Forensic AI tool aids brain injury investigations

Led by Oxford University, the study saw an AI model trained on real, anonymised police reports and forensic data from cases with traumatic brain injuries (TBIs). This resulted in a mechanics-informed machine learning framework to help police and forensic teams accurately predict TBI outcomes based on described assault scenarios. It achieved prediction accuracy exceeding 94 per cent for skull fracture, with 79 per cent for loss of consciousness and intracranial haemorrhage.

The work, published in Communications Engineering, featured input from Thames Valley Police, the National Crime Agency, Cardiff University, Lurtis Ltd., the John Radcliffe Hospital and other partner institutions.

“This research represents a significant step forward in forensic biomechanics,” said lead researcher Antoine Jérusalem, Professor of Mechanical Engineering at Oxford.

“By leveraging AI and physics-based simulations, we can provide law enforcement with an unprecedented tool to assess TBIs objectively.”

The framework uses a general computational mechanistic model of the head and neck, designed to simulate how different types of impacts - such as punches, slaps, or strikes against a flat surface - affect various regions. This provides a basic prediction of whether an impact is likely to cause tissue deformation or stress. An AI layer then incorporates this information with any additional relevant metadata, such as the victim’s age and height, before providing a prediction for a given injury.

“An 'Achilles heel' of forensic medicine is the assessment of whether a witnessed or inferred mechanism of injury, often the force, matches the observed injuries,” said forensics consultant and Cardiff University researcher, Dr Michael Jones.  

“With the application of machine learning, each additional case contributes to the overall understanding of the association between the mechanism of cause, primary injury, pathophysiology and outcome.”

According to the researchers, the AI tool is not intended to replace human forensic and clinical experts in investigations. Rather, the intention is to provide an objective estimate of the probability that a documented assault was the true cause of a reported injury.

"AI will never be able to identify without doubt the culprit who caused an injury,” said Professor Jérusalem.  

“All it can do is tell you whether the information provided to it is correlated with a certain outcome. Since the quality of the output depends on the quality of the information fed into the model, having detailed witness statements is still crucial.”