data using both Deninger and SEM methods (p > 0.05 for all cases/gases). We demonstrated the feasibility of the SEM-based approach using retrospective under-sampling, mimicking AF = 10/14 in a small-animal-cohort from the previously reported dynamic-lung studies. A pixel-by-pixel comparison of the Deninger-derived and SEM-derived fractional-ventilation-estimates obtained for AF = 10/14 (≤16% difference) has confirmed that even at AF = 14, the accuracy of the estimates is high enough to consider this method for prospective measurements.

Imaging biomarkers of lung ventilation in interstitial lung disease from 129Xe and oxygen enhanced 1H MRI

PURPOSE

To compare imaging biomarkers from hyperpolarised ¹²⁹Xe ventilation MRI and dynamic oxygen-enhanced MRI (OE-MRI) with standard pulmonary function tests (PFT) in interstitial lung disease (ILD) patients. To evaluate if biomarkers can separate ILD subtypes and detect early signs of disease resolution or progression.

STUDY TYPE

Prospective longitudinal.

POPULATION

Forty-one ILD (fourteen idiopathic pulmonary fibrosis (IPF), eleven hypersensitivity pneumonitis (HP), eleven drug-induced ILD (DI-ILD), five connective tissue disease related-ILD (CTD-ILD)) patients and ten healthy volunteers imaged at visit 1. Thirty-four ILD patients completed visit 2 (eleven IPF, eight HP, ten DIILD, five CTD-ILD) after 6 or 26 weeks.

FIELD STRENGTH/SEQUENCE

MRI was performed at 1.5 T, including inversion recovery T₁ mapping, dynamic MRI acquisition with varying oxygen levels, and hyperpolarised ¹²⁹Xe ventilation MRI. Subjects underwent standard spirometry and gas transfer testing.

ASSESSMENT

Five ¹H MRI and two ¹²⁹Xe MRI ventilation metrics were compared with spirometry and gas transfer measurements.

STATISTICAL TEST

To evaluate differences at visit 1 among subgroups: ANOVA or Kruskal-Wallis rank tests with correction for multiple comparisons. To assess the relationships between imaging biomarkers, PFT, age and gender, at visit 1 and for the change between visit 1 and 2: Pearson correlations and multilinear regression models.

RESULTS

The global PFT tests could not distinguish ILD subtypes. Percentage ventilated volumes were lower in ILD patients than in HVs when measured with ¹²⁹Xe MRI (HV 97.4 ± 2.6, CTD-ILD: 91.0 ± 4.8 p = 0.017, DI-ILD 90.1 ± 7.4 p = 0.003, HP 92.6 ± 4.0 p = 0.013, IPF 88.1 ± 6.5 p < 0.001), but not with OE-MRI. ¹²⁹Xe reported more heterogeneous ventilation in DI-ILD and IPF than in HV, and OE-MRI reported more heterogeneous ventilation in DI-ILD and IPF than in HP or CTD-ILD. The longitudinal changes reported by the imaging biomarkers did not correlate with the PFT changes between visits.

DATA CONCLUSION

Neither ¹²⁹Xe ventilation nor OE-MRI biomarkers investigated in this study were able to differentiate between ILD subtypes, suggesting that ventilation-only biomarkers are not indicated for this task. Limited but progressive loss of ventilated volume as measured by ¹²⁹Xe-MRI may be present as the biomarker of focal disease progresses. OE-MRI biomarkers are feasible in ILD patients and do not correlate strongly with PFT. Both OE-MRI and ¹²⁹Xe MRI revealed more spatially heterogeneous ventilation in DI-ILD and IPF.

The evolving role of radiological imaging in cystic fibrosis

PURPOSE OF REVIEW

Radiological imaging has a crucial role in pulmonary evaluation in cystic fibrosis (CF), having been shown to be more sensitive than pulmonary function testing at detecting structural lung changes. The present review summarizes the latest published information on established and evolving pulmonary imaging techniques for assessing people with this potentially life-limiting disorder.

RECENT FINDINGS

Chest computed tomography (CT) has taken over the predominant role of chest radiography in many centres for the initial assessment and surveillance of CF lung disease. However, several emerging techniques offer a promising means of pulmonary imaging using less ionizing radiation. This is of particular importance given these patients tend to require repeated imaging throughout their lives from a young age. Such techniques include ultra-low-dose CT, tomosynthesis, dynamic radiography and magnetic resonance imaging. In addition, deep-learning algorithms are anticipated to improve diagnostic accuracy.

SUMMARY

The recent introduction of triple-combination CF transmembrane regulator therapy has put further emphasis on the need for sensitive methods of monitoring treatment response to allow for early adaptation of treatment regimens in order to limit irreversible lung damage. Further research is needed to establish how emerging imaging techniques can contribute to this safely and effectively.

Novel imaging techniques for cystic fibrosis lung disease

With an increasing number of patients with cystic fibrosis (CF) receiving highly effective CFTR (cystic fibrosis transmembrane regulator protein) modulator therapy, particularly at a young age, there is an increasing need to identify imaging tools that can detect and regionally visualize mild CF lung disease and subtle changes in disease state. In this review, we discuss the latest developments in imaging modalities for both structural and functional imaging of the lung available to CF clinicians and researchers, from the widely available, clinically utilized imaging methods for assessing CF lung disease-chest radiography and computed tomography-to newer techniques poised to become the next phase of clinical tools-structural/functional proton and hyperpolarized gas magnetic resonance imaging (MRI). Finally, we provide a brief discussion of several newer lung imaging techniques that are currently available only in selected research settings, including chest tomosynthesis, and fluorinated gas MRI. We provide an update on the clinical and/or research status of each technique, with a focus on sensitivity, early disease detection, and possibilities for monitoring treatment efficacy.

Reproducibility of 19 F-MR ventilation imaging in healthy volunteers

Purpose

To assess the reproducibility of percentage ventilated lung volume (%VV) measurements in healthy volunteers acquired by fluorine (¹⁹F)‐MRI of inhaled perfluoropropane, implemented at two research sites.

Methods

In this prospective, ethically approved study, 40 healthy participants were recruited (May 2018‐June 2019) to one of two research sites. Participants underwent a single MRI scan session on a 3T scanner, involving periodic inhalation of a 79% perfluoropropane/21% oxygen gas mixture. Each gas inhalation session lasted about 30 seconds, consisting of three deep breaths of gas followed by a breath‐hold. Four ¹⁹F‐MR ventilation images were acquired per participant, each separated by approximately 6 minutes. The value of %VV was determined by registering separately acquired ¹H images to ventilation images before semi‐automated image segmentation, performed independently by two observers. Reproducibility of %VV measurements was assessed by components of variance, intraclass correlation coefficients, coefficients of variation (CoV), and the Dice similarity coefficient.

Results

The MRI scans were well tolerated throughout, with no adverse events. There was a high degree of consistency in %VV measurements for each participant (CoV_observer1 = 0.43%; CoV_observer2 = 0.63%), with overall precision of %VV measurements determined to be within ± 1.7% (95% confidence interval). Interobserver agreement in %VV measurements revealed a high mean Dice similarity coefficient (SD) of 0.97 (0.02), with only minor discrepancies between observers.

Conclusion

We demonstrate good reproducibility of %VV measurements in a group of healthy participants using ¹⁹F‐MRI of inhaled perfluoropropane. Our methods have been successfully implemented across two different study sites, supporting the feasibility of performing larger multicenter clinical studies.

MR properties of 19 F C3 F8 gas in the lungs of healthy volunteers: T 2 ∗ and apparent diffusion coefficient at 1.5T and T 2 ∗ at 3T

PURPOSE

To measure the transverse relaxation time ( T 2 ∗ ) and apparent diffusion coefficient (ADC) of ¹⁹ F-C₃ F₈ gas in vivo in human lungs at 1.5T and 3T, and to determine the representative distribution of values of these parameters in a cohort of healthy volunteers.

METHODS

Mapping of ADC at lung inflation levels of functional residual capacity (FRC) and total lung capacity (TLC) was performed with inhaled ¹⁹ F-C₃ F₈ (eight subjects) and ¹²⁹ Xe (six subjects) at 1.5T. T 2 ∗ mapping with ¹⁹ F-C₃ F₈ was performed at 1.5T (at FRC and TLC) for 8 subjects and at 3T (at TLC for seven subjects).

RESULTS

At both FRC and TLC, the ¹⁹ F-C₃ F₈ ADC was smaller than the free diffusion coefficient demonstrating airway microstructural diffusion restriction. From FRC to TLC, the mean ADC significantly increased from 1.56 mm² /s to 1.83 mm² /s (P = .0017) for ¹⁹ F-C₃ F_8, and from 2.49 mm² /s to 3.38 mm² /s (P = .0015) for ¹²⁹ Xe. The posterior-to-anterior gradient in ADC for FRC versus TLC in the superior half of the lungs was measured as 0.0308 mm² /s per cm versus 0.0168 mm² /s per cm for ¹⁹ F-C₃ F₈ and 0.0871 mm² /s per cm versus 0.0326 mm² /s per cm for ¹²⁹ Xe. A consistent distribution of ¹⁹ F-C₃ F₈ T 2 ∗ values was observed in the lungs, with low values observed near the diaphragm and large pulmonary vessels. The mean T 2 ∗ across volunteers was 4.48 ms at FRC and 5.33 ms at TLC for 1.5T, and 3.78 ms at TLC for 3T.

CONCLUSION

In this feasibility study, values of physiologically relevant parameters of lung microstructure measurable by MRI ( T 2 ∗ , and ADC) were established for C₃ F₈ in vivo lung imaging in healthy volunteers.

Evaluation of fluorine-19 magnetic resonance imaging of the lungs using octafluorocyclobutane in a rat model

PURPOSE

To test octafluorocyclobutane (OFCB) as an inhalation contrast agent for fluorine-19 MRI of the lung, and to compare the image quality of OFCB scans with perfluoropropane (PFP) scans THEORY AND METHODS: After normalizing for the number of signal averages, a theoretical comparison between the OFCB signal-to-noise ratio (SNR) and PFP SNR predicted the average SNR advantage of 90% using OFCB during gradient echo imaging. The OFCB relaxometry was conducted using single-voxel spectroscopy and spin-echo imaging. A comparison of OFCB and PFP SNRs was performed in vitro and in vivo. Five healthy Sprague-Dawley rats were imaged during single breath-hold and continuous breathing using a Philips Achieva 3.0T MRI scanner (Philips, Andover, MA). The scan time was constant for both gases. Statistical comparison between PFP and OFCB scans was conducted using a paired t test and by calculating the Bayes factor.

RESULTS

Spin-lattice (T₁ ) and effective spin-spin ( T 2 ∗ ) relaxation time constants of the pure OFCB gas were determined as 28.5 ± 1.2 ms and 10.5 ± 1.8 ms, respectively. Mixing with 21% of oxygen decreased T₁ by 30% and T 2 ∗ by 20%. The OFCB in vivo images showed 73% higher normalized SNR on average compared with images acquired using PFP. The statistical significance was shown by both paired t test and calculated Bayes factors. The experimental results agree with theoretical calculations within the error of the relaxation parameter measurements.

CONCLUSION

The quality of the lung images acquired using OFCB was significantly better compared with PFP scans. The OFCB images had higher a SNR and were artifact-free.

An 8-element Tx/Rx array utilizing MEMS detuning combined with 6 Rx loops for 19 F and 1 H lung imaging at 1.5T

PURPOSE

To firstly improve the attainable image SNR of ¹⁹ F and ¹ H C₃ F₈ lung imaging at 1.5 tesla using an 8-element transmit/receive (Tx/Rx) flexible vest array combined with a 6-element Rx-only array, and to secondly evaluate microelectromechanical systems for switching the array elements between the 2 resonant frequencies.

METHODS

The Tx efficiency and homogeneity of the 8-element array were measured and simulated for ¹ H imaging in a cylindrical phantom and then evaluated for in vivo ¹⁹ F/¹ H imaging. The added improvement provided by the 6-element Rx-only array was quantified through simulation and measurement and compared to the ultimate SNR. It was verified through the measurement of isolation that microelectromechanical systems switches provided broadband isolation of Tx/Rx circuitry such that the ¹⁹ F tuned Tx/Rx array could be effectively used for both ¹⁹ F and ¹ H nuclei.

RESULTS

For ¹ H imaging, the measured Tx efficiency/homogeneity (mean ± percent SD; 6.79 μ T / kW ± 26 % ) was comparable to that simulated ( 7.57 μ T / kW ± 20 % ). The 6 additional Rx-only loops increased the mean Rx sensitivity when compared to the 8-element array by a factor of 1.41× and 1.45× in simulation and measurement, respectively. In regions central to the thorax, the simulated SNR of the 14-element array achieves ≥70% of the ultimate SNR when including noise from the matching circuits and preamplifiers. A measured microelectromechanical systems switching speed of 12 µs and added minimum 22 dB of isolation between Tx and Rx were sufficient for Tx/Rx switching in this application.

CONCLUSION

The described single-tuned array driven at ¹⁹ F and ¹ H, utilizing microelectromechanical systems technology, provides excellent results for ¹⁹ F and ¹ H dual-nuclear lung ventilation imaging.

19 F MRI of human lungs at 0.5 Tesla using octafluorocyclobutane

PURPOSE

The aim of this study was to demonstrate the feasibility of fluorine-19 (¹⁹ F) MRI of the human lungs using octafluorocyclobutane (OFCB, C₄ F₈ ). This gas has 8 magnetically equivalent fluorine nuclei and relatively long T₁ and T₂ (˜50 ms), which render it suitable as an MRI contrast agent. Previous experiments in small laboratory animals showed that OFCB could be successfully used as an alternative to the gases often used for ¹⁹ F MRI (sulfur hexafluoride and perfluoropropane).

METHODS

One male volunteer participated in this study. Immediately before an MRI scan, the volunteer inhaled the gas mixture-80% OFCB with 20% oxygen-and held his breath. Experiments were performed on a 0.5T whole-body MR scanner with a customized transmit-receive coil tuned at ¹⁹ F frequency. Fast spin echo in 2D and 3D modes was used for image acquisition. 2D images were obtained with in-plane resolution of 10 × 10 mm² without slice selection. 3D images were obtained with the voxel size of 10 × 10 × 30 mm² . Breath-hold duration was 20 s for 2D and 40 s for 3D imaging, respectively.

RESULTS

Anatomically consistent ¹⁹ F MR images of the human lungs were obtained with SNR around 50 in 2D mode and 20 in 3D mode. 3D volumetric images of the lungs were reconstructed and provided physiologically reasonable volume estimates.

CONCLUSION

The application of OFCB enables informative ¹⁹ F lung imaging even at low magnetic field strengths. The OFCB gas shows promise as an inhalable contrast agent for fluorine lung MRI and has a potential for clinical translation.

1 H-guided reconstruction of 19 F gas MRI in COPD patients

PURPOSE

To reduce acquisition time and improve image quality and robustness of ventilation assessment in a single breath-hold using ¹ H-guided reconstruction of fluorinated gas (¹⁹ F) MRI.

METHODS

Reconstructions constraining total variation in the image domain, L1 norm in the wavelet domain, and directional total variation between ¹⁹ F and ¹ H images were compared in order to accelerate ¹⁹ F ventilation imaging using retrospectively undersampled data from a healthy volunteer. Using the optimal constrained reconstruction in 8 patients with chronic obstructive pulmonary disease (16-seconds breath-hold), ventilation maps of various acceleration factors (2-fold to 13-fold) were compared with maps of the full data set using the Dice coefficient, difference in volume defect percentage and overlap percentage, as well as hyperpolarized ¹²⁹ Xe gas MRI.

RESULTS

The reconstruction constraining total variation and directional total variation simultaneously performed best in the healthy volunteer (RMS error = 0.07, structural similarity index = 0.77) for a measurement time of 2 seconds. Using the same reconstruction in the patients with chronic obstructive pulmonary disease, the Dice coefficient of defect volumes was 0.86 ± 0.05, the mean difference in volume defect percentage was -1.0 ± 1.7 percentage points, and the overlap percentage was 87% ± 2% for a measurement time of 6 seconds. Between volume defect percentage of ¹⁹ F and ¹²⁹ Xe, a linear correlation (r = 0.75; P = .03) was found, with ¹⁹ F volume defect percentage being significantly higher (mean difference = 11%; P = .04).

CONCLUSION

¹ H-guided reconstruction of pulmonary ¹⁹ F gas MRI enables reduction of acquisition time while maintaining image quality and robustness of functional parameters.

Dynamic susceptibility contrast 19 F-MRI of inhaled perfluoropropane: a novel approach to combined pulmonary ventilation and perfusion imaging

PURPOSE

To assess alveolar perfusion by applying dynamic susceptibility contrast MRI to ¹⁹ F-MRI of inhaled perfluoropropane (PFP). We hypothesized that passage of gadolinium-based contrast agent (GBCA) through the pulmonary microvasculature would reduce magnetic susceptibility differences between water and gas components of the lung, elevating the T 2 ∗ of PFP.

METHODS

Lung-representative phantoms were constructed of aqueous PFP-filled foams to characterize the impact of aqueous/gas phase magnetic susceptibility differences on PFP T 2 ∗ . Aqueous phase magnetic susceptibility was modulated by addition of different concentrations of GBCA. In vivo studies were performed to measure the impact of intravenously administered GBCA on the T 2 ∗ of inhaled PFP in mice (7.0 Tesla) and in healthy volunteers (3.0 Tesla).

RESULTS

Perfluoropropane T 2 ∗ was sensitive to modulation of magnetic susceptibility difference between gas and water components of the lung, both in phantom models and in vivo. Negation of aqueous/gas phase magnetic susceptibility difference was achieved in lung-representative phantoms and in mice, resulting in a ~2 to 3× elevation in PFP T 2 ∗ (3.7 to 8.5 ms and 0.7 to 2.6 ms, respectively). Human studies demonstrated a transient elevation of inhaled PFP T 2 ∗ (1.50 to 1.64 ms) during passage of GBCA bolus through the lung circulation, demonstrating sensitivity to lung perfusion.

CONCLUSION

We demonstrate indirect detection of a GBCA in the pulmonary microvasculature via changes to the T 2 ∗ of gas phase PFP within directly adjacent alveoli. This approach holds potential for assessing alveolar perfusion by dynamic susceptibility contrast ¹⁹ F-MRI of inhaled PFP, with concurrent assessment of lung ventilation properties, relevant to lung physiology and disease.

Optimization of steady-state free precession MRI for lung ventilation imaging with 19 F C3 F8 at 1.5T and 3T

PURPOSE

To optimize 19 F imaging pulse sequences for perfluoropropane (C3 F8 ) gas human lung ventilation MRI considering intrinsic in vivo relaxation parameters at both 1.5T and 3T.

METHODS

Optimization of the imaging parameters for both 3D spoiled gradient (SPGR) and steady-state free precession (SSFP) 19 F imaging sequences with inhaled 79% C3 F8% and 21% oxygen was performed. Phantom measurements were used to validate simulations of SNR. In vivo parameter mapping and sequence optimization and comparison was performed by imaging the lungs of a healthy adult volunteer. T1 and T2* mapping was performed in vivo to optimize sequence parameters for in vivo lung MRI. The performance of SSFP and SPGR was then evaluated in vivo at 1.5T and 3T.

RESULTS

The in vivo T2* of C3 F8 was shown to be dependent upon lung inflation level (2.04 ms ± 36% for residual volume and 3.14 ms ± 28% for total lung capacity measured at 3T), with lower T2* observed near the susceptibility interfaces of the diaphragm and around pulmonary blood vessels. Simulation and phantom measurements indicate that a factor of ~2-3 higher SNR can be achieved with SSFP when compared with optimized SPGR. In vivo lung imaging showed a 1.7 factor of improvement in SNR achieved at 1.5T, while the theoretical improvement at 3T was not attained due to experimental SAR constraints, shorter in vivo T1 , and B0 inhomogeneity.

CONCLUSION

SSFP imaging provides increased SNR in lung ventilation imaging of C3 F8 demonstrated at 1.5T with optimized SSFP similar to the SNR that can be obtained at 3T with optimized SPGR.