A Deterministic Method to Calculate the AIS Trauma Score from a Finite Element Organ Trauma Model (OTM)

TL;DR

This paper introduces a deterministic Organ Trauma Model (OTM) that calculates the risk to life from organ injuries using a novel Peak Virtual Power metric, calibrated against AIS thresholds, and validated with real pedestrian accident data.

Contribution

The paper presents a new OTM that directly relates stress measures to AIS injury severity, improving injury prediction accuracy over traditional strain-based methods.

Findings

OTM accurately predicts head trauma severity in pedestrian accidents.

PVP correlates with AIS thresholds and injury severity.

Model underestimates some injuries due to post-impact hemorrhage.

Abstract

Traumatic injuries are measured using the Abbreviated Injury Scale (AIS), which is a risk to life scale. New human computer models use stresses and strains to evaluate whether serious or fatal injuries are reached, unfortunately these tensors bear no direct relation to AIS. This paper proposes to overcome this deficiency and suggests a unique Organ Trauma Model (OTM) able to calculate the risk to life based on the severity on any organ injury, focussing on real-life pedestrian accidents. The OTM uses a power method, named Peak Virtual Power (PVP), and calculates the risk to life of brain white and grey matters as a function of impact direction and impact speed. The OTM firstly calibrates PVP against the medical critical AIS threshold observed in each part of the head as a function of speed. This base PVP critical trauma function is then scaled and banded across all AIS levels using the confirmed property that AIS and the probability of death is statistically and numerically a cubic one. The OTM model has been tested against four real-life pedestrian accidents and proven to be able to predict pedestrian head trauma severity. In some cases, the method did however under-estimate the head trauma by 1 AIS level, because of post-impact haemorrhage which cannot be captured with the employed Lagrangian Finite Element (FE) solver. It is also shown that the location of the injury predictions using PVP coincide with the post mortem reports and are different to the predictions made using maximum principal strain.