Modeling human decomposition: a Bayesian approach

Abstract

Environmental and individualistic variables affect the rate of human decomposition in complex ways. These effects complicate the estimation of the postmortem interval (PMI) based on observed decomposition characteristics. In this work, we develop a generative probabilistic model for decomposing human remains based on PMI and a wide range of environmental and individualistic variables. This model explicitly represents the effect of each variable, including PMI, on the appearance of each decomposition characteristic, allowing for direct interpretation of model effects and enabling the use of the model for PMI inference and optimal experimental design. In addition, the probabilistic nature of the model allows for the integration of expert knowledge in the form of prior distributions. We fit this model to a diverse set of 2,529 cases from the GeoFOR dataset. We demonstrate that the model accurately predicts 24 decomposition characteristics with an ROC AUC score of 0.85. Using Bayesian inference techniques, we invert the decomposition model to predict PMI as a function of the observed decomposition characteristics and environmental and individualistic variables, producing an R-squared measure of 71%. Finally, we demonstrate how to use the fitted model to design future experiments that maximize the expected amount of new information about the mechanisms of decomposition using the Expected Information Gain formalism.

Figures (9)

• Figure 1: This graphic represents our generative model for decomposition based on a range of environmental and individualistic variables. Solid-line arrows represent the effect of one variable on the occurrence of a particular decomposition characteristic. A small subset of effects are allowed based on the stage of decomposition associated with each characteristic. The dashed-line arrows represent the use of the model to infer postmortem interval via Bayesian inference.
• Figure 2: Histogram of PMI values in the geoFOR dataset along with log-transformed PMI values discussed in Section \ref{['sec:likelihood']}.
• Figure 3: Performance of the "Empty" model. Figure \ref{['fig:r2_plot']}, left, shows the Predicted vs. True PMI values for the "empty" model. The markers represent the mean PMI predictions, and the error bars represent the 90% prediction interval from performing PMI posterior inference as described in Section \ref{['sec:pmi-inference']}. Figure \ref{['fig:calibration']}, right, shows the percentage of cases falling within the prediction interval as a function of the prediction interval size. A perfectly calibrated model would fall along the dashed black line.
• Figure 4: $\beta_{d0}$ posterior distributions for each decomposition characteristic. The distributions are ordered by the median value of $\beta_{d0}$.
• Figure 5: Probability distributions that show the relationship between the size of a body and desiccation. Negative samples correspond to desiccation being less likely in relation to normal body size. Positive samples correspond to desiccation being more likely.