Establishing models for postmortem interval estimation based on measuring surface temperature of corpses and ambient temperature by infrared thermography technology

Abstract

Body temperature is an important indicator for forensic scientists to estimate the early post-mortem interval (PMI). Traditionally, forensic practitioners have mostly used methods such as measuring the internal temperatures of the body, including rectal temperature and liver temperature, to infer the PMI. However, these methods rely on the experience of the operator, suffer from manual measurement errors, and cannot fully consider the interference of environmental factors, resulting in unstable accuracy of the inference results. Infrared thermography technology can use optoelectronic technology to dynamically and monitor the surface temperature of objects and ambient temperature. It has the advantages of non-contact, convenience, and high efficiency. This project utilized thermography technology to monitor the surface temperature of SD rat corpses and the real-time ambient temperature under five average environment temperature conditions of 25 °C, 20 °C, 15 °C, 10 °C, and 5 °C, aiming to explore the correlations among the changes in body surface temperature, ambient temperature changes, and PMI. By comparing the changes in body surface temperature of rabbits, SD rats, and C57 mice, the project investigated the relationship between body weight and the rate of body surface temperature decrease. Then, multivariate correlation analysis, multiple regression analysis, and machine learning analysis methods were employed to construct and evolve models for PMI estimation. Finally, the accuracy of the model was verified using the single-blind method. The results showed that using infrared thermography to detect body surface temperature and ambient temperature can more conveniently and accurately infer the PMI, demonstrating high application and transformation value.