A non-invasive electronic nose design based on virtual gas sensor array

1 Introduction

Lung cancer is one of the most common malignant tumors in the world, and its incidence rate has been increasing significantly in most countries. It is of great significance to design an electronic nose that can detect the breath of lung cancer patients. At present, the pathological basis for the diagnosis of lung cancer breath is not very clear, and no method has been developed internationally that can be used for accurate diagnosis of early lung cancer. In 1999, Michael Phillips and others used a large gas chromatograph to study the relationship between breath odor and lung cancer, and for the first time pointed out that 22 VOCs gases can be regarded as characteristic gases of lung cancer.

At present, the pathological explanation of lung cancer breath detection is: abnormal human body conditions can lead to excessive production of reactive oxidants in the body, including oxygen free radicals and hydrogen peroxide, which disrupts the balance between strong oxidants and antioxidants, leading to a potential harm - oxidative stress, which is the main cause of cell damage. In this case, lipid peroxidation of polyunsaturated fatty acids in the cell membrane will produce alkanes and aldehydes, which are excreted from the body through breathing. However, the specific characteristic gas has not yet been explained pathologically internationally.

This paper proposes a novel non-invasive electronic nose for lung cancer detection based on a virtual gas sensor array. The electronic nose contains a pretreatment device consisting of solid phase microextraction (SPME) and a capillary column (gas chromatography, GC) to achieve the concentration, adsorption, desorption and separation of VOCs in the patient's respiratory gas, and then a SAW sensor with a differential structure coated with a polyisobutylene film is used to quantitatively detect the separated organic gas components (Figure 1).

2 Design of the respiratory detection electronic nose system

The concentration of most VOCs contained in the respiratory gas of lung cancer patients is very low (10-9 ~ 10-12 mol / L). The experiment uses inert Tedlar gas collection bags and SPME to collect and pre-enrich the respiratory gas, and uses thin film technology to improve the sensitivity of SAW sensors. Finally, an improved artificial neural network algorithm combined with image technology is used to identify lung cancer patients.

The SAW sensor is made of widely used ST-cut LiNbO3 crystal. The center frequency of the sensor is 51.92 MHz, the bandwidth is 2.131 MHz, the good Q value is 24.36, and the attenuation is 7.85 dB. One of the sensors in a pair is coated with PIB film as the working sensor, while the other sensor without film coating is used as the reference sensor.

The experimental GC-SAW electronic nose system combined the concept of virtual SAW sensor array and image recognition method to detect the exhaled gas of lung cancer patients, normal people and chronic bronchitis patients, and identified 11 VOCs as characteristic gases for lung cancer, namely styrene, decane, isoprene, benzene, undecane, 1-hexene, hexanal, propyl benzene, 1,2,4-trimethyl benzene, heptanal and methyl cyclopentane.

Figure 2 shows the spectrum response of a SAW sensor when detecting the breath of a lung cancer patient. Its sensitivity to styrene is 3.24 ng/Hz.

3. Detection of VOCs in lung cancer cell metabolism

For the 11 selected lung cancer characteristic VOCs gases, the experiment confirmed that these VOCs gases are highly correlated with lung cancer. The results show that the electronic nose can well detect the types and contents of these VOCs gases in the respiratory gas of lung cancer patients, and the diagnosis rate of lung cancer patients is above 75%. In order to more accurately determine the characteristic gases of lung cancer patients, a comparative experiment at the cellular level was conducted with reference to the pathological interpretation, and the electronic nose was used to detect VOCs in the lung cancer cell culture fluid. Lung cancer tissue and normal lung tissue specimens of the same body were cultured separately. The culture fluid after cell culture was obtained, SPME extraction was performed, and GC was used for analysis. As shown in Figure 3, the spectra of lung cancer cell culture fluid and normal lung cell culture fluid obtained by GC are compared with normal cells. It is determined that there are specific gases in the metabolic gases of lung cancer cells. The analysis results of one lung cancer cell and the normal lung cell culture fluid of the same body are shown in Figure 4. Comparing the peak shape diagrams of the two, it can be found that although there are differences in peak height, the peak positions are almost all the same. The similar analysis results make people suspect that the normal lung tissue adjacent to the cancer has actually undergone canceration. The experiment used electron microscope to observe the cultured cells and found that the normal lung cell images judged by doctors had cell overlap, which is not the case with normal cells. It can be seen that the cancerous lung tissue that cannot be correctly judged by the naked eye can also be identified through our experimental method, so the results of this experiment are of great significance for the early screening of lung cancer by breathing gas.

4 Conclusion

This paper proposes an electronic nose with the advantages of high sensitivity, low cost and easy operation. Using this electronic nose system, the pathological aspects of the respiratory gases of lung cancer patients, chronic bronchitis patients, and healthy people were studied. At the same time, the electronic nose system was used to study the volatile gases in the metabolites of lung cancer cells, and preliminary evidence for early lung cancer detection was obtained.

As for the respiratory gas diagnosis of lung cancer, the medical physiology is still not very clear, and the technology that can be applied to the accurate clinical diagnosis of early lung cancer does not exist. The research goal is to use this new electronic nose system to diagnose lung cancer patients in the near future or to diagnose or screen early lung cancer patients as much as possible.