Modeling impulse oscillometry data: Using augmented-RIC and extended-RIC respiratory impedance model parameters to track pulmonary health and disease
Respiratory diseases are a great health concern. Early detection of disease could greatly improve quality of life and yet the most common pulmonary function test can be an unreliable diagnostic method, especially in the case of children. A reliable method of detecting and assessing small airway impairment could greatly help with timely diagnosis. This research approaches this problem by taking Impulse Oscillometry (IOS) data and its model-derived parameters and assessing their utility in tracking respiratory function and more specifically, small airway function. This research first describes the use of electric circuits as models to describe lung properties at distinct regions of the respiratory system. Analyzing one of the most successful of these models, Mead's respiratory impedance model, it was concluded that it was not applicable when using IOS data. It is reasoned that the very small volume displacements generated with this testing technique do not sufficiently stimulate the subjects' lung tissue and chest wall making these two parameters found in the Mead's model irrelevant. Therefore, the more parsimonious versions of the Mead's model namely the augmented-RIC (aRIC) and extended-RIC (eRIC) models are shown to be more suitable respiratory impedance models for IOS data. In order to estimate model parameters an algorithm was developed based on the nonlinear least squares minimization technique. A very large and diverse database was obtained consisting of 670 IOS test results from patients classified as normal (NL), asthmatic (ASTH) or small airway impaired (SAI). Using the developed algorithm and database, model parameters were estimated for both the aRIC and eRIC models. It was found that both models succeeded almost equally well in modeling quality assured IOS data providing good data fit for all data classifications (NL, ASTH or SAI). Statistical analysis showed that model-derived parameters are able to discriminate between health and disease while at the same time monitoring response to bronchodilator administration, with the peripheral compliance (Cp) model parameter showing an even better response than IOS-derived indices. The use of such a large database makes it possible for this study to provide a reliable range of what can be considered to be "normal" values for IOS indices, which can be used by themselves as useful indices of small airway impairment. However, it is concluded that model-derived parameters work equally well in differentiating health from disease while providing insight into the function of specific lung regions. This is especially true in the case of the aRIC model, which tracks lung function in a reliable and realistic manner.
Biomedical engineering|Electrical engineering
Ramos Rocha, Carlos D, "Modeling impulse oscillometry data: Using augmented-RIC and extended-RIC respiratory impedance model parameters to track pulmonary health and disease" (2011). ETD Collection for University of Texas, El Paso. AAI1498314.