The Development of a Tympanic Membrane Model and Probabilistic Dose-Response Risk Assessment of Rupture Because of Blast

Abstract

ABSTRACT Introduction There is no dose-response model available for the assessment of the risk of tympanic membrane rupture (TMR), commonly known as eardrum rupture, from exposures to blast from nonlethal flashbangs, which can occur concurrently with temporary threshold shift. Therefore, the objective of this work was to develop a fast-running, lumped parameter model of the tympanic membrane (TM) with probabilistic dose-dependent prediction of injury risk. Materials and Methods The lumped parameter model was first benchmarked with a finite element model of the middle ear. To develop the dose-response curves, TMR data from a historic cadaver study were utilized. From these data, the binary probability response was constructed and logistic regression was applied to generate the respective dose-response curves at moderate and severe eardrum rupture severity. Results Hosmer-Lemeshow statistical and receiver operation characteristic analyses showed that maximum stored TM energy was the overall best dose metric or injury correlate when compared with total work and peak TM pressure. Conclusions Dose-response curves are needed for probabilistic risk assessments of unintended effects like TMR. For increased functionality, the lumped parameter model was packaged as a software library that predicts eardrum rupture for a given blast loading condition.