After death human body is subject to the processes of autolysis and putrefaction. Notably, the changes in cadaver during decomposition complicate its forensic analysis and particularly the estimation of the post-mortem interval (PMI). The process and rate of decomposition is impacted by various intrinsic and extrinsic factors that vary across countries and regions. Studying the decomposition pattern in different regions in the world helps us to understand the process and improve the precision of the PMI estimation of decomposed bodies. With the aim to develop a taphonomic model of decomposition in the province of Barcelona (Catalonia, Spain), this study analyses the influence of several intrinsic and extrinsic factors in the pattern and rate of decomposition in this geographical area. Our statistical model concluded that the most significant factors affecting the decomposition pattern and rate are temperature and PMI. Nevertheless, there are other intrinsic factors such as cause, manner of death and underlying pathological conditions which also have an important role. Moreover, considering the various variables studied in this research, two predictive machine learning algorithms were developed as a probabilistic approach to estimate the PMI. Reliable classification results are obtained for three interval groups (1-2 days, 3-10 days, and > 10 days) and two interval groups (>1 week, < 1 week). Machine learning algorithm is a promising tool to gain objectivity in forensic PMI assessments. The results of this study could potentially assist further research in forensic taphonomy.
Keywords: Decomposition; Forensic anthropology; Forensic pathology; Machine learning; Post-mortem interval (PMI).
Copyright © 2022. Published by Elsevier B.V.