BioBreath Matrix Advancing Respiratory Biomarker AnalysisThrough Innovative Matrix Technology

Abstract

Non-communicable diseases (NCDs) are responsible for a substantial proportion of global mortality, highlighting the need for accessible and continuous health monitoring technologies. Human breath contains numerous volatile organic compounds (VOCs) that reflect physiological and metabolic conditions. This study aimed to develop BioBreath Matrix, a non-invasive breath analysis platform capable of integrating multiple VOC biomarkers with artificial intelligence to support comprehensive health assessment.

The proposed system utilizes a multivariate sensor array consisting of MQ-3, MQ-135, and MQ-138 sensors, together with a DHT22 sensor for environmental compensation. Four key breath biomarkers—acetone, nitric oxide, isoprene, and ammonia—were selected based on their associations with metabolic, respiratory, cardiovascular, and renal health conditions. Sensor signals were processed through feature extraction techniques and analyzed using Random Forest and Support Vector Machine (SVM) algorithms. The analytical outputs were subsequently integrated into a Composite Health Score and visualized through a real-time health monitoring dashboard.

The developed prototype successfully demonstrated the acquisition, processing, and interpretation of multidimensional breath biomarker data within a unified framework. The system enabled simultaneous biomarker monitoring, AI-assisted health classification, and real-time visualization of health indicators through an interactive dashboard.

In conclusion, BioBreath Matrix demonstrates the feasibility of combining breath biomarker sensing and machine learning for non-invasive health assessment. The platform provides a foundation for future clinical validation and may contribute to the development of personalized, real-time, and preventive healthcare technologies.