Noninvasive optical determination of partial pressure of carbon dioxide

Abstract

Herein is disclosed a sensor for noninvasive measurement of the partial pressure of CO2 (pCO2) in the skin of a human. The sensor comprises a housing, a gas measuring chamber for measuring gases, at least one chimney for communication of gases diffusing through the skin to the gas measuring chamber, a broad band light source transmitting light into the gas measuring chamber and a detector system comprising a first and a second photodetector. The first photodetector detects light at a wavelength wherein CO2 absorbs light and the second photodetector acts as a zero reference detector by measuring light in a freeband where no gases absorb light.

Claims

1. A sensor for noninvasive measurement of the partial pressure of CO.sub.2 (pCO.sub.2) in the skin of a human comprising: a housing, a gas measuring chamber for measuring gases, at least one chimney for communication of gases diffusing through the skin to the gas measuring chamber, a broad band light source transmitting light into the gas measuring chamber, and a detector system comprising a first and a second photodetector for detecting light transmitted from the light source through the gas measuring chamber, wherein the first photodetector detects light at a wavelength wherein CO.sub.2 absorbs light, wherein the second photodetector acts as a zero reference detector by measuring light in a freeband where no gases absorb light, wherein the light source is a light emitting diode, and wherein the second photodetector is connected to the light source in a feedback loop through a signal processor, for auto-calibrating the light source according to the amount of light received at the second photodetector.

2. A sensor according to claim 1, wherein the freeband is centered at 3.9 m.

3. A sensor according to claim 1, wherein the gas measuring chamber has an inner surface of a reflective material comprising gold, silver, or aluminum forming a reflective tube.

4. A sensor according to claim 3, wherein the roughness of the reflecting tube is less than of the wavelength of the light absorbed by the gas to be determined.

5. A sensor according to claim 1 further comprising a beam splitter for splitting the light from the collecting lens between the photodetectors of the sensor.

6. A sensor according to claim 1, wherein the measuring chamber has two opposing openings, wherein the first opening is closed by the light source and the second opening is closed by the detector system, thereby providing a straight line between the light source and the detector system.

7. A sensor according to claim 6, wherein the at least one chimney for communication of gases diffusing through the skin to the gas measuring chamber is perpendicular to the gas measuring chamber.

8. A sensor according to claim 1 wherein the light source is an infrared light source.

9. A sensor according to claim 1 wherein the signal processor is connected to the first photodetector and the second photodetector.

10. A sensor according to claim 9, wherein the signal processor calculates the partial pressure of CO.sub.2 based on the difference between the signals received at the first and the second photodetector.

11. A sensor according to claim 1, wherein the total volume of the gas measuring chamber and the at least one chimney is no more than 2 L.

12. A sensor according to claim 1, wherein the first photodetector has a band pass filter centered at 4.26 m.

13. A method for calculating the partial pressure of CO.sub.2 in the skin of a human comprising: irradiating a CO.sub.2 gas sample in a gas measuring chamber by a broad band light source, detecting the light transmitted through the gas measuring chamber by a first and a second photodetector, wherein the first photodetector detects light at a wavelength wherein CO.sub.2 absorbs light, and wherein the second photodetector detects light at a freeband where no gases absorb light, and calculating by means of a signal processor the CO.sub.2 partial pressure based on the difference between the signals received at the first and the second photodetector, wherein the light source is a light emitting diode, and wherein the second photodetector is connected to the light source in a feedback loop through the signal processor, for auto-calibrating the light source according to the amount of light received at the second photodetector.

14. A sensor according to claim 3, wherein the roughness of the reflecting tube is less than of the wavelength of the light absorbed by the gas to be determined.

15. A sensor according to claim 5, wherein the beam splitter is chosen from a reflective prism and a reflective grating.

16. A sensor according to claim 9, wherein the signal processor calculates the partial pressure of CO.sub.2 based on the difference between the signals received at the first and the second photodetector.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The above and other aspects will be apparent and elucidated from the embodiments described with reference to the drawing in which:

(2) FIG. 1 shows a schematic block diagram of a sensor for measuring the partial pressure of CO.sub.2 in the skin of a human patient.

(3) FIG. 2 shows detailed view of the Non Dispersive Infrared (NDIR) detector for measuring partial pressure of CO.sub.2.

(4) FIG. 3 shows a different embodiment of the NDIR detector for measuring partial pressure of CO.sub.2 wherein the prism (27) is replaced with an optical grating (28).

(5) FIG. 4a shows the absorption spectra for water and carbon dioxide in the spectrum 0-10 m.

(6) FIG. 4b shows a close-up of the absorption spectra of FIG. 4a focusing on the spectrum of 3.8 m to 4.6 m.

(7) FIG. 5 shows a diagram of the gas absorption spectra for CO.sub.2 and the filter function of the active channel and the reference channel, respectively.

(8) FIG. 6 shows a sensor for transcutaneous measurement of the partial pressure of CO.sub.2 in the skin, communicating with a monitor via a wired connection.

(9) FIG. 7a shows the warm-up time measured for the sensor versus the temperature.

(10) FIG. 7b shows the warm-up time of the sensor versus the relative PCO.sub.2[mmHg].

DETAILED DESCRIPTION

(11) FIG. 1 shows a schematic block diagram of an example of a sensor (10) for measuring the partial pressure of CO.sub.2 in the skin (1) of a human patient. The sensor (10) comprises a housing, a measuring unit (20), a patient interface in the form of a membrane (13) contacting the skin (1) of the patient when in use, and protecting the sensor (10) against e.g. moisture and dust. Furthermore, the sensor comprises thermistors (11) for measuring the sensor temperature, used for controlling a heating element (17). The temperature readout from the thermistors (11) is further used for signal processing when converting the detector signal into a partial pressure of CO.sub.2 e.g. by the Severinghaus equation. The two chimneys (18) connect a measuring chamber (23) with the skin surface through the membrane (13), whereby gases diffusing through the skin are transported through the membrane (13) and the chimneys (18) to the measuring chamber (23). The measuring chamber (23) is closed off at one end by a light source (21), and at the opposite end by a detector system in the shape of a dual channel detector (15). The skin (1) of the patient, the chimneys (18) and the measuring chamber (23) define a closed system. The electronic board (14) comprises amplification, filtering, A/D converter, and signal processing means. The electronic board (14) communicates with the dual channel detector (15), to convert the signals received from the dual channel detector (15), into a value representing the partial pressure of blood gases in the skin of the patient. Furthermore the signal processing at the electronic board (14) receives information from at least one of the thermistors (11) regarding the sensor (10) surface temperature to control the heating element (17) that heats the sensor (10).

(12) The heating element (17) is used to warm up the sensor (10), which again warms up the skin of the human patient. The thermistors (11) are also used to control the heating element (17), to achieve the correct applied sensor (10) temperature and avoid burning the skin of the patient. The design as described with respect to FIG. 1 has two thermistors (11). One thermistor (11) would be enough for the purpose, but it has become a standard within the industry to have two thermistors (11), doubling the sensor (10) temperature control since it is important to measure the correct temperature for calculating the blood gas partial pressure and to avoid burning the skin of the patient. The heating of the sensor (10) is programmed by the hospital staff, defining measuring time and temperatures. The electronic board (14) also takes care of communication with external equipment (not shown), e.g. a monitor, controller unit connected via electrical cables or smart phones, computers, or tablets connected via wireless means such as WiFi, Bluetooth, GSM or like network.

(13) In an embodiment, the electronic signal processing is performed outside the sensor (10). Hence the electronic board (14) is located in a device, e.g. a monitor external to the sensor (10). Thermistors (11), heating element (17), light source (21), and the dual channel detector (15) is connected to the electronic board (14) in the external device via electrical cabling e.g. in the form of a bus. Preferably A/D and D/A conversion are performed in the sensor (10), such that the communication between the sensor (10) and the external device is digital wired or wireless communication.

(14) The membrane (13) protects the internal parts of the sensor (10), e.g. the electronics, the chimneys (18), and the measuring chamber (23) against intruding particles such as moisture and dust. The membrane (13) is permeable to blood gases, hydrophobic, mechanically robust and does not change the permeability or structural behavior when in contact with alcohol swab, contact gel or sweat. It allows the blood gases to diffuse from the patient skin to penetrate the membrane (13), while blocking dust and moisture from penetrating the membrane (13) to potentially harm the sensor (10), measuring parts, electronics, or block the chimneys (18) and thus obstructing the passageway for the blood gases between the skin of the patient and the measuring chamber (23).

(15) In one embodiment of the invention the chimneys (18) have a diameter of 200 m and a length of no more than 4 mm. The length of the chimneys (18) should preferably be as short as possible, since the volume of gas within the chimneys (18) adds to the total volume of the gas measuring chamber (23), hence the volume here has negative effect on the sensitivity. Some length is however necessary to allow the photodetector and the light source to be positioned at either side of the measuring chamber (23), and allow the thermistors (11) to be fitted on the surface of the sensor as they need to be close to the skin to detect the skin temperature. The diameter of the chimneys (18) also adds to the total volume and should be chosen as small as possible to in order to keep the response time of the sensor low. On the other hand, larger diameter would decrease the risk of the chimneys (18) being blocked by intruding particles and would also allow a better flow of gas between the skin and the measuring chamber (23). Hence also here a diameter of 200 m is a compromise. Two chimneys (18) have been chosen since one chimney may increase the risk of partly or fully blocking the chimney, whereas three or more chimneys (18) increase the total volume and the response time. The measuring chamber (23) has a total volume (here including the volume of the two chimneys (18)) of no more than 2 L, due to the small volume of gas diffusing through the skin per time unit, a large volume would dramatically increase the response time of the sensor (10), which should preferably be no more than one minute. Hence a volume of no more than 2 L has been found to be a good compromise. The interaction length of the measuring chamber (23), i.e. the length of the measuring chamber (23) wherein the light may interact with the blood gas molecules is in the range of 1.5 mm to 30 mm, see also table I.

(16) FIG. 2 is a detailed schematic block diagram of the gas measuring unit (20) comprising the light source (21) transmitting light into the measuring chamber (23) via an IR transparent glass window (22) made of e.g. ruby or sapphire, protecting the light source (21) and sealing the measuring chamber (23), and a collimating lens (24) sealing the opposite end of the measuring chamber (23). The collimating lens (24) is collecting and rectifying the light such that all light are focused at a prism (27). The collecting lens defines the focal point and is chosen such that the prism (27) splits the light equally between two photodetectors (25) and (26). The collimating lens (24), The prism (27), and the two photodetectors (25, 26) together make up the dual channel detector (15). The geometry between the collimating lens (24), the prism (27) and the photodetectors (25, 26) is such that the majority of the light reaching the collimating lens (24) also reach the sensitive point at each of the photodetectors (25, 26) to optimize the sensitivity. The prism (27) is coated with an IR reflective coating, e.g. aluminum, gold, silver, or other suitable material.

(17) To further increase the sensitivity the gas measuring chamber (23) and the prism (27) should have a smooth finishing to ensure a good forward reflection, i.e. ensuring that as much light as possible travels from the light source (21) to the photodetectors (25, 26). Hence the roughness should be to 1/20 , at 632.8 nm. A smooth surface will provide a better signal-to-noise ratio, as a larger percentage of the light will be transmitted from the source (21) to the photodetectors (25, 26), whereas a rougher surface will give more back reflection, or scattering, and reduce the percentage of light reflected back towards the source. The gas measuring chamber (23) may be drilled out of a solid piece of metal or another suitable material. The desired roughness is achieved using suitable tooling and/or polishing. Alternatively, the gas measuring chamber (23) may be molded in plastic and then spray coated to achieve the desired reflectivity and roughness.

(18) FIG. 3 shows a schematic diagram of an alternative gas sensing unit. The only difference from FIG. 2 is that the prism (27) is replaced with a reflective grating (28) instead of the prism (27). The reflective grating (28) has the advantage compared with the prism (27) that it does not need to be aligned as accurate as the prism (27). However with a correct alignment, the prism (27) is more effective in distributing the light between the two photodetectors (25, 26).

(19) FIG. 4a shows the absorption spectra of water and carbon dioxide, in the mid IR spectrum, from 0-10 m. The spectrum is created using the Hitran database maintained by the Atomic and Molecular Physics Division, Harvard-Smithsonian Center for Astrophysics. Many more gases absorption spectra's are available in the database, but under normal circumstances only water vapour, carbon dioxide, and oxygen diffuses through the skin, and oxygen does not absorb light. Hence it is only relevant to look at CO.sub.2 and H.sub.2O. As it may be seen, water vapour dominates most of the spectrum, but there are narrow bands where CO.sub.2 dominates and also at least one band, where there is essentially no absorbance at all, both are located around the 4 m wavelength.

(20) FIG. 4b shows a close-up at the spectrum from FIG. 4a around the 4 m wavelength, where it is already determined that CO.sub.2 has a band with great absorption, i.e. around 4.26 m, and where there is also a free band where essentially no gasses absorb light, i.e. around the 3.9 m wavelength. This graph may further be used for designing the filters such that they cover a suitable bandwidth and has appropriate steep filter coefficients, such that the photodetector covers exactly the band of interest, e.g. 4.26 m90 nm for CO.sub.2 and 3.9 m45 nm for O.sub.2.

(21) FIG. 5 shows the absorption coefficient/cm.sup.1 on the left vertical axis and the filter transmission in percentage on the right vertical axis versus the wavelength in m on the horizontal axis. The spikes show the gas absorption for CO.sub.2 and the curved line shows the filter function of the reference channel (left) and the active channel (right). The reference channel is chosen in a freeband where no gas molecules absorb light and hence all light emitted at this wavelength will pass through to the detector and due to the rather steep gradient of the filter function, and the limited band pass filter, of the reference channel detector it will not receive light that is likely to be absorbed by gas molecules. The filter function of the active filter is chosen such that it covers the absorption spectrum for the gas to be determined. The active channel will hence receive light inversely proportional to the amount of gas in the measuring chamber (23). Referring to FIG. 4, it is further seen that there is also a freeband just above the CO.sub.2 absorption band where a reference channel could be placed.

(22) FIG. 6 shows a sensor (10) for optically measuring the partial pressure of CO.sub.2 in the skin of a patient. The sensor is connected to a monitor device (34) via electrical wires (35). The monitor is further connected to the power grid and hospital information infrastructure (not shown). Although the sensor (10) comprises signal processing means, some of the signal processing may still be conducted by the monitor (34).

(23) FIG. 7A shows the warm up time for a sensor according to the invention, wherein the sensor temperature [ C.] is plotted as a function of time [minutes]. The temperature is set to 43 C. and the power supplied to the sensor is limited to 500 mW (normally sensors are limited to 1000 mW supply power). After just 3.8 minutes (3 minutes and 48 seconds) the temperature is stable at 43 C. but already after 2.2 minutes (2 minutes and 12 seconds) stable CO.sub.2 levels can be determined. Increasing the power limit to 1000 mW, as the todays commercially available sensors, will approximately cut the warm-up time to half the time, i.e. under 120 seconds for reaching a stable temperature of 43 C. and approximately 60 seconds for reaching a stable CO.sub.2 level.

(24) FIG. 7B shows the relative CO.sub.2 partial pressure for the warm up time corresponding to FIG. 7A for a test gas with 10% CO.sub.2. It can be seen that the CO.sub.2 level stabilizes at around 2 minutes from start-up where also the sensor temperature stabilizes. The graph shows 4 different curves with approximately the same course but reaching different levels of CO.sub.2. The difference in CO.sub.2 level reflects that different light sources have been used giving different sensitivity of the sensor, i.e. the system provides the same stabilization point independent of the light source used. The differences in the levels are handled in the initial calibration process.

EXAMPLE

(25) The sensor (10) as described with reference to FIGS. 1 and 2 is to be placed at the patient's skin (1) for measuring the partial pressure of blood gases in the skin. The membrane (13) is permeable to such gases, whereas moisture, dust etc. is blocked from entering the sensor core. The blood gases travel through the two chimneys (18) and into the measuring chamber (23). The measuring chamber (23), the chimneys (18) and the skin define a closed system. The amount of blood gas molecules in the measuring chamber (23) will thus in time reflect the partial pressure of blood gases in the skin tissue of the patient. At one end of the measuring chamber (23), a light source (21) is located, emitting light into the measuring chamber (23). In the measuring chamber (23) light at 4.26 m is absorbed by CO.sub.2 molecules in the measuring chamber (23) which lead to a weakening of the light intensity at this wavelength, whereas light at 3.9 m is not absorbed and thus passes through the measuring chamber (23) to be detected by the detector (15). The difference between the light detected with the reference channel (3.9 m) of the detector and the light detected with the active channel of the detector (4.26 m) reflects the amount of gas in the measuring chamber (23) and hence the partial pressure of CO.sub.2 in the skin tissue of the patientsee FIGS. 2 and 3 for active and reference channel. The difference is calculated by the signal processing device at the electronic board (14). The heating element (17) controls the temperature of the sensor, which in turn controls the temperature of the skin of the patient. Depending on the conditions of the patient it may be necessary to warm up the skin of the patient more in some circumstances than in other. The skin is warmed to temperatures between 25 and 45 C. The temperature is controlled by the feedback system comprising at least one thermistor (11) measuring the sensor surface temperature and feeding back signals to the signal processor at the electronic board (14) controlling the heating element (17).

(26) Tests have been conducted to verify the sensor capability. The tests have been conducted according to Class II special controls guidance document: Cutaneous carbon dioxide (PcCO.sub.2) and Oxygen (PcO.sub.2) monitors; Guidance for industry and FDA (document issued 2002), and according to ASTM F984.

(27) The sensor was calibrated (initial calibration) by exposing the sensor to gases 1-9, Table II, with a gas flow in the range of 10-20 mL/min in 10 minutes (comprising stabilizing period for each test gas). The last 30 seconds of data is recorded for each gas to establish a median reading. The median values are used as single point for each concentration for the calibration curve. The sensors are further exposed to the gases 1-9 during 10 minutes, before reading the PCO.sub.2 values. The last 200 seconds are recorded to evaluate the mean value, the minimum and maximum value and the standard deviation.

(28) TABLE-US-00002 TABLE II Test gases Gas [CO.sub.2] [O.sub.2] [N.sub.2] Test Gas 1 0% CO.sub.2 Rest N.sub.2 Test Gas 2 1% CO.sub.2 90% O.sub.2 Rest N.sub.2 Test Gas 3 3% CO.sub.2 Rest N.sub.2 Test Gas 4 5% CO.sub.2 10% O.sub.2 Rest N.sub.2 Test Gas 5 7% CO.sub.2 12% O.sub.2 Rest N.sub.2 Test Gas 6 10% CO.sub.2 Rest N.sub.2 Test Gas 7 15% CO.sub.2 50% O.sub.2 Rest N.sub.2 Test Gas 8 20% CO.sub.2 20.9% O.sub.2 Rest N.sub.2 Test Gas 9 30% CO.sub.2 Rest N.sub.2

(29) The differences between the calculated values using the gas sensor and the test gas values was determined as can be seen in Table III.

(30) TABLE-US-00003 TABLE III Calculated gas values pCO.sub.2 pCO.sub.2 mean Standard Test calculated Min Max measured deviation Gas [CO.sub.2] % [mmHg] [mmHg] [mmHg] [mmHg] 3 [mmHg] Test 0 0 0 0.4 0.2 0.3 gas 1 Test 1 7.4 7.8 8.3 8.1 0.3 gas 2 Test 3 22.1 21.7 22.5 22.1 0.6 gas 3 Test 5 36.9 36.4 37.7 37.1 0.7 gas 4 Test 7 51.6 50.9 52.2 51.5 0.9 gas 5 Test 10 73.8 73.3 75.9 74.4 1.5 gas 6 Test 15 110.6 106.7 109.6 108.2 1.8 gas 7 Test 20 147.5 145.9 150.5 148.4 2.8 gas 8 Test 30 221.3 214.1 221.7 218.5 4.9 gas 9

(31) The response time of the sensor has been determined according to the class II Special Control Guidance for tcPCO.sub.2, ASTM F984 and IEC 60601-3-1. The maximum 10-90% response time is recorded as can be seen from Table IV. The response time was measured at 44 C., the temperature correction factor was set to 1 and the metabolic offset was 0 (zero) mm Hg. The in vivo correction is also 0 (zero) mm Hg. The test procedure was repeated three times. The test was conducted on a sensor as described herein with two chimneys (18) covered by a membrane (13). As can be seen from the 6.sup.th column, the acceptance criterion is a response time below 60 seconds. As can be seen the SMD5 component, has a significant longer response time than the rest. The reason for this slow response is that one chimney was blocked, and hence the gases were only able to enter the test measuring chamber (23) via one chimney. Despite the blockade of one chimney (18), the response time is still below the acceptance criteria.

(32) TABLE-US-00004 TABLE IV Response time 10-90% response time [s] Direc- System 1.sup.st 2.sup.nd 3.sup.rd Acceptance Test results tion No. Cycle Cycle Cycle Criterion Passed Failed Test SMD1 41 41 40 60 sec for x Gas 3 PI#71 PCO.sub.2 to Test SMD2 37 38 36 x Gas 6 PI#72 SMD3 36 40 34 x PI#73 SMD5 59 59 54 x PI#74 Test SMD1 44 42 44 x Gas 6 PI#71 to Test SMD2 39 40 41 x Gas 3 PI#72 SMD3 37 37 39 x PI#73 SMD5 58 57 56 x PI#74