Ability Impact Score (AIS) in Trauma Modeling

• Ability Impact Score (AIS) is a deterministic method that maps finite element outputs to clinical injury ratings using the Peak Virtual Power indicator.
• The method calibrates tissue-specific injury thresholds through explicit FE simulations and empirical data to assign precise AIS levels.
• It applies a cubic scaling law to relate injury severity to fatality probability, validated by reconstructions of real-world pedestrian impacts.

The Ability Impact Score (AIS), as operationalized in the Organ Trauma Model (OTM), is a deterministic biomechanical method for assigning clinical risk-to-life scores based on quantitative finite element (FE) outputs. Unlike traditional approaches that utilize qualitative or indirect metrics, the OTM directly links stress and strain-rate tensors from explicit FE simulations to specific AIS levels by means of the Peak Virtual Power (PVP) indicator. This methodology provides a physical and statistical mapping between injury mechanics and clinical injury ratings, with particular focus on predicting brain trauma in real-world pedestrian impacts (Bastien et al., 2019).

1. Theoretical Framework: Peak Virtual Power and Injury Quantification

The OTM establishes injury prediction via the Peak Virtual Power (PVP), which is derived from the local “virtual power” per unit volume: $\pi(t) = \sigma_{ij}(t)\,\dot\varepsilon_{ij}(t)$ where $\sigma_{ij}$ denotes the von Mises equivalent stress and $\dot\varepsilon_{ij}$ the elastic strain-rate tensor. PVP is the temporal maximum of this quantity: $$PVP = \max_{t\in[0,T]} \bigl[\sigma_{ij}(t)\,\dot\varepsilon_{ij}(t)\bigr]$$ This relation is grounded both in the Clausius-Duhem inequality and in kinematic/energetic arguments, with PVP scaling as $v^2$ for direct impacts and, for pedestrian “ride-down” scenarios, as %%%%3%%%%. The underlying hypothesis is that irreversible mechanical work effectively equates to clinically relevant injury.

2. Calibration: Mapping Mechanical Metrics to Clinical Severity

The translation from PVP to clinical AIS leverages organ- and direction-specific thresholds. Using explicit human-head FE models (e.g., THUMS AM50 with viscoelastic tissue properties), the authors determine the PVP value at which known injury correlates occur:

• Grey-matter contusion: $\sim30\%$ plastic strain $\rightarrow$ AIS 3
• White-matter diffuse-axonal-injury (DAI): $\sim21\%$ plastic strain $\rightarrow$ AIS 4

For each anatomical region and impact direction, $PVP_4(v)$ is fit to the form

$PVP_4(v) = \alpha_4\,v^3$

with $\alpha_4$ calibrated from simulation output. This procedure generates a base set of direction- and tissue-specific trauma risk functions.

3. Cubic Scaling Law: Probability of Death and Severity Indexing

Epidemiological studies demonstrate that the probability of death, $P_{\mathrm{death}}(x)$, associated with each AIS level $x$, follows a cubic relationship: $$P_{\mathrm{death}}(x) \approx Cx^3, \quad C \approx 0.4461$$ Consequently, the thresholds for lower and higher AIS levels are derived by scaling the AIS-4 threshold: $$PVP_k(v) = \left(\frac{k}{4}\right)^3 PVP_4(v), \quad k = 1,\dots,5$$ This cubic scaling partitions the range of injury metrics into statistically meaningful bands, each bounded by an empirical ±20% tolerance to accommodate biological and measurement variability.

4. Finite Element Simulation and Injury Assignment Workflow

The OTM workflow for AIS assignment utilizes detailed FE simulations with explicit modeling of tissue mechanical response and realistic loading conditions:

• Model: THUMS AM50, representing white and grey matter with viscoelastic incompressible elements
• Impactor: 200g rigid cylinder, applied to frontal, lateral, or occipital regions
• Boundary: Neck base free, head unconstrained, gravity off
• Loading: Impact speeds varied from 2 to 17 m/s

At each simulation timestep, PVP is computed per element; the maximal PVP among elements separately for grey and white matter is extracted: $$PVP_{\text{grey}} = \max_{e \in \text{grey}} PVP_e,\quad PVP_{\text{white}} = \max_{e \in \text{white}} PVP_e$$ Assignment to an AIS level $k$ proceeds by identifying the interval between PVP band thresholds: $PVP_{k}(v) \leq PVP_{\text{region}} < PVP_{k+1}(v)$ Accounting for the calibrated $\alpha_4$ for the specific tissue and impact direction, and including ±20% uncertainty corridors, a deterministic AIS can thus be assigned for each region. The Maximal AIS (MAIS) across grey and white matter serves as the injury severity for the organ.

5. Empirical Validation with Real-World Pedestrian Injury Cases

Validation involved reconstruction of four actual pedestrian accidents, utilizing known vehicle types, impact speeds, and independently determined post-mortem (PM) AIS assessments. The OTM’s predictions typically matched clinical outcomes within ±1 AIS level and located the correct lobe in over 75% of cases. Underestimation of trauma severity was attributed to the omission of post-impact hemorrhage and swelling—biomechanical processes not captured by Lagrangian solid solvers. In individual cases, the OTM could provide plausible indications of subclinical trauma zones not observed on PM, suggesting possible utility in detecting otherwise unrecorded injury patterns (Bastien et al., 2019).

6. Model Limitations and Extensibility

Several principal limitations constrain the OTM’s deterministic mapping:

• Lagrangian mesh FE methods cannot simulate bleeding or swelling; thus, injuries resulting from post-impact fluid dynamics (e.g., hemorrhage) are not predicted
• Bone fracture prediction is decoupled from tissue PVP and is not currently integrated
• PVP is sensitive to tissue material properties such as Young’s modulus, density, and viscosity; variation due to age, disease, or population requires re-calibration
• Extension to other organs is direct where failure thresholds and PVP–AIS calibrations exist

Future developments could incorporate Arbitrary Lagrangian-Eulerian (ALE) or Smoothed Particle Hydrodynamics (SPH) approaches for capturing fluid-structure injury mechanisms.

7. Significance for Clinical and Computational Injury Biomechanics

The OTM’s deterministic method provides a direct, quantitative link between FE-derived mechanical metrics and risk-to-life injury scales. By leveraging the cubic statistical relation between severity and fatality, this approach anchors clinical scoring in explicit physical processes and facilitates rigorous, repeatable injury predictions. A plausible implication is that this framework may enable improved forensic analysis, safety system validation, and organ-specific trauma assessment. However, integration with multiphysics solvers is necessary to capture the full spectrum of trauma mechanics, particularly those involving post-impact tissue response (Bastien et al., 2019).