Towards lung volume measurement by a rebreathing technique
The work contained in this thesis was concerned with rebreathing methods of measuring lung volume. In particular, one novel rebreathing technique which uses oxygen as the indicator gas was assessed. This technique appeared methodologically simple and readily applicable in a clinical environment. In essence, it relied on a graphical extrapolation of the time related changes in oxygen concentration to allow for oxygen uptake. This technique has been tested using a mass spectrometer which enabled nitrogen and argon as well as oxygen to be simultaneously used as indicator gases. Although the lung volumes as measured by the different indicator gases should have been the same, these were found to be different. These discrepancies were related to the concentration of the indicator gas which existed in the bag and lung prior to rebreathing. A hypothesis explaining these inconsistancies was formulated. This was based on an initial but non-sustained output of carbon dioxide into the bag-plus-lung system. A numerical model of idealised rebreathing showed that the hypothesis was sufficient to explain the discrepancies observed. A correction procedure was devised which performed successfully in the model. This correction was incorporated into an on-line computing procedure for calculating real lung volume. When tested in normal subjects this gave consistent results for lung volume, irrespective of indicator gas employed. The corrected lung volumes were unaffected by the initial gas compositions in the bag and lung, and were also independent of non-sustained gas exchange, whether this was due to carbon dioxide and/or nitrous oxide. This technique could, therefore, be use under anaesthetic conditions, since the uptake or output of nitrous oxide no longer upsets the calculation of lung volume. The use of more than one indicator gas, within the same manoeuvre, was shown to provide a valuable indication of the presence of errors in the system. When this approach was applied to more conventional rebreathing techniques of lung volume measurement, it also highlighted the presence of inaccuracies.