Machine Learning Model to Predict Chest Injury during Thorax Impact on Human Body Models

In recent years, virtual models have been extremely helpful in predicting potential injury risk to occupants in vehicle crashes. Virtual models offer detailed occupant anthropometry and closest possible bio-fidelity over existing test devices. This study focuses on the assessment of chest deflections in frontal thorax impacts using virtual human body models of a few anthropometries and transforming the assessment of injuries for a broader range of anthropometries (sections of the population). The study utilizes machine learning to enable injury assessment across a wide range of body types. A standard test scenario (Kroell load case) with a frontal blunt thoracic impact is considered for this study. Results from physical tests and simulations from various finite element human body models (HBMs) from literature are used to train supervised machine learning models. The combination of virtual simulation and machine learning reduces the reliance on physical prototypes and expands the reach of chest injury prediction for various populations. It provides a scalable, time-efficient approach to estimate injury risk and to protect a more diverse range of occupants. By integrating advanced simulation with data-driven modelling, this study offers a practical framework for evaluating chest injury risk in a wider population. The study highlights and demonstrates how predictive algorithms can enhance the versatility of simulation outcomes. Future work will explore expanding this approach to other impact types and body configurations, including vulnerable and underrepresented population groups.