Model to Predict Probability of Significant Skin Burn Injury From a Directed-Energy Source

Abstract

ABSTRACT Introduction Millimeter wave directed energy in the frequency regime of 94-95 GHz has potential for use in numerous military applications including crowd control and area denial. Although proven to be very safe, millimeter wave energy has the potential, because of accidental over exposure, to produce significant injuries. Currently, the Dynamic Thermal Model (DTM), developed by Beason and colleagues, is used to calculate the temperature profile in skin undergoing (millimeter wave) heating, using an all-or-nothing threshold of injury. Risk of significant injury (RSI) is determined by product of the probability of an injury outcome on a region of the body times the probability of that the injury will occur. Thus, a threshold injury determination may over predict burn probability and fail safety requirements. This work augments the DTM, replacing the current threshold value of injury with a probabilistic risk of injury to better quantify the risk of significant injury. Materials and Methods In this study, continuous probabilistic dose–response models using logistic regression analysis have been developed to account for mild second-degree, deep second-degree, and third-degree burn injuries based on a historic experimental burn dataset. Statistical analysis methods such as Hosmer–Lemeshow statistics, McFadden’s pseudo R2 and receiver operator characteristic were used to validate the models against an independent experimental burn dataset. Results Comparison of logistic models fit using damage coefficients from the literature showed that Mehta and Wong provided the best fits historic burn data, which was corroborated by the McFadden pseudo R2 statistic for mild second-degree, deep second-degree, and third-degree severity. Conclusion The dose–response models developed in this study are shown to be an excellent predictor of burn injury for each severity. The DTM was repackaged with the probabilistic burn models to more accurately determine the risk of significant burn injury.