TECHNICAL FIELD

This disclosure relates to techniques for measuring in a non-invasive manner transcutaneous carbon dioxide of a person and more in particular to a wearable device and a related method for non-invasive measuring the partial pressure of transcutaneous CO2 of a person.

BACKGROUND

Telemedicine is an area that includes interactive medicine, data collection and sharing between specialists, and remote patient monitoring. This field is constantly growing and there is an increasing interest in the sector also because of the recent COVID- 19 pandemic, which has further encouraged countries to act in the sector. Telemedicine minimizes costs and ensures that you may keep your health under control even when you are unable to go to the hospital or to your doctor. Among the main support tools to make telemedicine and, in particular, remote monitoring possible we find wearable devices. They are devices that may be attached to the skin, such as patches or smart fabrics, or worn, like watches, belts, or socks. These are equipped with specific sensors for the detection of particular biological variables of clinical interest. The collected data are then made available to the user and the doctor. Lightweight and compact, they allow the subject to carry out daily activities without limiting the movements and with the minimum occupied space. In the case of the monitoring of the respiratory system, we go from systems for the analysis of oxygen saturation to those for the measurement of lung ventilation, also through devices that evaluate the quality of the ambient air we breathe.

One of the most important chemical compounds in living bodies is carbon dioxide (CO2). Its equilibrium with other chemical species and is absolute value must be in a specific range in order to guarantee life. Blood pH must be kept between 7.35 and 7.45 in order to keep a physiological condition. The value of pH in blood may be evaluated using the Henderson- Hasselbach Equation: In order to keep pH in the predefined range, there must be a physiological arterial partial pressure of CO2 (PaC02) which is 40 mmHg, and a physiological amount of bicarbonates (24 mEq/L). Formula 2.4 shows the reversible acid-base reaction in blood and any change in the amount of species involved shifts the reaction towards the right or the left. In particular, if the arterial partial pressure of CO2 (PaCO2) increases, there may be a condition of respiratory acidosis, with a pH lower than 7.35. On the other hand, if PaCO2 decreases too much there is a condition of respiratory alkalosis with a pH higher than 7.45. While over the years studies and products for the analysis of blood oxygenation have had a great success, less has been done regarding carbon dioxide. The tools available to date for this type of analysis are mainly three. Blood gas analysis (BGA) is an invasive and painful method, which does not allow continuous monitoring. Alternatively, in the context of invasive instrumentation, there is the End Tidal CO2 (EtCO2) monitoring. Transcutaneous CO2 monitoring is an alternative measurement which may be performed non-invasively by means of electrochemical sensors (Stow-Severinghaus-type PCO2 sensors). Using this technology, we have carbon dioxide electrodes produced by private companies such as SenTec. This technology requires specialized personnel and a hospital facility. In addition, continuous remembranization and calibration are other critical aspects of the apparatus. Optical sensors, already used for the analysis of carbon dioxide in EtCO2 monitoring, have been used for transcutaneous CO2 monitoring in scientific literature but no device for the analysis of the partial pressure of carbon dioxide has been commercialized.

Both electrochemical sensors and optical sensors evaluate the transcutaneous pressure of carbon dioxide (PtCO2). The mechanism through which the partial pressure of carbon dioxide is made available outside the skin must be found in the permeability to these gases to the two most superficial layers of the skin itself, which are stratum corneum and epidermis. In the deepest layer, dermis, there is a dense network of capillaries organized in predominantly vertical structures of about 0.2 - 0.4 mm. Transcutaneous CO2 measurement methods require the heating of the skin to arterialize the measurement area and increase the diffusion capabilities through the skin. If a local hyperemia is induced by warming the skin till a temperature of 42.0 °C, the supply of arterial blood to the dermal capillary bed below the sensor increases and the flow of 02 and CO2 through the skin may become 30 times higher than normal. Even at a sensor temperature of 37°C, a good correlation with PaCO2 has been reported but it takes more time to react to fast PaCO2 changes and there is an initial over-shooting of the PtCO2. In general, PtCO2 correlates well with the corresponding PaCO2 value. However, the skin carbon dioxide partial pressure is generally higher than PaCO2 and a correction factor needs to be applied.

An optical approach may be used, exploiting near infrared light.

The optical method exploits the spectroscopy approach, which is defined as the study of the interaction between matter and electromagnetic radiation as a function of the wavelength or frequency of the radiation. In particular, in the case of carbon dioxide the perfect range of wavelength is the near-infrared range, centred at 4.25 pm, because it is invisible to the human eye and safe for human beings because it has low energy. Moreover only chemical compounds that have two or more different atoms may absorb infrared light, so oxygen is not involved in this process but other species such as nitrous oxide can. However, each compounds absorb infrared at a specific wavelength so if 4.25 pm is used, only carbon dioxide may absorb it.

The milestone of spectroscopy is the Lambert-Beer Law, which basically relates the attenuation of light to the properties of the material through which the light is traveling. For this specific application, the Lambert-Beer law states that the number of infrared rays absorbed is proportional to the concentration of the infrared absorbing substance so the more CO2 is present, the more infrared light is absorbed. The system is composed by three main elements: an infrared source which emits wavelengths within a narrow band around 4.25 pm, a sample chamber and a detector. There is one side effect called "collision broadening”. In air and body diffused gasses, there are also oxygen and nitrous oxide in addition to CO2. These molecules collide with the CO2 molecules slightly altering the way they absorb infrared waves, leading to the absorption pattern broadening which may cause a potential source of error in measurement. EtCO2 systems measure the amount of nitrous oxide and oxygen present and use this information to correct for errors due to collision broadening.

In the patent publication WO2016/173877, another PtCO2 device is presented, with its schematic block diagram shown in Figure 1. This prior sensor is composed by a measuring chamber, which is separated from the skin through a membrane. Gasses diffuse from the skin through two chimneys. The measuring system is made by a LED and a photodetector. There are two thermistors and a heating element in order to warm the skin and control the temperature. Every component of this sensor is specifically customized for the application, creating a functional but expensive sensor.

The article by Vishal Varun Tipparaju, Sabrina Jimena Mora, Jingjing Yu, Francis Tsow, and Xiaojun Xian. Wearable transcutaneous co monitor based on miniaturized nondispersive infrared sensor. IEEE Sensors Journal, 21(15):17327- 17334, Aug. 2021, discloses a wearable device for continuous monitoring of PtCO2 using a Cozir NDIR CO2 sensor. This prior wearable wristband device, schematically represented in Figures 2 and 3, has been validated against EtCO2 in human test. The profile of the PtCO2 measurement correlates very well with the EtCO2 trend in different test conditions. The output signal of Cozir CO2 in ppm is converted in mmHg by using the calibration plot. There is not a heating element and it suffers from a delayed response to rapid variations of PaCO2. In fact, the wristband has a lag of 5 minutes with respect to EtCO2. This wearable device is equipped with a sealing O-ring and a hydrophobic membrane made of PDMS in order to protect the CO2 sensor from the water vapor diffused by the skin. Indeed, water molecules interfere NDIR sensors because water has a high extinction coefficient at wavelengths near 4.26 pm.

In order to deeply understand how the prior non-invasive optical devices for measuring the PtCo2 work, a first prototype has been realized considering what is disclosed in the prior documents WO 2016/173877 and in the cited paper by Vishal Varun Tipparaju et al..

In this first prototype, not according to this invention, the main chosen components were the following:

- a Cozir-A non-dispersive infrared (NDIR) sensor for the analysis of carbon dioxide.

- a NiChrome wire with a resistivity of 2.308 Q/m as heater.

- a module mounting IRF520 power MOSFET to increase the supplied current for heating.

Additionally, two NTC thermistors with a nominal resistance of 100 KQ were chosen, one to detect the temperature of the skin and one for that of the wire. The system was also equipped with a LED and a momentary switch to facilitate user interaction. The microcontroller used is Arduino Nano 33 BLE Sense which includes the module for communication via Bluetooth Low Energy. A case was built with the use of the SOLIDWORKS® program to host the various components and a firmware on Arduino IDE was implemented.

A fabric cuff was provided to attach the device to the skin.

A sensitivity to the presence of the skin is highlighted through the comparison between the output values of the CO2 sensor in air and the one’s in contact with the skin.

To obtain a reliable measurement of PtCO2, the skin temperature has to be increased, ideally up to 42 °C. The case had a measurement chamber where the air passes before entering the sensor. The sensitivity to the presence of the skin could be detected with this first prototype, shown in detail in Figures 4a, 4b and 4c. This first prototype, which was functionally similar to the device disclosed in WO 2016/173877 and in the article by Vishal Varun Tipparaju et al., surprisingly was largely inaccurate for no apparent reason.

SUMMARY

Deep studies carried out by the inventors about why the first prototype gave so largely inaccurate measurements, lead the inventors to investigate about what takes place into the chamber. Without being bound to a theory, it has was made the hypothesis that carbon dioxide saturated the chamber after a few minutes of testing and this could have been a main cause for inaccuracy of the first prototype.

To solve this problem, it has been designed a wearable device for non-invasive measurement of the partial pressure of transcutaneous CO2 of a person as defined in the enclosed claim 1.

A method of measuring in a non-invasive manner a partial pressure of transcutaneous CO2 of a person is also disclosed. This method may be implemented by running a software in a microcontroller of the wearable device of this invention.

Preferred embodiments are defined in the dependent claims.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 depicts a prior device for measuring partial pressure of CO2 at a skin of a person.

Figures 2 and 3 depict another prior wearable wristband device for measuring partial pressure of CO2 at a skin of a person, comprising a membrane for separating the person's skin to a chamber for CO2 collecting gases emitted by the skin.

Figures 4a, 4b and 4c illustrate a first prototype, not according to the present invention, of a device for measuring partial pressure of CO2 at a person's skin.

Figures 5 and 6 are sectional views of a wearable device of this invention highlighting the channel and the inlet for putting the chamber of the device into fluid communication with ambient air.

Figures 7 and 8 show the outer case of the second prototype of the wearable device of this invention, highlighting the cavity in which the valve for closing/opening the channel is installed.

Figure 9 is a picture of the second prototype after having removed a part of the outer case to show how the valve is installed.

Figure 10 is a view of the outer case of the second prototype of the wearable device of this invention, highlighting a seat in which a temperature sensor is installed around a perimeter of the opening of the chamber, destined to be in contact with a person's skin.

Figure 11 illustrated the three phases of the method for measuring a partial pressure of CO2 using the wearable device of this invention.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

A wearable device of this invention for non-invasive measurement of the partial pressure of transcutaneous CO2 (PtCO2) of a person will be disclosed referring to the enclosed figures from 5 to 11. It is a self-contained device that comprises:

- a housing (1) for a CO2 sensor defining a chamber (2) for measuring gases, having an opening at one bottom wall configured to be airtight closed by the skin of the person when the wearable device is worn so as CO2 emitted by the person’s skin, within the perimeter of the chamber, enters in the chamber (2) through the opening.

- a CO2 sensor (not shown in the figures), installed into the housing (1) to close the chamber (2) from one top wall opposite to the opening and oriented toward the opening in order to detect CO2 emitted by the person's skin, within the perimeter of the chamber, that closes the opening, configured to generate a signal representative of a CO2 concentration in a portion of the chamber (2) delimited by the CO2 sensor and the person's skin;

- one heater (not shown in the figures) installed into the housing (1) and configured to warm the portion of the person's skin, within the perimeter of the chamber, that closes the opening of the chamber (2).

- a microcontroller (not shown in the drawings) functionally coupled with the CO2 sensor for receiving the signal representative of a CO2 concentration, and functionally coupled with the one heater to warm the person's skin, the microcontroller being configured to output a value of CO2 concentration, measured into the chamber, corresponding to the signal representative of a CO2 concentration.

Differently from what is disclosed in WO2016/173877, the chamber (2) is not closed by a membrane, but it is closed by the person's skin in direct contact with the opening of the chamber (2).

Differently from to the prior device disclosed in the cited paper by Vishal Varun Tipparaju et al., the housing (1) further defines an inlet (3) to the chamber (2) and a channel (4) configured to put into fluid communication the inlet (3) of the chamber (2) with ambient air. More in detail, the housing (1), the CO2 sensor and the inlet (3) are configured so as, when the person’s skin closes the opening of the chamber (2), ambient air outside the wearable device may enter in the chamber (2) only passing throughout the channel (4) and the inlet (3). In practice, when the wearable device is worn by the person to carry out a measurement, the channel (4) is the only way for putting the chamber (2) into fluid communication with the ambient air.

As stated hereinafter, it has been discovered that the prior device disclosed in the cited paper by Vishal Varun Tipparaju et al. may give largely inaccurate values of CO2 concentration because the chamber may saturate with CO2, thus the sensor becomes unable to carry out new measurements. To overcome this unexpected problem, the wearable device of this invention comprises a controlled valve installed into the channel, configured to close or open the channel respectively in order to seal or to put in fluid communication to ambient air the chamber when the wearable device is worn by the person. This valve is controlled by the microcontroller, which is configured to open/close the controlled valve and to display a measured value of CO2 concentration into the chamber. The controlled valve must be opened to allow recirculation of ambient air before a new measure of CO2 is to be started, and to be closed when the CO2 measurement is performed. In practice, when the wearable device is worn with the opening being airtight closed by the skin of the person, air in the chamber is flushed only when the controlled valve is open; moreover, air remains sealed in the chamber only when said controlled valve is closed.

To test the effectiveness of the new device, a second exemplary prototype according to this invention has been realized. With respect to the first prototype, the carbon dioxide sensor was replaced with a more compact Cozir-LP NDIR CO2. A normally open solenoid valve, an antiparallel diode and another MOSFET module were used to power the valve. A rechargeable 3.7 V Lithium-ion battery and a latching switch to turn the device on and off were selected. Everything was then soldered on a custom-made Printed Circuit Board. A housing was designed in SOLIDWORKS® and a bracelet was realized to attach the device to the person's wrist and to hold the battery.

The main components used for the second prototype are the following:

• Arduino Nano 33 BLE as microcontroller

• A Cozir-LP NDIR CO2 sensor, used to sense CO2

• A NiChrome wire used as a heater to warm the skin of the subject with a resistivity of 5.755 Q/m

• One NTC thermistor, as a temperature sensor for temperature sensing, with a nominal resistance of 100 KQ

• A normally-open (NO) solenoid valve

• One green LED that blinks during the calibration phase

• One red LED that is on when the solenoid valve is closed

• One blue LED that is on when the heater reaches the desired temperature

• A diode to discharge the electrical energy accumulated by the valve

• A momentary switch

• Two MOSFET modules (SSM3K615R produced by Toshiba) to increase the amount of current in the hot wire and for the solenoid valve • One 3.7 V Lithium-Ion battery

• A miniature latching switch

The results of the first prototype highlighted the need of an output air channel. At the same time, during the acquisition period of the PtCO2, preferably the chamber connected to the CO2 sensor must be as closed as possible in order to be more sensitive to the presence of the skin.

The miniature pneumatic solenoid valve with two ports and normally open (NO) by Parker, was the chosen solution. Its low power design reduces heat generation and power consumption (0.36 W). It is compact with an economical design to reduce size and cost of integration.

A MOSFET module is used to deliver the sufficient amount of current. When the valve passes from the active to the inactive phase, a flyback diode is used for discharging the electrical energy accumulated by the valve during the active phase. The diode in antiparallel configuration with respect to the valve protects from damages that may occur.

The 3-D housing is designed in the Fusion 360 software. The chosen material for the 3D-printed case is the DraftGrey material used on a Stratasys PolyJet Printer, mostly because of its resistance to high temperatures.

The volume of the chamber where CO2 diffuses from the skin to the CO2 sensor had to be as small as possible to create a sensitive device to analyse PtCO2.

The height of the chamber was not reduced further because it must accommodate the air inlet (3) of the channel (4), shown in Figures 5 to 7. This channel (4) is closed by the solenoid valve (5), which may be either open or closed, according to the specific phase of the measurement. In Figure 8 it may be observed the part of the case that hosts the solenoid valve (5). Figure 9 depicts the second prototype in which the solenoid valve (5) is installed.

According to an aspect, the heater, for example a NiChrome wire, for heating the person's skin that closes the opening of the chamber (2) may be located in the groove (6) as shown in Figure 10 around the perimeter of the opening, in contact with the person’s skin and covered by thermal paste. In this case, the temperature sensor, that may be for example a thermistor, may be installed too in the groove 6 in direct contact with the person's skin and the heater. As an alternative, in case the heater (NiChrome wire) is not in direct contact with the person's skin, as shown in Figure 4a making reference to the first prototype not according to the present invention, there may be a first temperature sensor in direct contact with the person's skin, installed in the housing at the position 7, and a second temperature sensor installed in direct contact with the heater to sense the temperature of the heater itself.

The wearable device according to the present invention allows to measure in an accurate manner the partial pressure of transcutaneous CO2 and thus to obtain a surrogate of the partial pressure of CO2 in the person's blood. This is done by programming the microcontroller of the wearable device to implement three phases, summarized in Figure 11: the first phase involves the analysis of environmental carbon dioxide; the second is the heating of the skin up to a nominal temperature, that may be for example 38.5 °C, or for a nominal time interval (in case said nominal temperature is not attained at the expiring of the nominal time interval); and a last stage of analysis of the carbon dioxide diffused by the skin.

According to a basic embodiment of the method of this disclosure, once the person has worn the wearable device of the invention, the microcontroller commands the controlled valve of the wearable device, for example by pushing a button, so as to cause the CO2 sensor measure an ambient CO2 concentration in the environment, to provide an offset value of CO2 concentration. Once the ambient CO2 concentration has been measured, the microcontroller commands, for example by pushing the same button again or a second button of the wearable device, the controlled valve for closing the channel to seal the chamber. Then, the NiChrome wire starts heating the person's skin up to the nominal temperature (for example 38.5 °C) and/or for a nominal time interval. At the end of this nominal time interval, the CO2 sensor measures a CO2 concentration in the chamber, which is sealed by the person's skin and by the closed valve.

The ambient CO2 concentration may be measured before wearing the device of this invention, or even after the device has been worn. In the latter case, the microcontroller may command the controlled valve of the wearable device to open the channel and to flush the chamber with fresh ambient air from an environment in which the wearable device is located. For example, this may be done when a first measurement has just been carried out and a new measurement is desired. Before carrying out a new measurement, with the wearable device still worn on the person's wrist, the controlled valve may be opened to flush the chamber with ambient air to repeat the above method steps.

Preliminary results showed the great impact of the ambient CO2 in the measurement of skin CO2. There is a different offset in the analysis of skin CO2 according to different ambient CO2 concentrations. For this reason, in the second prototype the first state of the algorithm is aimed to evaluate the CO2 concentration near the subject before passing to the heating phase.

In the wearable device realized according to the second prototype, the wearable device may be connected, using wires or in wireless mode, to a smartphone or PC to display the results of measurements collected by the microcontroller of the wearable device. After having connected the device to a smartphone or PC, the digital CO2 values are read at 1 Hz, with the CO2 sensor in polling mode. When the difference between 10 following samples is less than 10 ppm, the signal is considered stable and is indicated by the green LED stopping its blinking. The average of the 10 values is read by the measurement system. It may be used to consider the offset in CO2 skin measurement. Now the device may be put on the wrist and, after pressing the button, the system passes in the heating phase.

The heating process starts by setting the Duty Cycle (DC) of the analog pin connected to the NiChrome wire to 35%. Now the system enters in a loop: the device measures skin and wire temperature (note: the wire is in contact with the skin) with a frequency of 1 Hz. All the specification of this heating process have been determined after several experimental attempts. If the skin/wire temperature reaches 44 °C, the DC is set to 0% to avoid burning of the skin.

When the skin temperature reaches the target temperature, for example 38.5°C, for over 10 seconds, the third phase may begin. The time needed for this phase is in the order of minutes, according to the subject and to the initial skin temperature. The process may be stopped at any moment by pressing the button and the subject may put off the device whenever he wants, so as a prototype of a biomedical device it meets the basic safety conditions.

In this state the PtCO2 measurement from the skin takes place and the output values of the sensor are streamed to the smartphone/PC paired at the beginning. The sampling frequency is 1 Hz and the CO2 sensor is used in polling mode. According to the different tests, carried out with the second prototype, this phase has different duration, but there is an initial phase lasting 3 minutes in order to have a stable signal. The functioning of the device was evaluated with the above-mentioned device for

PtCO2 monitoring marketed by SenTec and it has been proven able to detect an increase in PtCO2.