FI102569B - Measurement of gas components with high precision - Google Patents

Description

102569

The invention relates to a method for the non-dispersive analysis of gas mixtures by radiation absorption, wherein the analysis is directed to at least one predetermined gas component and the gas mixture may contain interfering gaseous components. to reduce these said effects for the measurement result, the method comprising the steps of: determining, prior to analysis, the magnitude and direction of the effect of each significant interference component affecting the radiation absorption of the gas component being measured at at least one concentration of both the gas component and the interference component. In particular, the invention is directed to the use of a method for measuring the gas components of a patient's respiratory gas by the infrared method by compensating for the cross-effects due to collisions of different gas components.

When measuring the gas concentration by infrared technology, a non-dispersive method is most often used, i.e. the absorption signal is measured in a relatively narrow wavelength range through an optical bandpass filter. The width of the transmission band of the optical filter is typically of the same order of magnitude as the width of the spectral line utilized by the gas component to be analyzed. The measured radiation signal is then an integrated value over the transmissions prevailing at different wavelengths of the transmission band. Such devices are described in U.S. Pat. No. 3,745,349 and HEWLETT-PACKARD 25 JOURNAL, September 1981, pp. 3-5: R.J. Solomon - "A Reliable, Accurate CO2 Analyzer for Medical Use". The former publication uses two infrared radiation sources which emit through the sample gas, but the radiation from the second radiation source is made narrower than the wavelength range used in the measurement, for example by means of an optical gas filter. The purpose is to provide a reference source whose radiant intensity 30 would not be affected by the gas mixture being examined in the detector. The aim is to compensate for the effect of dirt and other substances accumulating in the measuring chamber, which have an absorption in the same area as the gas under test, and to avoid matching problems between different detectors. Infrared sources are used modulated, so that one detector alternates between the actual signal and the reference-35 signal, which have thus traveled the same path and see the contaminants in the same way. In this publication, these problems are solved with a structure without moving parts. The latter publication has sought to solve the same problems, but by using a rotating disk containing different filters to obtain a reference value.

\ 2 102569

The absorption spectrum of a gas in molecular form normally consists of absorption spectral lines caused by molecular vibrations and a fine structure due to rotational displacement within them, i.e. absorption lines. Thus, measured with sufficient resolution, the gas absorption spectrum consists of a large number of very narrow ab-5 sorption lines. For example, carbon dioxide has a vibration absorption spectral band with a center wavelength of 4260 nm. A more detailed analysis shows that the region consists of more than 80 narrow absorption lines caused by rotation. The half-width-width of these lines, which means the line plate at the point where the absorption is half maximum, and the absorption intensity depend on many factors such as temperature, self-absorption caused by the long measurement paths, and collisions of other molecules in the gas mixture. Collision flattening thus means an increase in the width of the absorption lines, in particular the half-width, in connection with which the intensity of the absorption peak of the line simultaneously decreases. The first two can generally be easily taken into account in the compensation of the measurement signal by measuring the temperature and the linearization effects due to the measurement geometry on the gas in question.

Instead, the change due to collisions with other gas components, which can sometimes be significant, must be taken into account in particular to minimize concentration errors. In APPLIED OPTICS, Vol. 25, no. 14/1986, pp. 2434-2439:

Cousin, Le Doucen, Houdeau, Boulet, Henry - "Air broadened linewidths, intensities-20 ties, and spectral line shapes for CO2 at 4.3 pm in the region of the AMTS instrument" describes changes in the half-width of carbon dioxide in a nitrogen and oxygen mixture. The half-width of the carbon dioxide line of the gas mixture at normal pressure (serial number 67) is 0.055 cm -1 (0.10 nm) in the oxygen mixture and 0.060 cm -1 (0.11 nm) in the nitrogen mixture for a concentration of 5% CO 2. The proportion of carbon dioxide itself caused by this widening in these figures is only about 0.003 cm -1.

Polar gases such as nitrous oxide and anesthetic gases have a much greater effect on half-width. For this reason, for example, the carbon dioxide measurement result of a patient's respiratory gas is corrected by also measuring the nitrous oxide concentration as in U.S. Pat. No. 4,423,739 and using this nitrous oxide measurement result computationally to correct the carbon dioxide concentration measurement result. It is known that the effect of oxygen on the measurement result can be compensated in a corresponding manner, although the error caused by oxygen is smaller than the error caused by nitrous oxide. Oxygen does not have infrared ab sorption, so the concentration information required for compensation must be measured by some other method, such as paramagnetically. Nitrogen gas also has no infrared absorption and, in addition, is very difficult to measure, the analyzer is usually calibrated in the presence of nitrogen, assuming that its need for compensation is eliminated and other gases are compensated for the carbon dioxide measurement thus obtained. This results, for example, in the effect of oxygen on carbon dioxide reducing the concentration value, because the collision propagation caused by oxygen is smaller than that caused by nitrogen. All other gases normally present in the patient's breath during anesthesia contribute to the initial measurement of carbon dioxide. Rarely used noble gases, such as helium and xenon, have a small collision propagation effect, which means that they have a concentration-reducing effect compared to nitrogen. Similar compensatory needs for the collision propagation of the gas components present in the breathing gas mixture also exist for other gases, with the largest interacting gases being carbon dioxide, the anesthetic gases previously listed.

In addition to the error caused by the collision propagation described above, the measurement signal may have overlapping infrared absorption of the gas components. It may cause an error in the measured gas reading even if there is no gas to be measured. This invention does not relate to this well-known phenomenon.

The gas concentration is proportional to the number of gas molecules in the measurement volume and pressure. The number of molecules involved in infrared absorption in the collision process remains more or less the same unless the conditions otherwise change. The energy distribution of 20 alone has changed slightly, causing the absorption lines to widen. The absorbance value integrated over the absorption line thus remains virtually unchanged. However, infrared technology is not able to directly measure absorbance but measures transmission. Especially when measuring non-dispersively using a wavelength band wider than the width of a single ab-25 sorption line, the transmission measurement signal which will be integral over the above-mentioned wide wavelength band of the optical bandpass filter is erroneous, giving too high a concentration.

If the transmission band of the optical bandpass filter extends over several absorption lines, as is usually the case when measuring the gas components of the respiratory gases, then there is a need for correction because the concentration reading increases as the collision width increases. The concentration error is also affected by the length of the measuring channel, i.e. the dimension of the measuring chamber in the direction in which the radiation passes through the chamber. In the short channel, the need for compensation is lower, while in the longer channel, where the absorption of the gas to be measured has reduced the transmission substantially more, the need for compensation may be appreciable. In a carbon dioxide measurement of a patient's breathing air, where, for example, a substantial portion of the gas mixture may be nitrous oxide, i.e., nitrous oxide, the calculated concentration value may be up to 15% too high due to the transmission integral error described above. The error is further exacerbated by the anesthetic gases used during anesthesia, especially desflurane and sevoflurane due to their high concentration of use or high absorption capacity, but also other anesthetic gases such as halothane, enflurane and isoflurane and xenon et al.

It is an object of the present invention to provide a method for compensating as completely as possible for a collision propagation error in a normal non-dispersive gas measuring system in which the transmission bandwidth of the optical bandpass filter is greater than the width of a single absorption line. Another object of the invention is such a method by which the error caused by a collision propagation caused by gas components other than nitrous oxide could be compensated as simply and reliably as possible without having to calibrate the device separately for each disturbance component causing the collision propagation. A third object of the Kek-15 invention is such a method which is particularly suitable for the analysis of a patient's respiratory gases.

The disadvantages described above can be eliminated and the objects defined above can be achieved by the method according to the invention, which is characterized by what is defined in the characterizing part of claim 1 and by the use according to the invention, characterized by what is defined in the characterizing part of claim 5.

The main advantage of the invention is that the method according to it can yk-j. 25 simple correction terms to compensate for a measurement error caused by multiple collision propagation caused by, for example, a gas component in a patient's breathing air. The invention also has the advantage that the method according to it can be used to compensate not only the measurement result of carbon dioxide but also the measurement result of nitrous oxide or, if the measurement is applied to other gas components, to compensate the measurement results of these other gas components.

In the method of the invention, at least all essential gas components with infrared absorption are measured from the patient's breathing gas. The initial measured concentration value of a gas component 35 measured by infrared technology, such as carbon dioxide and nitrous oxide, is generally compensated for by the collision propagation caused by at least two gas components having infrared absorption and present in a gas mixture such as breathing gas. In order to obtain an accurate measurement result, the measuring sensor is individually calibrated using a final optical bandpass filter, due to the fact that the transmission bands of the different individuals of the bandpass filters are always located at slightly different wavelengths. However, according to the invention, it is sufficient to use one gas component, in particular anesthesia-5 gas, to calibrate the device and to use the result obtained to determine the compensation parameters of several other, in particular anesthetic, gases indirectly. This is done by utilizing the dependence coefficients between the above-mentioned one gas component and the other gas components, which have been determined in another context in accordance with the invention. These dependence factors are used to correct the collision propagations caused by a gas component other than the above-mentioned one, while the collision propagation caused by this one gas component is compensated in a conventional manner by means of a compensation function. Compensation by means of dependence factors and a compensation function is most preferably performed in the order of first compensating the proportion of gas components whose concentration measurement is not affected or is affected as little as possible by the actual gas component or components to be analyzed. The gas to be measured is a component in the gas mixture to be analyzed for which strict requirements have been set for the measurement accuracy and reliability, in which case the effect of possible sources of error must be eliminated.

20

In the following, the invention will be described in detail with reference to the accompanying figures.

Figure 1 schematically shows how the different gas components of the breathing gas interact through the collision-expansion.

25

Figures 2A and 2B show some examples of errors caused by desflurane (Des) and Sevo-fluran (Sev), respectively, in the carbon dioxide concentration reading as the center wavelength of the passband of the optical bandpass filter changes. The vertical axis has a percentage error and the horizontal axis has a carbon dioxide content.

30

Figure 3 shows the interdependence of the errors caused by desflurane and sevoflurane in the carbon dioxide reading in Figures 2A and 2B.

Figures 4A and 4B show some examples of errors caused by desflurane (Des) and Sevo-35 flurane (Sev), respectively, in the nitrous oxide concentration reading as the center wavelength of the passband of the optical bandpass filter changes. The vertical axis has a percentage error and the horizontal axis has a nitrous oxide content.

6 102569

Figure 5 shows the interdependence of the defects caused by desflurane and sevoflurane in Figures 4A and 4B.

As shown in Figure 1, substantially all of the components 5 present in the breathing gas cause a collision propagation when measured non-dispersively through an optical bandpass filter using infrared technology, the passband of which is substantially wider than the width of the individual rotational absorption lines. The arrows indicate which gases cause the collision propagation and which gas it targets. Carbon dioxide requires the most compensation, but the nitrous oxide reading also depends on many gases. 10 The readings for anesthetic gases (desflurane, sevoflurane, halothane, enflurane and iso-fluran) are also to some extent dependent on other gas components, although the need for compensation is so small that it is not necessary in practice. Measuring the compensation parameters individually for all gas measuring sensors completed in production would be time consuming and expensive, especially if it has to be done for all anesthetic gases to achieve accurate compensation.

The bandpass filters used generally have a bandwidth of 1 to 3% of their center wavelength. For example, if the center wavelength is 4260 nm suitable for carbon dioxide, then the bandwidth is generally about 100 nm. For manufacturing reasons, there is a slight scatter in the wavelength of mid-20. For this reason, the passband of the filter does not always coincide with the gas absorption line. In most cases, the situation is even such that the required collision propagation compensation parameters depend on the center wavelength of the filter. It is therefore necessary to perform an individual calibration to find the parameters. In Figures 2A and 2B, the wavelength dependence is clearly shown. In Fig. 2A, the error caused by desflurane (10% concentration) in the carbon dioxide reading is shown for three different bandpass filters suitable for measuring carbon dioxide. The errors shown in D% CO2 also depend on the measurement geometry. The results in Figures 2A and 2B have been measured using a 30 mm long sample chamber and would be smaller with a shorter chamber. Figure 2 B shows that sevoflurane (7.5% concentration) 30 also has a similar effect on the carbon dioxide reading. The other anesthetic gases used are halothane, enflurane and isoflurane. However, they have a smaller error reading because e.g. application rates are lower.

Desflurane causes the same type of compensation for carbon dioxide as se-35 voflurane. Figure 3 shows the errors shown in Figure 2B as a function of the errors shown in Figure 2A. As shown in the figure, the errors are practically linearly interdependent and wavelength independent. In other words, if a suitable compensation parameter for desflurane is found, then it is also validly suitable for 7 102569 converted to sevoflurane. The linear dependence is good enough in practical breathing gas measurements, but other more complex dependence functions can also be considered.

5 Pleasure gas also requires collision propagation compensation not only for carbon dioxide, especially for high-concentration or high-absorption anesthetic gases such as desflurane and sevoflurane. Figure 4 shows similar errors as in Figure 2 but for nitrous oxide. Here, too, there is a wavelength dependence, albeit less. In the example, the mean wavelength of band 10 at 3900 nm with a certain scattering is used to measure nitrous oxide. The behavior for desflurane in Figure 4 A is similar to that for sevoflurane in Figure 4 B. A similar linear dependence as in Figure 3 can now also be found, as shown in Figure 5. Thus, it is sufficient to measure, for example, the compensatory function of desflurane. Compensations for other anesthetic gases can be calculated from desflurane values.

15

Figures 2-5 showed only examples of how the collision propagation compensation parameters required to accurately measure respiratory gas can be determined using a single gas. The collision propagation caused by this one interfering gas, e.g. desflurane, is then compensated by using a compensation function f {w, k * ci, cz} of the type described in Figures 2A and 4A, the value of which thus depends on the center wavelength w of the optical bandpass filter, desflurane and the concentration cz of the gas component to be measured. In the case of Figures 2A and 2B, the object of measurement is carbon dioxide, where cz denotes the CO2 content and cj denotes desflurane and sevo-: fluran, respectively. Figures 4A and 4B show similar compensation functions f {w, k * ci, cz} with the measurement target being nitric oxide, where cz denotes the N 2 O content and ci denotes desflurane and sevoflurane, respectively. If this one interfering gas is a gas other than desflurane, this means, of course, the concentration of the respective interfering gas, as can be seen from the above. Carbon dioxide 30 and / or nitrous oxide are typical target gases for medical use. According to the invention, at least one of the two is the gas to be measured, generally carbon dioxide, the initial measured concentration of which is corrected by means of one of the interference gas compensation functions discussed here. When a single measuring device is calibrated using the exact optical bandpass filter unit that remains in the device when it is put into use, i.e. when performing analyzes, the compensation function in the form f {k * cj, cz} is simplified, according to which the measured value of the device is corrected. In other applications, of course, there may be other gas 102569 su or other gases as the measuring object, in which case the previously determined compensation function f {w, k * cj, cz} or f {k * cj, cz}, respectively, is used.

When the concentration of at least one measured gas component is corrected as described above by one of the gas components causing the collision propagation, i.e. the compensating function of the first disturbance component, the error of the collision propagation taking into account the concentration of each of the other 10 interference components in the gas mixture. In most cases, the dependence of the dependence coefficient on the concentration of the disturbance component in question is so small that in practice the dependence coefficient can be approximated by a constant k, where the function k {ci} is of the form k * cj. For example, the coefficient of interdependence of sevoflurane and desflurane ka = 2 with sufficient accuracy for most applications, if their mixture ratio is kept approximately constant, i.e. 10%: 7.5% = 4: 3, as can be seen in Figures 3 and 5. In this case, the dependence factor ka is also independent of the measured gas, which means that the carbon dioxide content and nitric oxide content can be corrected for desflurane by the same factor ka = 2 if the first interference component was sevoflurane and the second interference component was greater than content. If the mixture ratio of sevoflurane and desflurane in the gas mixture is different, e.g. 1: 1, a factor ka of about 1.5 is obtained, which value is obtained by converting the original factor 4 = 3 by the quotient of the original mixture ratio 4 = 3 and the new mixture ratio = 2. For different gases : 25 therefore, it is usually not necessary to determine the dependence coefficients separately, but they can be obtained computationally. In exactly the same way as defined above, the interdependence coefficient of sevoflurane and desflurane according to the invention can also be used to determine the interdependence coefficients of other interfering gases, i.e. the interfering components and the first interfering component. In the case of medicine shown in the figures, the dependence coefficients would be e.g. kj, (sevoflurane - isoflurane), kc (sevoflurane - halothane), kj (sevoflurane - nitrous oxide), ke (sevoflurane - carbon dioxide), kf (sevoflurane - oxygen), etc., depending on which gas components are used. In general, it is reasonable to assume that the measured concentrations of said anesthetic gases are correct, i.e. that nitrous oxide and carbon dioxide do not cause collision propagation (in the case where their concentration is measured by infrared absorption), although this is not the case. Thus, the concentration measurement result of a non-first disturbance component is corrected by a quantity obtained by measuring the disturbance component in question or otherwise known and the dependence factor, e.g., k * c, where c is the concentration of the respective disturbance component (other than the first disturbance component). This method according to the invention e.g. minimizes the need for gases in the sensor parameter determination phase and thus saves both money and effort. It is clear that 5 gases other than those mentioned above could also be used, applying the same principle.

Of course, in order to determine the dependence factors described above, the magnitude of the effect of each interference component and the direction of the effect on each measured gas concentration must first be measured, but it is sufficient to do so for one concentration of this interference component and the target component. This means that the value of the compensation function f {k * cj, cz} of this disturbance component is measured by one value of the concentrations C] and cz, whereby the dependence coefficient k is formed by the ratio of the functions. Again, it is not necessary to determine the dependence coefficients for each device unit separately, but the dependence coefficients ka, kfo, kc, kj, etc. of each disturbance component can be determined before the start of device manufacturing. The dependence factors are thus independent of the device unit. If a basic mapping of a gas mixture system shows that the coefficient of dependence is not sufficiently constant but is of a concentration-dependent form k {cj} or k {cj, cz}, the compensation functions can be determined for two or more concentrations ci and / or cz before the start of equipment manufacturing. calculate the equation that determines the dependence coefficient k in each case. Often, however, the equation can be approximated by a constant dependence coefficient.

: 25 The concentrations of anesthetic gases are measured by any suitable method, such as infrared absorption, or by other means, such as their mixing ratio and mixing pressure. The aim is to obtain their concentration with such precision that they can be considered correct. It should also be noted that the gas component causing the collision propagation may at the same time be the gas constituting the actual measuring object, such as nitrous oxide and carbon dioxide. The above *; For one of the measurement concentrations cj and cz, a value is selected in practice from a range later in the analyzes, preferably a concentration which is in practice equal to or greater than the average of the values occurring in practice during the analyzes. It will be appreciated that the significant interference component may be any gas or the like which therefore causes a large widening of the absorption lines, so that the effect of this widening on the measurement result must be compensated at least in some way.

Claims (10)

10 102569 A method for non-dispersively analyzing gas mixtures by radiation absorption, wherein the analysis is directed to at least one predetermined gas component, and the gas mixture may contain interfering substance components that affect the radiation absorption of said comprising the steps of: (a) determining, prior to analysis, the magnitude and direction of the effect of each significant interference component 10 affecting the radiation absorption of the gas component being measured at at least one concentration of both the gas component and the interference component, characterized by (b) forming than the measurable gas component, the interference components interact with their effects 15 dependence factors; (c) prior to analysis, measuring the effect of the first non-measurable interference component of the non-measurable gas component acting on the gas component to be measured at the specified concentration; and (d) correcting the initial measurement result obtained from the radiation absorption during the analysis computationally by known concentrations of the interference components, the effect of said first interference component and said dependence coefficients for the actual measurement result. A method according to claim 1, characterized in that said known concentration of the interference component is obtained by measurement and in said step of correcting the measured measurement the measured concentration of each interference component and the coefficient of interdependence between it and the first interference component. Method according to claim 2, characterized in that said compensation coefficient is typically a non-linear coefficient with respect to concentration and said dependence coefficients are typically linear correlation coefficients or are at least approximated by linear correlation coefficients. 102569 Method according to Claim 1 or 2, characterized in that there are at least two gas components to be measured and two or more disturbance components, at least one of which is one of the gas components to be measured, the concentration measurement result of the measured gas component being compensated for by the other gas component. by a function depending on the measured concentration. Use of a method according to claim 1 for compensating and / or calibrating a measuring device, in particular for non-dispersively analysis of respiratory gas mixtures by radiation absorption, wherein the analysis is directed to at least one predetermined gas component and the gas mixture may contain interfering components. in order to reduce these said effects for the measurement result by means of predetermined effect factors of the interference components, characterized in that the effect of the interference component at a certain preselected concentration for k ytössä actual value; and during the analysis, the initial measurement result obtained on the basis of the radiation absorption is computationally corrected at least by means of the known concentrations of the interference components and their interdependence coefficients for the actual measurement result. Use according to Claim 5, characterized in that a value greater than the average of the concentrations in use is selected for a certain concentration of the interfering component · and that the interfering components are at least anesthetic gases whose concentrations in the breathing gas mixture are measured by non-dispersive infrared absorption. and whose measured concentration values are assumed to be correct. Use according to Claim 5, characterized in that the gases to be measured are nitrous oxide and carbon dioxide, the measurement results of which are compensated by the concentration of anesthetic gases, one of them by a compensation function and their interdependence factors. Use according to Claim 5 or 7, characterized in that the interference components of the nitrous oxide and carbon dioxide to be measured consist of carbon dioxide and nitrous oxide, respectively, and the measurement results of both are compensated for by means of a concentration and a compensation function. 12 102569 Use of a method according to claim 1 for compensating and / or calibrating a measuring device, in particular for non-dispersively analyzing respiratory gas mixtures by radiation absorption, wherein the analysis focuses on at least one predetermined gas component and the gas mixture may contain interfering components. to reduce these said effects for the measurement result, characterized in that the initial measurement result of the concentration of the gas component to be analyzed based on radiation absorption is corrected for the measured concentration of at least one anesthetic gas, desflurane, sevoflurane, isoflurane and enofuran, enflurane, haloth. Use according to Claim 9, characterized in that the gas component to be analyzed is carbon dioxide and / or nitrous oxide. 15