In silico modeling of oxygen‐enhanced MRI of specific ventilation

• Published in: Physiological reports Vol. 6; no. 7; pp. e13659 - n/a
• Main Authors: Kang, Wendy, Tawhai, Merryn H., Clark, Alys R., Sá, Rui C., Geier, Eric T., Prisk, G. Kim, Burrowes, Kelly S.
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.04.2018; John Wiley and Sons Inc
• Subjects: Computational model; Gas exchange; Gas Exchange and Transport; Lung; Lungs; Magnetic resonance imaging; Original Research; oxygen‐enhanced; Perfusion; Physiology; proton MRI; Pulmonary Circulation; specific ventilation; Ventilation; oxygen-enhanced; Computational model; proton MRI; specific ventilation
• ISSN: 2051-817X; 2051-817X
• DOI: 10.14814/phy2.13659

Specific ventilation imaging (SVI) proposes that using oxygen‐enhanced 1H MRI to capture signal change as subjects alternatively breathe room air and 100% O2 provides an estimate of specific ventilation distribution in the lung. How well this technique measures SV and the effect of currently adopted approaches of the technique on resulting SV measurement is open for further exploration. We investigated (1) How well does imaging a single sagittal lung slice represent whole lung SV? (2) What is the influence of pulmonary venous blood on the measured MRI signal and resultant SVI measure? and (3) How does inclusion of misaligned images affect SVI measurement? In this study, we utilized two patient‐based in silico models of ventilation, perfusion, and gas exchange to address these questions for normal healthy lungs. Simulation results from the two healthy young subjects show that imaging a single slice is generally representative of whole lung SV distribution, with a calculated SV gradient within 90% of that calculated for whole lung distributions. Contribution of O2 from the venous circulation results in overestimation of SV at a regional level where major pulmonary veins cross the imaging plane, resulting in a 10% increase in SV gradient for the imaging slice. A worst‐case scenario simulation of image misalignment increased the SV gradient by 11.4% for the imaged slice. In this study, we have assessed some of the underlying assumptions of an oxygen‐enhanced proton MRI technique to measure specific ventilation (SV) via in silico modeling. Simulation results show that imaging a single slice is representative of whole lung SV and the contribution of oxygen in the venous circulation on SV measurements is minimal. In addition, simulated misalignment of the 220 image set showed minimal impact on SV suggesting that this imaging technique is robust in healthy individuals.