A software program for quantitative analysis of alveolar oxygen partial pressure (pAO2) with oxygen-sensitive 3He-MRI

Purpose: To develop a software tool for quantitative analysis of alveolar oxygen partial pressure (p(A)O(2)) as well as its time course during apnea. Material and Methods: T-1-relaxation times of hyperpolarized He-3 are reduced by paramagnetic oxygen rendering He-3-MRI sensitive to oxygen and thus allowing the assessment of the local oxygen partial pressure in the pulmonary airspaces. Oxygen-related relaxation and loss of polarization by RF-excitation can be discriminated by acquiring two image series with varying interscan delay and/or flip angles. Software was developed to calculate the p(A)O(2) and the decay rate in user-defined regions of interest (ROI) automatically. Moreover, parameter maps can be calculated. In addition to the analysis of 2-dimensional data sets, the software allows the evaluation of 3-dimensional measurements for the first time. Artifacts due to lung motion were reduced by implementing a motion correction algorithm. Results: The software was successfully applied to data sets from healthy volunteers and from patients with various lung diseases. The parameter maps demonstrated a more homogeneous distribution of p(A)O(2) for the volunteers than for the patients. A regional increase in p(A)O(2) was found in a few patients. Conclusion: The described software allows the absolute quantification of p(A)O(2) as well as its variation over time. In the future, therefore, the software may gain importance for detecting mismatches between ventilation and perfusion, e. g., in patients with pulmonary embolism or chronic obstructive lung diseases.