Search

The merits of using physiologic sampling pumps (PSPs) instead of using constant-flow sampling pumps, i.e., "traditional sampling pumps" (TSPs), are discussed. A novel PSP that overcomes shortcomings of previous PSP designs is presented. Calibrated valves are used to obviate pump inertia that has limited the system response and accuracy of prior work. Technologies that provide minute ventilation (V[combining dot above](E)) of subjects in real time may therefore be used to the limit of their own accuracies to sample inhalation exposures. Analysis of the design and data from a prototype are presented to show how air sampling can be modulated to follow breathing.

A study was conducted to evaluate a portable respiratory inductive plethysmograph (RIP) as a means to estimate minute ventilation (V(E)) for use in controlling the flow rate of a physiologic sampling pump (PSP). Specific aims were to: (1) evaluate the ability of the portable RIP system to measure V(E) using a direct (individual) fixed-volume calibration method (Direct RIP model), (2) develop and evaluate the performance of indirect (group) regression models for V(E) prediction using output data from the portable RIP and subject demographic characteristics (Indirect RIP model), and (3) compare V(E) estimates from indirect and direct portable RIP calibration with indirect estimation models published previously. Nine subjects (19-44 years) were divided into calibration (n = 6) and test (n = 3) datasets and performed step-tests on three different days while wearing the portable RIP and breathing through a pneumotachometer (reference). Minute ventilation and portable RIP output including heart rate, breathing rate, and a motion index were recorded simultaneously during the 80 minute sessions. Calibration data were used to develop a regression model for V(E) prediction that was subsequently applied to the test dataset. Direct calibration of the portable RIP system produced highly variable estimates of V(E) (R2 = 0.62, average % error = 15 +/- 50) while Indirect RIP model results were highly correlated with the reference (R2 = 0.80-0.88) and estimates of total volume were within 10% of reference values on average. Although developed from a limited dataset, the Indirect RIP model provided an alternative approach to estimation of V(E) and total volume with accuracy comparable to previously published models.