oa A Self-Calibrated Gas Sensing System for Breath Analysis

A self-calibrated gas sensing system for breath analysis is introduced. The system is composed of an array of high sensitivity gas sensor combined with temperature, humidity and flow sensor for calibration. The system is able to diagnose common diseases and help people stop them at early stages.

1. Introduction

Breath analysis is a new method for point-of-care health monitoring and diagnosis. Compared to traditional diagnostic techniques like blood test, urine test, biopsy, endoscopy and imaging, it has at least two advantages: complete non-invasiveness and limitless repeatability. Exhaled breath contains thousands of volatile compounds that can serve as biomarkers of metabolism processes [1]. By analyzing the pattern of the exhaled compounds, various diseases, including airway inflammation, renal failure and lung cancer, can be detected at early stages [1].

Gas chromatography (GC), mass spectrometry (MS) or the combination of the two (GC-MS) are the most frequently adopted methods in breath analysis given their high reliability and sensitivity. Nevertheless, these techniques require bulky equipment not suitable for point-of-care application. Gas sensor array based electronic nose (eNose) is a promising alternative to GC and MS. It has the advantage of portability and low cost which makes it suitable for home-based healthcare solutions. However, the response of gas sensor array also varies with different temperature, humidity and flow rate. To address this problem, in-system calibration techniques needs to be adopted.

2. Implementation

2.1 Diagram of Self-calibrated Gas Sensing System

As shown in Fig. 1, the proposed gas sensing system is composed of three major parts, which are the multi-sensor stage, the analog front end (AFE) and the digital processing unit. The sensor part includes a gas sensor array and three other sensors for calibration, which are flow, temperature and humidity sensor, respectively. The AFE part consists of two individual blocks. One is the frequency shift detection circuit for Film Bulk Acoustic Resonator (FBAR) gas sensor array and the other is a voltage detection circuit shared by the flow, temperature and humidity sensors. The digital processing unit takes charge of the sensor data sampling, processing and transmission.

Figure 1: Diagram of Self-calibrated Gas Sensing System

2.2 FBAR Gas Sensor Array and AFE

2.2.1 FBAR Gas Sensor Array

The gas sensor array is the core of the system and directly determines the overall performance of the eNose. Here, FBAR gas sensor is chosen as the element of the array to achieve high sensitivity. The selectivity of the FBAR-based sensor is drastically affected by the sensing material. By depositing different sensing materials or changing the properties of the sensing material on each individual sensor in the array, a unique pattern response to a specific gas mixtures can be generated by the array. The gas composition and concentration of the mixture can be acquired from the sensor array. For breath analysis, a set of specific sensing materials targeted at sensing biomarkers of common diseases are selected. For example, palladium is sensitive to hydrogen which is related to indigestion [1].

2.2.2 Analog Front End of FBAR

The response of FBAR gas sensor is illustrated in a frequency shift caused by mass loading effect when exposed to target gases. Two different methods can be used to detect the output signal. One is to count the high frequency signal directly with a proper ratio pre-scaling. The other method is to count the intermediate frequency (IF) signal after moving the original signal to the IF band. Because the design targets at mobile application, power consumption of the system should be one of the major considerations when designing the system. As direct counting at high frequency consumes a great deal of power, the second scheme is preferred.

Figure 3 shows a brief diagram of the frequency shift detection circuit for the FBAR gas sensor. Through RF multiplexer, each FBAR is alternatively connected to the first stage of the circuit to build an oscillator. Another reference FBAR without sensing material is used to build the reference oscillator. By mixing the sinusoidal signal of the two oscillators and low pass filtering the mixed signal, an intermediate frequency (IF) signal contains the frequency shift information of the sensor is acquired. At last, the frequency of the IF signal is read out using a simple counter.

Figure 3: Frequency shift detection circuit for FBAR gas sensor

2.3 Sensors Calibration and AFE

The response of the gas sensor array can drift due to the variation of flow rate, temperature and humidity. Calibration of the drift can be achieved by subtracting the influence of those variations. To obtain the calibration information, flow, temperature and humidity sensors need to be implemented on the same die. A BJT temperature sensor with ± 0.4 accuracy, a capacitive humidity sensor with ± 4% RH accuracy and a thermal mass flow meter was adopted. With signal conditioning interfaces, all the three sensors share the same 16 bit, 90 Hz Σ-Δ ADC.

2.4 Digital Processing Unit

The digital processing unit is designed to control the sampling rate and transmits the sensor data off-chip through a I2C transceiver. The processor also performs data fusion as multiple sensing data are obtained in the proposed system. Data processing and wireless communication are accomplished by the host controller.

3. Conclusion

A self-calibrated gas sensing system with high sensitivity and good reliability is introduced to achieve diagnosis of multiple common diseases like diabetes, airway inflammation and indigestion. The system is power efficient and small enough to be integrated in mobile devices like smart phone and tablets. The system features a self-calibration methodology through the use of multi-sensing platform namely: temperature, humidity and flow.

References

[1] De Lacy Costello, B., et al. “A review of the volatiles from the healthy human body.” J Breath Res 8 (2014): 014001.

[2] Wilson, Alphus D., and Manuela Baietto. “Advances in electronic-nose technologies developed for biomedical applications.” Sensors 11.1 (2011): 1105-1176.

[3] Benetti, M., et al. “Microbalance chemical sensor based on thin-film bulk acoustic wave resonators.” Applied Physics Letters 87.17 (2005): 173504.