Chronic obstructive pulmonary disease (COPD) is heterogeneous, with structural and functional changes that vary across patients and over time. Spirometry confirms airflow limitation but reflects global function, and computed tomography (CT) characterises emphysema and airway remodelling with ionising radiation. Magnetic resonance imaging (MRI) offers radiation-free assessment of lung structure and function using contrast-based perfusion mapping, inhaled gases and free-breathing analysis. Quantitative MRI (qMRI) biomarkers from these approaches relate to clinical status and treatment response, supporting longitudinal monitoring and therapy evaluation. Progress depends on standardised methods and robust validation to define where MRI complements or substitutes existing pathways.
Evolving MRI Techniques for Structure and Function
Dynamic contrast-enhanced MRI (DCE-MRI) visualises pulmonary perfusion after intravenous contrast and provides depiction comparable with contrast-enhanced CT. Breath-hold acquisitions are common, though free-breathing protocols are feasible. Gadolinium chelates remain standard but require attention to safety with repeated exposure.
Must Read: CT Imaging Sharpens Phenotyping in COPD
Hyperpolarised-gas MRI (HP-MRI) uses inhaled ^3He or ^129Xe to map ventilation. ^129Xe also assesses gas exchange because it partitions into tissue and blood with separable spectral peaks. HP-MRI yields functional information analogous to nuclear medicine techniques yet relies on specialised equipment and supply chains that can limit access.
Fluorine-gas MRI (^19F-MRI) images ventilation with perfluorinated gases without a polariser, typically during breath-hold or free breathing. It is more accessible than HP-MRI but still needs inhaled gas and multinuclear hardware, and comparative validation remains limited.
Conventional proton MRI (¹H-MRI) has advanced with ultrashort echo time (UTE) sequences that capture rapidly decaying signal to reveal parenchyma and airways. Phase-resolved functional lung MRI (PREFUL) derives ventilation and perfusion from free-breathing ¹H-MRI without exogenous agents. These approaches avoid contrast and specialised gas hardware, improving feasibility for repeated assessments, though coil dependence, bias fields and arbitrary signal units complicate standardisation.
Biomarkers Linking Imaging to Clinical Outcomes
Structural MRI measures are developing alongside simple read-outs on ultrashort echo time (UTE) images. Visual grading of emphysema on UTE broadly matches CT and repeats well. A summary number called low signal volume (LSV) counts the share of lung with very low signal and tracks with emphysema seen on CT, separating groups by severity. Because lung volume at the time of scanning can affect LSV, teams need aligned protocols and confirmed cut-offs. Airways can be seen down to small branches on UTE with consistent visual grading. Early tools that partly automate airway tracing and wall measurements look workable, but COPD-specific airway metrics still need wider development and testing.
Measures of function are further along. Ventilation defect percent (VDP) estimates how much of the lung is poorly ventilated. It can be produced from hyperpolarised-gas MRI, fluorine-gas MRI or ¹H-based free-breathing methods after aligning with structural images. Despite differences in how images are processed, VDP from hyperpolarised-gas MRI relates to spirometry, symptoms, exercise capacity and longer term outcomes, including decline in lung function, flare-ups and death. VDP also captures improvement after long-acting bronchodilators. VDP from fluorine-gas MRI is in line with hyperpolarised results, separates COPD from health and responds to short-acting bronchodilators, with most reported links focused on spirometry. VDP from ¹H-based free-breathing methods agrees with both, is repeatable, separates disease from health and is sensitive to dual bronchodilator therapy, with growing evidence beyond basic lung function.
Perfusion defect percent (QDP) highlights areas with low blood flow. On contrast-based scans, QDP aligns with expert visual reads and relates to spirometry and CT measures tied to small airways disease and emphysema, and it improves after dual bronchodilation. The influence of scan settings and repeatability point to a need for protocol tuning. A contrast-free version of QDP from ¹H-based free-breathing methods matches contrast-based results, shows strong repeatability, rises with worsening COPD and responds to treatment. Broader checks across scanners and standard settings are still required.
Other MRI measures add depth. With hyperpolarised gases, the red blood cell to barrier ratio reflects gas transfer and may help identify larger ventilation gains after bronchodilators. Diffusion-weighted hyperpolarised scans provide an apparent diffusion coefficient that increases with alveolar damage and aligns with diffusion capacity and CT emphysema. Flow measures from contrast-based or ¹H-based free-breathing scans can be derived, and ventilation–perfusion matching can be calculated from combined maps, with signs of better matching after bronchodilation.
Standardisation, Automation and the Path to Clinical Utility
Acquisition and analysis standards lag technical progress. Automated deep-learning pipelines for parenchymal segmentation on UTE are emerging, but open-source availability is limited and performance across scanners, field strengths, sequences and breathing protocols needs rigorous evaluation. Comparable development is needed for airway and vascular segmentation, as well as automated structure–function registration with agreed ventilation and perfusion defect definitions. Synthetic image generation could expand training datasets but requires careful validation against real-world variability.
Clinical validation is equally important. Structural qMRI measures should be harmonised with established CT metrics and, where feasible, supported by histopathological correlation. Large, multicentre, longitudinal COPD cohorts designed to test repeatability, reproducibility, minimal clinically important differences and associations with patient-centred outcomes would enable biomarker qualification and regulatory acceptance. Low-field MRI systems offer potential advantages for lung imaging, including improved field homogeneity, reduced susceptibility artefacts, smaller footprints and lower costs, with oxygen-enhanced protocols as a possible complement. Comparative studies versus 1.5 T, 3 T and CT are needed to define appropriate use.
MRI enables radiation-free quantification of COPD structure and function across UTE-based anatomy and contrast-free or contrast-based functional mapping. Ventilation and perfusion biomarkers show disease discrimination, associations with lung function and responsiveness to bronchodilators, while structural measures converge with CT surrogates. Adoption depends on standardised acquisition and analysis, cross-platform validation and longitudinal studies linking imaging changes to outcomes, positioning qMRI to support targeted therapy selection, safer monitoring and more precise evaluation of treatment response in COPD.
Source: British Journal of Radiology
Image Credit: iStock
