Real-Time MRI With Deep Learning for Efficient Evaluation of Neuromuscular Breathing Impairment

Abstract

Efficient detection of breathing impairment is critical for treatment and prognosis in neuromuscular disorders. However, standard pulmonary function tests often yield ambiguous results. This prospective study evaluates whether advanced real-time MRI (RT-MRI) combined with deep learning-based image segmentation provides sensitive outcome measures for respiratory dysfunction in late-onset Pompe disease (LOPD), a model disease for diaphragmatic weakness. Eleven Pompe patients (mean age 52.2 years; 55% female) and 11 controls (mean age 50.9 years; 55% female) were included. RT-MRI with a temporal resolution of 50 ms, combined with U-Net-supported lung segmentation, revealed significantly reduced diaphragmatic motion in Pompe patients compared to controls and unmasked paradoxical diaphragmatic motion in Pompe patients (7 of 11). Reduced diaphragmatic sniff velocity and pathological diaphragmatic/thoracic synchronicity were detected in Pompe patients with still normal results in standard pulmonary function tests. Fatty involution of the diaphragm as quantified by fast T1 mapping correlated significantly with functional parameters from RT-MRI and pulmonary function tests. RT-MRI combined with deep learning-based lung segmentation offers novel biomarkers for early detection of respiratory muscle weakness. This new technique provides useful outcome measures for clinical care as well as treatment studies in patients with neuromuscular breathing impairment. The technique can also be used to characterize physiologic breathing patterns in healthy individuals.

Keywords

breathing pattern / convolutional neural network / diaphragm / dynamic imaging / quantitative MRI / respiratory muscle weakness

References

[1] M. Boentert, S. Wenninger, and S. VA, “Respiratory Involvement in Neuromuscular Disorders,” Current Opinion in Neurology 30, no. 5 (2017): 529–537.

[2] N. Patel, I. M. Howard, and A. Baydur, “Respiratory Considerations in Patients With Neuromuscular Disorders,” Muscle & Nerve 68, no. 2 (2023): 122–141.

[3] J. O. Benditt, “Respiratory Care of Patients with Neuromuscular Disease,” Respiratory Care 64, no. 6 (2019): 679–688.

[4] P. Laveneziana, A. Albuquerque, A. Aliverti, et al., “ERS Statement on Respiratory Muscle Testing at Rest and During Exercise,” European Respiratory Journal 53, no. 6 (2019): 1801214.

[5] J. A. Evans and W. A. Whitelaw, “The Assessment of Maximal Respiratory Mouth Pressures in Adults,” Respiratory Care 54, no. 10 (2009): 1348–1359.

[6] M. Georges, D. Nguyen-Baranoff, L. Griffon, et al., “Usefulness of Transcutaneous PCO2 to Assess Nocturnal Hypoventilation in Restrictive Lung Disorders,” Respirology 21, no. 7 (2016): 1300–1306.

[7] B. K. Smith, M. Corti, A. D. Martin, D. D. Fuller, and B. J. Byrne, “Altered Activation of the Diaphragm in Late-Onset Pompe Disease,” Respiratory Physiology & Neurobiology 222 (2016): 11–15.

[8] J. Spiesshoefer, C. Henke, S. D. Herkenrath, et al., “Noninvasive Prediction of Twitch Transdiaphragmatic Pressure: Insights From Spirometry, Diaphragm Ultrasound, and Phrenic Nerve Stimulation Studies,” Respiration 98, no. 4 (2019): 301–311.

[9] A. Fayssoil, A. Behin, A. Ogna, et al., “Diaphragm: Pathophysiology and Ultrasound Imaging in Neuromuscular Disorders,” Journal of Neuromuscular Diseases 5, no. 1 (2018): 1–10.

[10] J. G. Houston, M. Fleet, M. D. Cowan, and N. C. McMillan, “Comparison of Ultrasound With Fluoroscopy in the Assessment of Suspected Hemidiaphragmatic Movement Abnormality,” Clinical Radiology 50, no. 2 (1995): 95–98.

[11] M. Gaeta, E. Barca, P. Ruggeri, et al., “Late-Onset Pompe Disease (LOPD): Correlations Between Respiratory Muscles CT and MRI Features and Pulmonary Function,” Molecular Genetics and Metabolism 110, no. 3 (2013): 290–296.

[12] S. C. A. Wens, P. Ciet, A. Perez-Rovira, et al., “Lung MRI and Impairment of Diaphragmatic Function in Pompe Disease,” BMC Pulmonary Medicine 15 (2015): 54.

[13] K. Mogalle, A. Perez-Rovira, P. Ciet, et al., “Quantification of Diaphragm Mechanics in Pompe Disease Using Dynamic 3D MRI,” PLoS ONE 11, no. 7 (2016): e0158912.

[14] A. Mankodi, W. Kovacs, G. Norato, et al., “Respiratory Magnetic Resonance Imaging Biomarkers in Duchenne Muscular Dystrophy,” Annals of Clinical and Translational Neurology 4, no. 9 (2017): 655–662.

[15] C. A. Bishop, V. Ricotti, C. D. J. Sinclair, et al., “Semi-Automated Analysis of Diaphragmatic Motion With Dynamic Magnetic Resonance Imaging in Healthy Controls and Non-Ambulant Subjects With Duchenne Muscular Dystrophy,” Frontiers in Neurology 9 (2018): 9.

[16] B. K. Smith, S. Allen, S. Mays, A. D. Martin, and B. J. Byrne, “Dynamic Respiratory Muscle Function in Late-Onset Pompe Disease,” Scientific Reports 9, no. 1 (2019): 19006.

[17] L. Harlaar, P. Ciet, G. van Tulder, et al., “Chest MRI to Diagnose Early Diaphragmatic Weakness in Pompe Disease,” Orphanet Journal of Rare Diseases 16, no. 1 (2021): 21.

[18] S. Zhang, K. T. Block, and J. Frahm, “Magnetic Resonance Imaging in Real Time: Advances Using Radial FLASH,” Journal of Magnetic Resonance Imaging 31, no. 1 (2010): 101–109.

[19] M. Uecker, T. Hohage, K. T. Block, and J. Frahm, “Image Reconstruction by Regularized Nonlinear Inversion–Joint Estimation of Coil Sensitivities and Image Content,” Magnetic Resonance in Medicine 60, no. 3 (2008): 674–682.

[20] M. Uecker, S. Zhang, D. Voit, A. Karaus, K.-D. Merboldt, and J. Frahm, “Real-Time MRI at a Resolution of 20 Ms,” Nmr in Biomedicine 23, no. 8 (2010): 986–994.

[21] S. Zhang, M. Uecker, and J. Frahm, “T1 Mapping in Real Time: Single Inversion-Recovery Radial FLASH With Nonlinear Inverse Reconstruction [Conference Presentation Abstract]: Annual Meeting ISMRM, Salt Lake City 2013,” Proceedings of the International Society of Magnetic Resonance in Medicine 21 (2013): 3700.

[22] H. Bulluck, V. Maestrini, S. Rosmini, et al., “Myocardial T1 Mapping,” Circulation Journal 79, no. 3 (2015): 487–494.

[23] B. Marty, B. Coppa, and P. G. Carlier, “Monitoring Skeletal Muscle Chronic Fatty Degenerations With Fast T1-Mapping,” European Radiology 28, no. 11 (2018): 4662–4668.

[24] O. Ronneberger, P. Fischer, and T. Brox. U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, ed. N. Navab, J. Hornegger, W. M. Wells, and A. F. Frangi (Springer International Publishing, 2015): 234–241.

[25] N. Siddique, S. Paheding, C. P. Elkin, and V. Devabhaktuni, “U-Net and Its Variants for Medical Image Segmentation: A Review of Theory and Applications,” IEEE Access 9 (2021): 82031–82057.

[26] U. Mellies and F. Lofaso, “Pompe Disease: A Neuromuscular Disease With respiratory Muscle Involvement,” Respiratory Medicine 103, no. 4 (2009): 477–484.

[27] J. Frahm, D. Voit, and M. Uecker, “Real-Time Magnetic Resonance Imaging: Radial Gradient-Echo Sequences With Nonlinear Inverse Reconstruction,” Investigative Radiology 54, no. 12 (2019): 757–766.

[28] F. A. Laghi, M. Saad, and H. Shaikh, “Ultrasound and Non-Ultrasound Imaging Techniques in the Assessment of Diaphragmatic Dysfunction,” BMC Pulmonary Medicine 21, no. 1 (2021): 85.

[29] T. S. FitzMaurice, C. McCann, D. S. Nazareth, and M. J. Walshaw, “Characterisation of Hemidiaphragm Dysfunction Using Dynamic Chest Radiography: A Pilot Study,” ERJ Open Research 8, no. 1 (2022): 00343–2021.

[30] M. Gaeta, O. Musumeci, S. Mondello, et al., “Clinical and Pathophysiological Clues of respiratory Dysfunction in Late-Onset Pompe Disease: New Insights From a Comparative Study by MRI and Respiratory Function Assessment,” Neuromuscular Disorders 25, no. 11 (2015): 852–858.

[31] T. Suwatanapongched, D. S. Gierada, R. M. Slone, T. K. Pilgram, and P. G. Tuteur, “Variation in Diaphragm Position and Shape in Adults With Normal Pulmonary Functiona,” Chest 123, no. 6 (2003): 2019–2027.

[32] D. Reyes-Leiva, J. Alonso-Pérez, M. Mayos, et al., “Correlation Between Respiratory Accessory Muscles and Diaphragm Pillars MRI and Pulmonary Function Test in Late-Onset Pompe Disease Patients,” Frontiers in Neurology 12 (2021): 621257.

[33] P. G. Carlier and H. Reyngoudt, “The Expanding Role of MRI in Neuromuscular Disorders,” Nature Reviews Neurology 16, no. 6 (2020): 301–302.

[34] A. Paszke, S. Gross, F. Massa, et al., “PyTorch: An Imperative Style, High-Performance Deep Learning Library,” Advances in Neural Information Processing Systems 32 (2019), https://doi.org/10.48550/arXiv.1912.01703.

[35] I. Loshchilov and F. Hutter. Decoupled Weight Decay Regularization; 2017.

[36] G. Van Rossum and F. Drake, Python 3 Reference Manual (CreateSpace, 2009).

[37] C. R. Harris, K. J. Millman, S. J. van der Walt, et al., “Array Programming With NumPy,” Nature 585, no. 7825 (2020): 357–362.

[38] P. Virtanen, R. Gommers, T. E. Oliphant, et al., “SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python,” Nature Methods 17, no. 3 (2020): 261–272.

[39] G. Bradski, “The openCV Library,” Doctor Dobb´s Journal of Software Tools 120 (2000): 122–125.

[40] J. D. Hunter, “Matplotlib: A 2D Graphics Environment,” Computing in Science & Engineering 9, no. 3 (2007): 90–95.

[41] A. T. S./.E. R. Society, “ATS/ERS Statement on respiratory Muscle Testing,” American Journal of Respiratory and Critical Care Medicine 166, no. 4 (2002): 518–624.

[42] A. Boussuges, Y. Gole, and P. Blanc, “Diaphragmatic Motion Studied by M-Mode Ultrasonography: Methods, Reproducibility, and Normal Values,” Chest 135, no. 2 (2009): 391–400.

[43] A. Testa, G. Soldati, R. Giannuzzi, S. Berardi, G. Portale, and N. Gentiloni Silveri, “Ultrasound M-Mode Assessment of Diaphragmatic Kinetics by Anterior Transverse Scanning in Healthy Subjects,” Ultrasound in Medicine & Biology 37, no. 1 (2011): 44–52.

[44] A. K. Simonds, “Chronic Hypoventilation and Its Management,” European Respiratory Review 22, no. 129 (2013): 325–332.

RIGHTS & PERMISSIONS

2026 The Author(s). MedComm published by Sichuan International Medical Exchange & Promotion Association (SCIMEA) and John Wiley & Sons Australia, Ltd.