Contactless metabolism estimation of small animals using high-frequency millimeter-wave radar

Abstract

Animals flexibly adapt to internal and external environmental changes by utilizing energy produced from oxygen as fuel. By non-invasively monitoring an animals oxygen consumption, it becomes possible to understand an individuals metabolic state. Calorimeters are known for directly measuring oxygen consumption but come with the issue of high initial costs. Despite the development of non-invasive techniques for measuring vital signs--including respiration rate (RR), heart rate (HR), and body temperature (Tb)--as indicators of metabolism, conventional methods encounter difficulties in estimating oxygen consumption rate (VO2). In this study, we developed a system that estimates the oxygen consumption of small animals using signals obtained from millimeter wave (mm-wave) radar technology processed through machine learning. By identifying frequency bands within the mm-wave signals contributing to VO2, our system is capable of predicting oxygen consumption several minutes in advance. Our system enables contactless, low-cost and multiplexed measurements of oxygen consumption, presenting a significant advancement in the field.