Evaluation of various models for respiratory inductance plethysmography calibration

We evaluated one nonlinear and two linear models of the ventilatory system while calibrating the respiratory inductance plethysmograph (RIP) against a pneumotachometer. A calibration method involving voluntary varying rib cage and abdominal contributions to tidal volume in a single body position was utilized. The influence on accuracy of the choice of respiratory phase during calibration was assessed. Both tidal and intratidal volumes were evaluated. Ten adults with no history of respiratory disorders went through RIP calibration and validation in the sitting and supine positions. A linear calibration model, relating lung volume changes from the start of inspiration or expiration to rib cage and abdominal excursions from initiation of respiratory motion, had the best accuracy. The choice of respiratory phase for calibration did not affect accuracy. RIP generally underestimated lung volume at the start of inspiration and overestimated lung volume at the end of inspiration. RIP was more accurate in the supine than the sitting position, probably because of limited spine flexion in the supine position.