ICA-enabled oxygen-enhanced MRI (OE-MRI) correlates with pulmonary function tests in cystic fibrosis

Motivation: There is a clinical need for non-ionising methods to assess heterogeneous lung function in cystic fibrosis (CF). Dynamic oxygen-enhanced MRI (OE-MRI) can assess regional lung function, however OE-MRI analysis is impaired by confounding signals and poor SNR. Goal(s): To evaluate the sensitivity of OE-MRI measures to the lung clearance index (LCI) in CF, with and without independent component analysis (ICA) to reduce noise. Approach: We used ICA to reduce noise in the OE-MRI measures. We evaluated the correlation between OE-MRI measures, LCI, and pulmonary function tests. Results: OE-MRI measures demonstrated significant correlation with LCI. OE-MRI measures extracted using ICA displayed clear oxygen-enhancement responses. Impact: Dynamic lung OE-MRI measures extracted using independent component analysis (ICA) exhibited significant correlation with lung clearance index (LCI2.5) in cystic fibrosis (CF) patients, suggesting a potential application of ICA-extracted OE-MRI measures to assess regional disease severity in CF.