Multivariate optoacoustic sensor for longitudinal diabetes monitoring

MOSAIC will radically miniaturize non-invasive optoacoustic technology and make it portable by coupling optoacoustic sensing to novel explainable artificial Intelligence (xAI) to monitor a major health threat in the 21st century: Diabetes Mellitus. MOSAIC is based on recent advances on optoelectronic miniaturization and uses the skin as a window to dermal microvasculature and resolves a large number of biomarkers associated with diabetes. The technology is offered as a fast sensor that is inexpensive and reliable, entirely non-invasive and portable and supported by elaborate quality control monitoring so that it can be employed for long-term patient monitoring. Based on proof-of-concept human measurements, xAI will further independently characterize the predictive power of each biomarker and other clinical measurements to improve accuracy for continuous grading of disease status. Impact: With 537 million diabetics and a large at-risk population, MOSAIC potentially relates to 2B people and aims to reach more individuals than is practical to screen in medical facilities. It offers early diabetes detection for early interventions, as well as treatment monitoring, improving life expectancy, quality of life and reducing healthcare costs. Innovation: The sensor addresses 5 barriers to successful adoption: it is non-invasive, senses highly multiplexed biomarkers, it can be easily integrated into daily routine, and uses fully automated data-analytics. MOSAIC synergistically integrates 3 teams on sensor developments, data analytics and xAI, one preclinical and one clinical team for sensor validation in mice and humans, respectively, and an SME for exploitation.