Résumé de recherche
Objective
Recent studies in respiratory system impedance (Zrs) with single-frequency oscillometry have demonstrated the utility of novel intra-breath measures of Zrs in the detection of pathological alterations in respiratory mechanics. In the present work, we addressed the feasibility of extracting intra-breath information from Zrs data sets obtained with conventional oscillometry.
Approach
Multi-frequency recordings obtained in a pulmonology practice were re-analysed to track the 11 Hz component of Zrs during normal breathing and compare the intra-breath measures to that obtained with a single 10 Hz signal in the same subjects. A non-linear model was employed to simulate changes in Zrs in the breathing cycle. The values of resistance (R) and reactance (X) at end expiration and end inspiration and their corresponding differences (R and X) were compared.
Main results
All intra-breath measures exhibited similar mean values at 10 and 11 Hz in each subject; however, the variabilities were higher at 11 Hz, especially for R and X. The poorer quality of the 11 Hz data was primarily caused by the overlapping of modulation side-lobes of adjacent oscillation frequencies. This cross-talk was enhanced by double breathing frequency components due to flow nonlinearities.
Significance
Retrospective intra-breath assessment of large or special data bases of conventional oscillometry can be performed to better characterise respiratory mechanics in different populations and disease groups. The results also have implications in the optimum design of multiple-frequency oscillometry (avoidance of densely spaced frequencies) and the use of filtering procedures that preserve the intra-breath modulation information.
