Pulmonary models have been used to capture and discriminate the respiratory mechanics of patients with pulmonary dysfunction [1]. These models vary in nature for highly detailed and complex models [2] to simple, lumped parameter models [3, 4]. Both the simple and complex modelling approaches have contributed to the state of the art in decision support for respiratory care. The NARX model of respiratory mechanics is a comparatively complex lumped parameter model that captures changes in elastance during pulmonary recruitment [4]. It has exhibited strong ability to interpolate and extrapolate across mechanical ventilation settings.