Deep Learning Approaches for Quantifying Ventilation Defects in Hyperpolarized Gas Magnetic Resonance Imaging of the Lung: A Review

This paper provides an in-depth overview of Deep Neural Networks and their application in the segmentation and analysis of lung Magnetic Resonance Imaging (MRI) scans, specifically focusing on hyperpolarized gas MRI and the quantification of lung ventilation defects. An in-depth understanding of Deep Neural Networks is presented, laying the groundwork for the exploration of their use in hyperpolarized gas MRI and the quantification of lung ventilation defects. Five distinct studies are examined, each leveraging unique deep learning architectures and data augmentation techniques to optimize model performance. These studies encompass a range of approaches, including the use of 3D Convolutional Neural Networks, cascaded U-Net models, Generative Adversarial Networks, and nnU-net for hyperpolarized gas MRI segmentation. The findings highlight the potential of deep learning methods in the segmentation and analysis of lung MRI scans, emphasizing the need for consensus on lung ventilation segmentation methods.

A 3D stack-of-spirals approach for rapid hyperpolarized 129 Xe ventilation mapping in pediatric cystic fibrosis lung disease

PURPOSE

To demonstrate the feasibility of a rapid 3D stack-of-spirals (3D-SoS) imaging acquisition for hyperpolarized ¹²⁹ Xe ventilation mapping in healthy pediatric participants and pediatric cystic fibrosis (CF) participants, in comparison to conventional Cartesian multislice (2D) gradient-recalled echo (GRE) imaging.

METHODS

The 2D-GRE and 3D-SoS acquisitions were performed in 13 pediatric participants (5 healthy, 8 CF) during separate breath-holds. Images from both sequences were compared on the basis of ventilation defect percent (VDP) and other measures of image similarity. The nadir of transient oxygen saturation (SpO₂ ) decline due to xenon breath-holding was measured with pulse oximetry, and expressed as a percent change relative to baseline.

RESULTS

¹²⁹ Xe ventilation images were acquired in a breath-hold of 1.2-1.8 s with the 3D-SoS sequence, compared to 6.2-8.8 s for 2D-GRE. Mean ± SD VDP measures for 2D-GRE and 3D-SoS sequences were 5.02 ± 1.06% and 5.28 ± 1.08% in healthy participants, and 18.05 ± 8.26% and 18.75 ± 6.74% in CF participants, respectively. Across all participants, the intraclass correlation coefficient of VDP measures for both sequences was 0.98 (95% confidence interval: 0.94-0.99). The percent change in SpO₂ was reduced to -2.1 ± 2.7% from -5.2 ± 3.5% with the shorter 3D-SoS breath-hold.

CONCLUSION

Hyperpolarized ¹²⁹ Xe ventilation imaging with 3D-SoS yielded images approximately five times faster than conventional 2D-GRE, reducing SpO₂ desaturation and improving tolerability of the xenon administration. Analysis of VDP and other measures of image similarity demonstrate excellent agreement between images obtained with both sequences. 3D-SoS holds significant potential for reducing the acquisition time of hyperpolarized ¹²⁹ Xe MRI, and/or increasing spatial resolution while adhering to clinical breath-hold constraints.

Comparison of single breath hyperpolarized 129 Xe MRI with dynamic 19 F MRI in cystic fibrosis lung disease

Purpose

To quantitatively compare dynamic ¹⁹F and single breath hyperpolarized ¹²⁹Xe MRI for the detection of ventilation abnormalities in subjects with mild cystic fibrosis (CF) lung disease.

Methods

Ten participants with stable CF and a baseline FEV1 > 70% completed a single imaging session where dynamic ¹⁹F and single breath ¹²⁹Xe lung ventilation images were acquired on a 3T MRI scanner. Ventilation defect percentages (VDP) values between ¹⁹F early‐breath, ¹⁹F maximum‐ventilation, ¹²⁹Xe low‐resolution, and ¹²⁹Xe high‐resolution images were compared. Dynamic ¹⁹F images were used to determine gas wash‐in/out rates in regions of ventilation congruency and mismatch between ¹²⁹Xe and ¹⁹F.

Results

VDP values from high‐resolution ¹²⁹Xe images were greater than from low‐resolution images (P = .001), although these values were significantly correlated (r = 0.68, P = .03). Early‐breath ¹⁹F VDP and max‐vent ¹⁹F VDP also showed significant correlation (r = 0.75, P = .012), with early‐breath ¹⁹F VDP values being significantly greater (P < .001). No correlation in VDP values were detected between either ¹⁹F method or high‐res ¹²⁹Xe images. In addition, the location and volume of ventilation defects were often different when comparing ¹²⁹Xe and ¹⁹F images from the same subject. Areas of ventilation congruence displayed the expected ventilation kinetics, while areas of ventilation mismatch displayed abnormally slow gas wash‐in and wash‐out.

Conclusion

In CF subjects, ventilation abnormalities are identified by both ¹⁹F and HP ¹²⁹Xe imaging. However, these ventilation abnormalities are not entirely congruent. ¹⁹F and HP ¹²⁹Xe imaging provide complementary information that enable differentiation of normally ventilated, slowly ventilated, and non‐ventilated regions in the lungs.

Pulmonary ventilation imaging in asthma and cystic fibrosis using oxygen-enhanced 3D radial ultrashort echo time MRI

BACKGROUND

A previous study demonstrated the feasibility of using 3D radial ultrashort echo time (UTE) oxygen-enhanced MRI (UTE OE-MRI) for functional imaging of healthy human lungs. The repeatability of quantitative measures from UTE OE-MRI needs to be established prior to its application in clinical research.

PURPOSE

To evaluate repeatability of obstructive patterns in asthma and cystic fibrosis (CF) with UTE OE-MRI with isotropic spatial resolution and full chest coverage.

STUDY TYPE

Volunteer and patient repeatability.

POPULATION

Eighteen human subjects (five asthma, six CF, and seven normal subjects).

FIELD STRENGTH/SEQUENCE

Respiratory-gated free-breathing 3D radial UTE (80 μs) sequence at 1.5T.

ASSESSMENT

Two 3D radial UTE volumes were acquired sequentially under normoxic and hyperoxic conditions. A subset of subjects underwent repeat acquisitions on either the same day or ≤15 days apart. Asthma and CF subjects also underwent spirometry. A workflow including deformable registration and retrospective lung density correction was used to compute 3D isotropic percent signal enhancement (PSE) maps. Median PSE (MPSE) and ventilation defect percent (VDP) of the lung were measured from the PSE map.

STATISTICAL TESTS

The relations between MPSE, VDP, and spirometric measures were assessed using Spearman correlations. The test-retest repeatability was evaluated using Bland-Altman analysis and intraclass correlation coefficients (ICC).

RESULTS

Ventilation measures in normal subjects (MPSE = 8.0%, VDP = 3.3%) were significantly different from those in asthma (MPSE = 6.0%, P = 0.042; VDP = 21.7%, P = 0.018) and CF group (MPSE = 4.5%, P = 0.0006; VDP = 27.2%, P = 0.002). MPSE correlated significantly with forced expiratory lung volume in 1 second percent predicted (ρ = 0.72, P = 0.017). The ICC of the test-retest VDP and MPSE were both ≥0.90. In all subject groups, an anterior/posterior gradient was observed with higher MPSE and lower VDP in the posterior compared to anterior regions (P ≤ 0.0021 for all comparisons).

DATA CONCLUSION

3D radial UTE OE-MRI supports quantitative differentiation of diseased vs. healthy lungs using either whole lung VDP or MPSE with excellent test-retest repeatability.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2018;47:1287-1297.

Pulmonary functional magnetic resonance imaging: asthma temporal-spatial maps

RATIONALE AND OBJECTIVES

Hyperpolarized (3)He magnetic resonance imaging (MRI) previously revealed the temporal and spatial heterogeneity of ventilation defects in asthmatics, but these findings have not been used in treatment studies or to guide personalized therapy. Our objective was to exploit the temporal and spatial information inherent to (3)He MRI and develop image processing methods to generate pulmonary ventilation temporal-spatial maps that could be used to measure, optimize, and guide asthma therapy.

MATERIALS AND METHODS

In this proof-of-concept study, seven asthmatics provided written informed consent to an approved protocol and underwent spirometry and (3)He MRI on three occasions, each 5 ± 2 days apart. A registration and segmentation pipeline was developed to generate three-dimensional, temporal-spatial, pulmonary function maps. Briefly, (3)He ventilation images were segmented to generate ventilation masks that were coregistered and voxels classified according to their temporal behavior. This enabled the regional mapping of temporally persistent and intermittent ventilation defects that were normalized to the (1)H MRI thoracic cavity volume to generate persistent ventilation defect percent (VDPP) and intermittent ventilation defect percent (VDPI).

RESULTS

(3)He temporal-spatial pulmonary function maps identified temporally persistent and intermittent ventilation defects. VDP(I) was significantly greater in the posterior (P = .04) and inferior (P = .04) lung as compared to the anterior and superior lung. Persistent and intermittent ventilation defect percent were strongly correlated with forced expiratory volume in one second/forced vital capacity (VDP(P): r = -0.87, P = .01; VDP(I): r = -0.96, P = .0008).

CONCLUSIONS

Temporal-spatial pulmonary maps generated from (3)He MRI can be used to quantify temporally persistent and intermittent ventilation defects as asthma intermediate end points and targets for therapy.

Hyperpolarized (3) He and (129) Xe MRI: differences in asthma before bronchodilation

PURPOSE

To compare hyperpolarized helium-3 ((3) He) and xenon-129 ((129) Xe) MRI in asthmatics before and after salbutamol inhalation.

MATERIALS AND METHODS

Seven asthmatics provided written informed consent and underwent spirometry, plethysmography, and MRI before and after salbutamol inhalation. (3) He and (129) Xe ventilation defect percent (VDP) and ventilation coefficient of variation (COV) were measured. To characterize the airways spatially related to ventilation defects, wall area percent (WA%) and lumen area (LA) were evaluated for two subjects who had thoracic x-ray computed tomography (CT) acquired 1 year before MRI.

RESULTS

Before salbutamol inhalation, (129) Xe VDP (8 ± 5%) was significantly greater than (3) He VDP (6 ± 5%, P = 0.003). Post-salbutamol, there was a significant improvement in both (129) Xe (5 ± 4%, P < 0.0001) and (3) He (4 ± 3%, P = 0.001) VDP, and the improvement in (129) Xe VDP was significantly greater (P = 0.008). (129) Xe MRI COV (Pre: 0.309 ± 0.028, Post: 0.296 ± 0.036) was significantly greater than (3) He MRI COV (Pre: 0.282 ± 0.018, Post: 0.269 ± 0.024), pre- (P < 0.0001) and post-salbutamol (P < 0.0001) and the decrease in COV post-salbutamol was significant ((129) Xe, P = 0.002; (3) He, P < 0.0001). For a single asthmatic, a sub-segmental (129) Xe MRI ventilation defect that was visible only before salbutamol inhalation but not visible using (3) He MRI was spatially related to a remodeled fourth generation sub-segmental airway (WA% = 78%, LA = 2.9 mm(2) ).

CONCLUSION

In asthma, hyperpolarized (129) Xe MRI may help reveal ventilation abnormalities before bronchodilation that are not observed using hyperpolarized (3) He MRI.