Institute of Fundamental Technological Research

Conventional biometric identification methods relying on Personally Identifiable Information (PII) pose significant challenges concerning privacy and security. Volatile organic compounds (VOCs) in exhaled breath are unique to individuals and can serve as biomarkers for various diseases, making them a promising tool for both bioidentification and clinical diagnostics. This research investigates the feasibility of utilizing VOCs in exhaled breath as biometric identifiers, employing machine learning algorithms for analysis. Additionally, the research investigates the use of VOCs as non-invasive indicators of health status, specifically for estimating BMI and gender. The study involved 94 participants with an average age of 67 years and an average BMI of 28 kg/m². Exhaled breath samples were collected using a portable electronic nose (e-nose) device, which analyzed the VOCs. Machine learning algorithms were applied to the data to assess the feasibility of identifying individuals and estimating BMI and gender based on VOC patterns. The results indicated that VOC patterns could reliably estimate BMI and gender, and potentially distinguish between individuals, suggesting VOCs as a viable tool for bioidentification. The application of machine learning to VOC data showed promise in non-invasive identification and diagnostics. VOCs in exhaled breath offer a novel, non-invasive method for biometric identification and health monitoring. This approach could overcome the limitations of traditional PII, providing a new avenue for personalized medicine. Further research is needed to enhance the accuracy and applicability of this method in diverse populations.

Exhaled breath pattern, Breath analysis, Volatile organic compounds, Algorithms, Artificial intelligence, Biometrics