A study of IOS data using the aRIC+I(p) model of respiratory impedance

Development of better methods to assess human lung function has been continuing since the existing standard lung function test of spirometry requires subjects to inhale and exhale with maximum effort, which may be troublesome especially for the elderly and young children, leading to unreliable results. Therefore, the method of forced oscillation, and the Impulse Oscillometry System (IOS) in particular, has been developed to lessen the effort of the patients while obtaining valid measurements. The applied pressure waves and the resulting airflow responses are recorded to provide information about the respiratory system's input impedance, which can be fit by electric circuit models to possibly serve as a means to detect and diagnose respiratory diseases. Presently, research continues to find a more accurate model that also provides reasonable component values. This paper proposes the augmented RIC+I(p)(aRIC+I(p)) model and compares it to five other well-known models (the RIC, extended RIC, augmented RIC, DuBois and Mead models) in fitting the IOS data from adult COPD patients and healthy subjects. While the aRIC+I(p) model yielded slightly higher fitting error than the Mead and DuBois models, it did not produce unphysiologically large values for any of its components, unlike the Mead and DuBois models. Hence, the aRIC+(p) model appears to be the most reasonable one for use, at this point in time, in studying IOS-based computer-aided detection and diagnosis of COPD.