0002

3D Magnetic Resonance Spirometry
Tanguy Boucneau1,2,3, Brice Fernandez4, Peder Larson5, Luc Darrasse1,2,3, and Xavier Maître1,2,3

1IR4M, CNRS, Orsay, France, 2IR4M, Univ. Paris-Sud, Orsay, France, 3IR4M, Université Paris-Saclay, Orsay, France, 4Applications & Workflow, GE Healthcare, Buc, France, 5Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States

Synopsis

Spirometry is a routine tool in pulmonology to challenge lung function. It is sensitive and specific to several common lung diseases. Nevertheless, spirometry provides a global measurement at the mouth that only characterizes the pulmonary response as a whole, and cannot specifically address regional affections. We developed a 3D MRI-based technique to non-invasively perform local spirometry throughout the lung. The final outcome is a 3D map with voxel-wise flow-volume loops over the organ. The potential of this technique is demonstrated by its sensitivity to normal, thoracic, and diaphragmatic breathings and its related regional specificity in a healthy subject.

Introduction

Spirometry is today the most frequently-used tool for pulmonary function testing. It usually consists in measuring, through a mouthpiece, air volumes and flows in and out the lung along several forced respiratory cycles1. The main outcome of this measurement is the flow-volume loop, a graphical representation of the recorded air flow as a function of the air volume variation along a full respiratory cycle. The shape of the loop is analyzed by the pulmonologist to characterize and diagnose different families of lung diseases. However, this loop is obtained for the whole organ and it does not provide any local information, which would differentiate affected and healthy regions within the organ.

Here we propose 3D MR spirometry to map local flow-volume loops within the lung. It sets the grounds for regional spirometry and comparative studies of the pulmonary function throughout the lung.

Principles

From retrospectively-gated 3D thoracic images, deformation fields can be inferred over the set of selected gates. The jacobian of those deformation fields, namely the determinant of the jacobian matrices calculated for each voxel, can then be processed to keep track of the relative volume expansions at the voxel scale throughout the lung. This metric is proportional to the volume ratio between corresponding pieces of lung tissue before image registration. If we assume that any pulmonary volume expansion results from local air admittance or expulsion, the air flow evolution along the respiratory cycle can be computed voxel-by-voxel from the first time derivative of those volume expansions over the gates. The full processing pipeline is illustrated in Figure 1.

Methods

MR acquisitions were performed in a GE Signa PET/MR 3.0 T on a human volunteer, lying supine, along either normal, thoracic, or diaphragmatic breathings. Data sets were acquired by a 3D radial UTE sequence2,3, with an imaging matrix of (212×212×142) isotropic voxels of 1.5 mm, TE = 12 µs, TR = 2 ms, BW = ±100 kHz and 11 min scan time. A 30-channel thoracic coil was used as receiver. The respiratory motion was monitored by self-navigation upon the k-space center4 so 32 temporal gates were defined along the respiratory cycle with respect to the gate for the end of expiration, taken as time reference. Retrospectively self- and soft-gated5 images were reconstructed with a l1-ESPIRiT algorithm provided by the BART toolbox6. On the reference image, at the volunteer’s current functional residual capacity (FRC), a 3D mask of the lung was processed.

Each gated 3D image was non-rigidly registered onto the FRC image with the Elastix toolbox7. The resulting 31 3D deformation fields were filtered8 along the respiratory cycle to ensure spatio-temporal smoothing and continuity.

For each voxel of the FRC image, 31 estimates of volume expansion and flow were thus obtained along a gated-average respiratory cycle of the pseudo-periodic lung motion during the MR acquisition. It led to the representation of voxel-by-voxel flow-volume loops over the entire lung.

Results

Figure 2 represents volume expansion and flow maps along a gated-average respiratory cycle. Figure 3 shows the resulting 3D MR spirometric image obtained over thoracic breathing. To compare the three types of breathing, flow-volume loop area maps and projections on the anterior-posterior axis are represented on Figure 4.

Discussion

Volume expansions and air flows are noticeably non-uniform over the lung. Particularly-low values are revealed in regions where the pulmonary alveolar density is expected to be weak (large blood vessels, large airways …). With normal breathing as reference, flow-volume loop areas are relatively larger in the anterior region of the lung for a thoracic breathing and relatively larger in the posterior region for diaphragmatic breathing. Area maps could interestingly be inferred for normal breathing by combining maps for thoracic and diaphragmatic breathings as normal breathing advantageously makes use of both the diaphragm and the intercostal muscles.

Previous studies were undertaken several years ago to perform MR-based spirometry9–11. Nevertheless, to our knowledge, this is the first time this technique is developed in 3D, from a direct and quantitative measurement of regional lung deformation in the three dimensions of space and without additional image contrast modification.

Conclusion

The feasibility and relevance of 3D Magnetic Resonance Spirometry were assessed. The preliminary study carried for a healthy volunteer on various types of breathing corroborates the expected enriched data information mapping could bring to pulmonology beyond standard clinical pulmonary function tests, usually performed in forced respiration. Regional quantification and differentiation of flow-volume loops during normal breathing should support more sensitive and specific diagnoses of lung diseases, which are obliterated by forced tests and global information integrated over healthy and affected pulmonary regions.

Acknowledgements

PET/MR platform affiliated to the France Life Imaging network (grant ANR-11-INBS-0006). We sincerely acknowledge Dr. Jean-Luc Gennisson, Dr. Thomas Similowski, Dr. Pierantonio Laveneziana and Dr. Christian Straus for fruitful discussions on this project.

References

1. Miller, M. R. et al. Standardisation of spirometry. Eur. Respir. J. 26, 319–338 (2005).

2. Tyler, D. J., Robson, M. D., Henkelman, R. M., Young, I. R. & Bydder, G. M. Magnetic resonance imaging with ultrashort TE (UTE) PULSE sequences: Technical considerations. J. Magn. Reson. Imaging 25, 279–289 (2007).

3. Burris, N. S. et al. Detection of Small Pulmonary Nodules with Ultrashort Echo Time Sequences in Oncology Patients by Using a PET/MR System. Radiology 278, 239–246 (2015).

4. Tibiletti, M. et al. Respiratory self-gated 3DUTE for lung imaging in small animal MRI. Magn. Reson. Med. 78, 739–745 (2017).

5. Jiang, W. et al. Motion robust high resolution 3D free-breathing pulmonary MRI using dynamic 3D image self-navigator. Magn. Reson. Med. (2017). doi:10.1002/mrm.26958

6. Martin Uecker. mrirecon/bart: version 0.4.03. (Zenodo, 2018). doi:10.5281/zenodo.1215477

7. Klein, S., Staring, M., Murphy, K., Viergever, M. A. & Pluim, J. P. W. lt;emphasis emphasistype="mono" gt;elastix lt;/emphasis gt;: A Toolbox for Intensity-Based Medical Image Registration. IEEE Trans. Med. Imaging 29, 196–205 (2010).

8. smoothn - File Exchange - MATLAB Central. Available at: https://www.mathworks.com/matlabcentral/fileexchange/25634. (Accessed: 24th October 2018)

9. Voorhees, A., An, J., Berger, K. I., Goldring, R. M. & Chen, Q. Magnetic resonance imaging-based spirometry for regional assessment of pulmonary function. Magn. Reson. Med. 54, 1146–1154 (2005).

10. Voskrebenzev, A. et al. Imaging-Based Spirometry in Chronic Obstructive Pulmonary Disease (COPD) Patients using Phase Resolved Functional Lung Imaging (PREFUL). in 26, (2018).

11. Feng, L. et al. Simultaneous Evaluation of Lung Anatomy and Ventilation Using 4D Respiratory-Motion-Resolved Ultrashort Echo Time Sparse MRI. J. Magn. Reson. Imaging 0,

12. Milic-Emili, J., Henderson, J. A., Dolovich, M. B., Trop, D. & Kaneko, K. Regional distribution of inspired gas in the lung. J. Appl. Physiol. 21, 749–759 (1966).

Figures

Flow chart describing the processing pipeline of 3D Magnetic Resonance Spirometry – from top left-hand corner to bottom right-hand corner.

Regional relative volume expansion and air flow maps along a gated-average respiratory cycle. These maps were processed from 3D MR data acquired while the volunteer was performing a thoracic breathing. Only 6 out of the 32 defined gates are represented here. As expected, we observe in the entire lung increasing values of regional volume expansion (positive value of flow) during inspiration and a decreasing value (negative value of flow) during expiration. Moreover, we observe higher values of regional volume expansions in regions close to the diaphragm in comparison to apical regions, as demonstrated in previous studies12.

3D MR spirometry presented along three orthogonal slices for a human volunteer performing a thoracic breathing. On the bottom right-hand corner, a diagram represents the convention along which flow-volume loops are usually represented in spirometry and the color conventions used for MR spirometry maps. The flow-volume loop for the entire lung, computed as the mean loop over a 3D lung mask, is reproduced in green over the image for reference.

Sagittal slices of normalized flow-volume loop area maps and projections of these 3D area maps on the anterior-posterior axis for a human volunteer performing normal, thoracic, or diaphragmatic breathings. Normalization was performed separately over each slice for sagittal views of area maps and over the entire lung volume for area maps projections to compensate for global tidal volume and air flow differences between the three types of breathing.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
0002