Analysis of Forced Expiratory Flow Signals Using the New Luus-Jaakola Optimization Procedure.

Abstract

The aim of this study was to estimate forced vital capacity (FVC) by analyzing incomplete forced expiratory flow-time (FEFT) curves using the new Luus-Jaakola (NLJ) optimization procedure. Complete FEFT curves from 90 subjects with and without respiratory disease were used to validate the method, and its application performance was evaluated by FEFT curves containing incomplete curves from nine obstructive patients. First, a lumped parameter model of the FEFT curve, taking FVC as one of the model parameters, is chosen. The 0- 1 s, 0-1.5 s, . . . , 0-6 s curves chosen from the accepted FEFT curves are, respectively, taken as the local system observations. Then, the NLJ optimization procedure was used to get the estimate of FVC (FVCest) by estimating the model parameters. Finally, FVCest was compared with the measured FVC (FVCmea). The difference between FVCest and FVCmea decreased with the lengths of observed curves increasing. When the lengths of observed curves reached 3 s, the mean difference (±SD) between FVCest and FVCmea was -11 ± 58 mL. Estimated FVC derived from the incomplete curves were close to FVC from accepted maneuvers. The method can be used to estimate FVC by analyzing an incomplete FEFT curve, providing the length of the curve is more than 3 s. This method can help the subjects who cannot complete the forced expiratory maneuver to provide a reliable estimate of FVC and also can be used to analyze the unacceptable spirometry because of early termination of exhalation.