A digital system for generating dynamic sinusoidal gas concentration signals

A fast, digitally controlled flow proportional gas injection system for studies in lung function

The aim of this paper is to describe a device for flow proportional injection of tracer gas in the lungs of mechanically ventilated patients. This device may then be used for the study of the multiple breath indicator gas washout technique to determine the end-expiratory lung volume. Such a tracer gas injection device may also be used in the study of other techniques that rely on uptake and elimination of tracer gas by the lungs. In this paper, an injector is described which enables injection of indicator gas at a predetermined concentration in a breathing circuit independent of the type of breathing. The presented setup uses a control computer to produce steering signals to a multivalve array in proportion to the input breathing signals. The multivalve array consists of ten circular valves, each with a different diameter, which can be opened or closed individually according to the input signal of the array. By opening of a certain combination of valves an amount of sulphur hexafluoride gas proportional to the inspiratory breathing signal is released. The rate of transmission between the components of the injection system was 80 Hz. The injector has a full flow range between 0-10 L/min. The delay time between the breathing signal and the flow response was 70 ms. The aimed washin gas concentration of 1% SF6 was achieved after 0.5 s. The study describes the results of tests to determine valve-flow ratios, step response and dynamic response of the injector. The flow output response of the injector system was shown to increase in input frequencies above 3 Hz. The valve flow ratios showed the largest relative deviation in the two smallest valves of the 10 valve array, respectively 0.005 L/min (25%) and 0.002 L/min (20%). We conclude that the injector can achieve a stable concentration of indicator gas in a breathing system with an accuracy of 0.005 L/min to execute the multiple breath indicator washout test in human subjects. The results of the study indicate that the injector may be of use in other application fields in respiratory physiology in which breathing circuit injection of indicator gas is required.

A real-time algorithm to improve the response time of a clinical multigas analyser

OBJECTIVE

An algorithm to improve the response time of a clinical respiratory multigas analyser is presented.

METHODS

The algorithm involves the application of a second order differential equation to the analyser gas output signals in real-time. The adjusted analyser output signals are compared with those of a quadrupole respiratory mass spectrometer sampling and analysing simultaneously.

RESULTS

Our results show a close correlation between the adjusted clinical gas analyser and the mass spectrometer signals. Lung volumes derived from a non-invasive sinusoidal inert gas forcing technique, in a model test lung, using the adjusted clinical gas analyser and the mass spectrometer signals demonstrated comparable results.

CONCLUSIONS

The algorithm provides an improvement on the relatively slow response times of the clinical gas analyser for breath-by-breath time-dependent applications. The same algorithm can also be applied to other instruments which have slow response times.

Pulmonary blood flow measured by inspiratory inert gas concentration forcing oscillations

The aim of this study was to discover if the forced inspired inert gas sinewave technique could be used to measure pulmonary blood flow, using nitrous oxide as the indicator gas, following inotropic stimulation of the heart by dobutamine, in the presence of a constant alveolar ventilation. Cardiac output (range 1-4.5 L min(-1)) was measured in six dogs by thermodilution and by calculation from the sinusoidal expired partial pressures of argon and nitrous oxide using: (i) analytical equations and a conventional continuous ventilation three-compartment lung model, which did not include recirculation; and (ii) a digital simulation tidal ventilation lung model (Gavaghan and Hahn, 1996. Respir. Physiol. 106, 209-221) which was adapted to include nitrous oxide mixed-venous recirculation from a combined single viscera compartment. The continuous ventilation model calculations always underestimated thermodilution cardiac output, with the bias error increasing to almost -1 L min(-1) at the longest forcing periods, 4-5 min. In contrast, the tidal ventilation model calculations were in close agreement to thermodilution cardiac output, with biases of -0.04 and -0.26 L min(-1) at forcing periods of 2 and 3 min, respectively.

Alveolar and dead space volume measured by oscillations of inspired oxygen in awake adults

Forced sinusoidal oscillations in the inspired concentration of a low-solubility inert gas can be used to measure airways dead space and alveolar volume. When inspired oxygen is oscillated about its mean value in the same way, the ratio between the amplitudes of the resulting end-expired and inspired oxygen oscillations is the same as that of an inert gas such as argon. Thus, oxygen forcing oscillations can be used to measure lung volume. In nine healthy spontaneously breathing adults, the FIO2 (mean FIO2 = 0.26, mean minute volume = 8.5 L/min) was forced to sinusoidally oscillate with an amplitude of +/- 0.04. The mean airways dead space measured using FIO2 oscillations with a forcing period of 3 min was 0.17 +/- 0.04 L, and the airways dead space measured by a single-breath C02 technique was no different at 0.19 +/- 0.03 L. An oxygen oscillation of the same period measured the mean end-expired alveolar volume at 3.1 +/- 0.7 L. Adding together the airways dead space and end-expired alveolar volume, obtained by the oxygen oscillation technique, provided a measure of FRC that at 3.3 +/- 0.7 L matched the FRC of 3.3 +/- 0.8 L measured by whole-body plethysmography. A third measure of FRC using a multiple-breath nitrogen washout technique gave a smaller volume of 3.00 +/- 0.85 L. The advantage of using FIO2 oscillations is that accurate FRC measurements can be made continuously, without interfering with the subject's natural breathing rhythm.

Measurement of dead-space in a model lung using an oscillating inspired argon signal

In a model lung, airways dead-space can be accurately measured using a forced inspired oscillating argon signal, which varies sinusoidally about a mean concentration of 6% v/v with an amplitude of +/- 4% v/v. With sinusoid forcing periods longer than 120 seconds, and at a breathing rate of 13.4 breaths minute-1, the mean airways dead-space can be measured with a standard error of less than 5%. Sinusoid forcing periods shorter than 120 s provided inaccurate estimates of dead-space and so should not be used with this technique.