Breath Odor Biometrics by Artificial Olfaction Sensor Array and Machine Learning

Exhaled breath odor contains thousand chemical compounds and serves as abundant information source to characterize the human. In this research, we investigate a breath odor biometrics by artificial olfaction sensor array and machine learning. Starting from the feasibility study of machine-learning based breathprint analysis by the molecular fragments data in a gas chromatography-mass spectrometry, we constructed the chemiresistive sensor array and examined the breath odor sensing for the biometrics applications. In the breath odor sensing, the larger number of sensors and sensing features were utilized, the more complex breath odors could be discriminated. On the basis of these finding, we demonstrated the primary studies of breath odor sensing based individual authentication and physiological data collection/monitoring. Our results provide an important foundation towards complex odor sensing in smart farming, food quality control, healthcare and telemedicine.