Abstract
PURPOSE: To present a method for quantifying dissolved (129)Xe spectroscopy using singular value decomposition (SVD) and a dynamic red blood cell (RBC)/membrane ratio as a biomarker of disease. METHODS: A spectroscopic sequence was performed in 45 subjects (27 healthy, 12 dyspnea of unknown origin [DUO], and 6 pulmonary hypertension [PH]) consisting of 499 pulse/acquire experiments. SVD was used to construct a low-noise approximation of FID data, and time-domain curve-fitting was performed on all free induction decays allowing calculation of RBC (218 ppm) and membrane (197 ppm) signal amplitudes and the RBC/membrane ratio. RBC/membrane oscillation amplitudes were assessed using independent t-tests. An Analysis of Covariance (ANCOVA) test was performed to control for age and sex, followed by post hoc Tukey tests for pairwise comparisons. RESULTS: Independent t-tests demonstrated statistically significant differences in RBC oscillation amplitudes and RBC/membrane oscillation amplitudes among healthy subjects and DUO patients (p-value = 0.003 and p-value = 0.0008, respectively). An ANCOVA test was performed to control for age and sex and resulted in statistically significant differences among diseases (p = 3.12 × 10(-6)). A post hoc pairwise Tukey test demonstrated statistical significance among healthy subjects and DUO patients (p = 3.12 × 10(-5)) and among healthy and PH patients (p = 2.56 × 10(-5)). CONCLUSION: Dynamic RBC-to-membrane ratio measurements may yield useful physiological information related to overall lung health.