Novel Approach for Asthma Detection Using Carbon Monoxide Sensor ^†

Abstract

Around 339 million people suffer from asthma worldwide. An acute asthma attack causes difficulties in daily life activities and can sometimes be fatal. The unnecessary challenges faced by asthmatics signifies the need for a device that helps people monitor and control asthma to prevent possible attacks. A number of studies have reported an elevation of carbon monoxide in exhaled breath (eCO) of asthma patients and suggest that this can be used as an effective biomarker of lung inflammation. By making use of the reported results, this project aims to make use of the eCO biomarker to design a carbon monoxide (CO) asthma monitoring system. The system consists of a Raspberry Pi 3 microcontroller and a MQ 7 CO sensor for processing and detecting the carbon monoxide concentration in parts per million. For accurate results, a face mask is attached to the sensor to mitigate environmental CO. The working of the sensor circuit is validated using a carbon monoxide source. With more researchers focusing on the threshold level of CO for an imminent asthma attack, this CO sensor could eventually save lives and improve standards of living while being an affordable and user-friendly device for active lifestyles.

1. Introduction

Asthma, a respiratory ailment, is categorized as chronic inflammation in the pulmonary airways, along with oxidative stress in the bronchial tubes. A significant characteristic is the swelling in both proximal and distal lung airways [1]. An asthma patient can experience sudden or unexpected asthma attacks that can be triggered by any factor, and the factors vary from person to person. The intensity of an acute asthma attack is unforeseeable and has the possibility to be fatal [2].

Asthma affects human beings of all ages, and it generally begins in childhood. Asthma is a medical issue increasingly affecting people globally. As reported by the Global Asthma Report in 2018, it was believed that around 339 million people had asthma, and there were 417,918 deaths due to asthma worldwide [3]. India has around 15–20 million asthmatics, 6% of the child and 2% of the adult population. The phase 3 International Study of Asthma & Allergy in Children (ISAAC) investigated the nationwide predominance of the presence of wheeze as being 7% in Indian children aged between 6 and 7 years and between 13 and 14 years. Certainly, more than 50% of the people had terrible asthma [4].

Asthma is an incurable disease. While some asthma cases are mortal, the majority are less serious and cause hardships in day-to-day life. It is very challenging for an asthmatic to maintain a healthy lifestyle since most people with asthma encounter symptoms of exercise-induced lung constriction while performing exercises, and engaging in exercise and sports can be very burdensome and may sometimes even be dangerous. The inability to participate in daily life activities can make them inactive, affecting their social well-being, which leads to stress, sleep deprivation, depression, etc. [5].

Spirometry and Peak Expiratory Flow are employed in diagnosing and monitoring the severity of asthma exacerbation. However, physician’s supervision is mandatory for both techniques, yet visiting a healthcare center is impracticable and tiresome for asthmatics. This inefficacy paved the way for advanced portable device technology [6]. These portable devices assist in diagnosing and continuously monitoring disease symptoms at earlier stages, reducing unnecessary hospitalization costs, allowing remote monitoring and helping maintain better patient-to-doctor ratios.

In recent times, chemical biomarkers like exhaled nitric oxide (eNO) and exhaled carbon monoxide (eCO) have been employed by healthcare professionals as another capable means of analyzing, evaluating and framing asthma treatment plans [2]. Carbon monoxide (CO) is a biomarker of oxidative stress induced by the stress protein heme-oxygenase [(HO)-1] and also due to inflammation [1]. Exhaled CO (eCO) is like exhaled NO, which has been assessed as a candidate breath biomarker of pathophysiological states elevated in asthmatics and decreases with steroid therapies. As mentioned in [1], NO is of airway origin, whereas CO is of alveoli origin. The main advantage of utilizing eCO as a breath biomarker is that it is observed in greater concentrations than nitric oxide (NO). This allows for less sophisticated and more reliable monitoring devices [7]. Many of the currently available asthma monitoring devices are either unaffordable, unreliable or not suitable for an active lifestyle [5].

When a person is experiencing an asthma attack, the body might be triggered by a stimuli, which results in bronchial inflammation. One method to detect these phenomena is by monitoring eCO concentrations before, during and after an acute attack [5]. Prior studies [8,9,10,11,12,13,14,15] have proven that CO levels increase during an asthma attack. This project utilizes the results found in these studies to develop an asthma monitor using a CO sensor.

Thus, the detection of eCO could be achieved via an effortless, non-invasive device for monitoring the acute exacerbation of asthma. This study aims to develop a programmed device that monitors eCO levels by using a CO sensor, which could possibly reduce acute asthma attacks and help improve the quality of life for asthmatics with varying lifestyles by reducing costs and extending portability, ultimately saving lives.

2. Literature Survey

The study conducted by Kiyoshi et al. (1997) shows an elevation of eCO in asthmatics that reduces with corticosteroid therapy, and it further indicates that the changes in the concentration of exhaled CO were remarkably associated with the number of eosinophil cells present in the sputum [8]. P. Paredi et al. (1999) further inspected the possibility that allergen challenge can raise exhaled carbon monoxide (eCO) levels as an indication of HO activation in 15 asthmatics and found out that the eCO increases in asthmatic reactions occur independently of the change in lung calibre [9]. The research conducted by M. Yamaya et al. (2001) has shown that eCO concentrations in people with unstable severe asthma were notably higher than in those with stable conditions. eCO concentrations in slight and non-extreme asthma patients did not differ appreciably from those patients who are controlled and do not have a smoking habit (p > 0.20). The research further states a remarkable proportionality between eCO concentration and forced expiratory volume [10].

Susumu Sato et al. (2003) estimated the optimal cutoff level of exhaled CO concentration to differentiate actual smokers from nonsmokers among 161 asthmatics and 170 COPD patients. The resulting analysis demonstrated that in asthmatics and COPD patients, the eCO levels were potentially affected by underlying lung inflammation, leading to the miscategorization of smoking status due to exhaled CO. This shows that CO levels increase in asthma [11]. The research of Ildiko Horvath et al. (2015) indicates that in asthmatic lungs, the induction of HO-1 may cause increasing eCO concentrations, suggesting that it is medically applicable as a diagnostic tool in terms of evaluating asthma [12]. The prevalence of the CO biomarker was studied by Yoichiro et al. (2016), and the result signifies that exhaled CO levels in children were considerably increased while encountering asthma exacerbation and were reduced after inhalation therapy with β2-agonist and SCG in those with intermittent asthma [13].

Amanda A. Pereira et al. (2018) speculated an increase in the levels of exhaled CO in asthmatic children, specifically in children with partially stable or unstable asthma, and the mean-adjusted eCO level was 0.56 ppm greater in asthmatic children, suggesting that exhaled CO may serve as an inexpensive biomarker for asthma control [14]. Yoichiro et al. (2020) further stated that in asthmatic infants and toddlers, asymptomatic asthma had exhaled CO levels of 2.0 (1.0–2.0) ppm, and those experiencing an asthma attack had exhaled CO levels of 2.0 (2.0–3.25) ppm (p < 0.0001) [15].

3. Materials and Methods

3.1. Overall Architecture of the Implemented System

As shown in Figure 1, the proposed system consists of two working units. One is the hardware architecture block and another is the software architecture block.

3.2. Hardware Architecture

(1)

Signal Acquisition Unit: Figure 2 shows a pictorial view of the signal acquisition unit. For accuracy and in order to avoid environmental CO, the sensor needs to be held close to the mouth. For this purpose, a face mask is used. The MQ-7 sensor contains 4 pins, namely A0, D0, Vcc and GND. The A0 pin is connected to AIN0 of PCF8591, Vcc and GND are connected to pin 2 and pin 6 of the Raspberry Pi. The CO is acquired through MQ-7 and transmitted to PCF8591. Figure 3 shows the connection of the MQ 7 sensor with the Raspberry Pi 3b+ and PCF8591.

(2)

Analog-to-Digital Conversion: The MQ-7 sensor gives output in the form of analog values. These analog data are given to PCF8591 ADC. PCF8591 works on I2C communication.

The obtained Analog value is converted to digital volts by

V_OUT = A_OUT × 5/256

(1)

A_OUT—analog value

V_OUT—digital value in (0–5 V).

3.3. Software Architecture

The concentration of CO in parts per million (ppm) was determined in accordance with the resistance ratio (RS/R0). RS is the estimated resistance altered when the sensing channel encounters gas, and R0 is the stable sensor resistance in clean air or in the absence of gas. Employing Ohm’s law and the sensor circuit diagram, the following can be obtained

RS = VC − RL/V_OUT − RL

(2)

• VC—voltage current (in this condition 5Volts from pi)
• V_OUT—output voltage (calculated analog/digital value)
• RL—load resistance (here the value is 10 K).
• R0 can then be computed with the below equation,
• R0 = RS/fresh air ratio value from sensitivity graph of MQ-7.

For the purpose of transforming the digital signal values to CO concentration in ppm, the MQ-7 datasheet is used. The correlation coefficients (A and B) are found by performing a power regression. The sensitivity characteristic graph of the MQ-7 shown in Figure 4 is loaded on Web Plot Digitizer software v4.6. The obtained data points are then loaded to R script code to perform a power regression.

Calculation of CO ppm concentration: To calculate the CO ppm concentration, two codes are written. The first section of the code is for calibrating the MQ-7 sensor in fresh air to acquire the RS and R0 values in atmospheric air, and the second section of code is for sensing CO in the testing environment. For the purpose of calibrating in clean air, the mean of 500 values obtained by running the first part of code was taken into account. The initial value begins with 0 and allows the addition of every read to each other till the given condition fails. Thus, from the mean value, the RS value in air and the R0 value in fresh air are calculated. This code runs a single time.

The next section of the code senses the presence of carbon monoxide, calculates the ratio and reads out the concentration of CO in ppm with respect to the change in voltage value. Figure 4 shows the flowchart for calculating CO concentration in real time.

Figure 5 shows the entire setup of the developed system with a face mask attached to the sensor.

4. Results and Discussion

The estimated values of correlation coefficients A and B using R-Script are found to be 83.69 and −1.63 approximately. The sensor has been calibrated based on the datasheet. The sensor was calibrated to detect 40 ppm CO in air. The response from the sensor in terms of ppm, as shown in Figure 6, is obtained for healthy (six numbers) and chronic asthma (three numbers) patients, and the level of CO was higher than the threshold for asthma patients, whereas for normal patient is below the threshold. This shows the sensitivity of the designed system against CO sources. The

Table 1. Exhaled CO concentration in healthy and individual with asthma.

S.No	Type of Subject	Exhaled CO Concentration Range in Ppm
1	Healthy	1.5 to 2.0
2	Asthma Patient	6.0 to 6.5

below shows the converted analog reading, as shown above in terms of ppm.

5. Conclusions and Future Scope

The designed CO sensor circuit shows high sensitivity to the CO sources. The target use of this system is to alert the patient to take precautions to avoid an asthma attack in the first place by monitoring their CO levels. Once the threshold level of CO that is dangerous to an asthma patient is found via clinical trials, the system can alert patients when their CO levels cross the threshold range. In addition, the design of the system can be enhanced further by reducing the size of the system as small as possible for handy use, travel ability and wearability. This project can be improved further by incorporating an IoT structure, giving the physician access to the data anywhere at any time to have an insight into the patient’s health condition. This project can be further enhanced to develop an accurate and fully functional device that justifies the necessity of every asthmatic who would make use of it.