Ask the Expert: Computer Modeling to Help Diagnose Head Injuries

Awareness of sports-related traumatic brain injuries will likely reach new heights with the Christmas Day release of Concussion, a feature film about brain damage in football players who suffer repeated concussions. Since 2000, UNC Professor Igor Szczyrba has been working on computer models to predict when a brain injury might occur. Szczyrba addresses his research in this Q&A:

Q: Can you describe the models you've helped develop for traumatic brain injuries?

Professor Szczyrba: Traumatic brain injury models that I have been developing, in collaboration with various researchers for the last 15 years, use powerful computers to predict dynamic changes within the brain tissue when a human head forcefully accelerates or decelerates. Specifically, we have introduced sophisticated partial differential equations to model brain dynamic, designed custom-made software to solve these equations, and conducted hundreds of numerical simulations mimicking realistic traumatic scenarios.

Q: How are the models being used for diagnosing head injuries?

Professor Szczyrba: The models describe the changes in physical properties within the brain — such as the strain of the brain matter during and some time after an accident. By comparing the models' predictions with the experimental critical values regarding when neurons or blood vessels in the brain can be damaged, the models can be used to estimate the potential severity of the brain trauma.

In a general situation, such as a skiing or a car accident, it's difficult to establish the specific type of head acceleration that took place. Nevertheless, the models' predictions based on estimating the type of head acceleration using the accident characteristics, such as the speed of a skier or the type of an obstacle hit with the head, can be a valuable aid to medical professionals in making a decision whether an MRI is necessary, for example. However, the problem is that we don't know exactly how the head was rotating.

Inserting acceleration sensors in football helmets, which has been recently done, allows for a much better utilization of the models' predictions even during a game. If we have sensors in the helmet our input is precise and the model's predictions are more accurate.

Q: What do the models suggest for making contact sports safer?

Professor Szczyrba: The particularly dangerous hits, according to our model, are hits to the head like a left hook in boxing and mixed martial arts fights followed by the right hook. Such a sequence of hits leads to a resonance effect that enhances the brain matter strain. The current football rules that penalize the 'sandwiching' are coherent with our predictions.

It is well known that rapid head rotational accelerations are especially dangerous since they can lead to a severe neuronal damage even several days after an accident. Thus, if such an acceleration is reported by sensors embedded in a football helmet, the sensors' reading might justify taking a football player off the field based not only on the existing general brain injury criteria but also on the model's predictions, even if the player passes the concussion test.

Q: What's the latest on your research?

Professor Szczyrba: One of the most difficult issues is to model the way the energy is transferred from a moving skull to the brain. It takes place in a nonhomogenous layer of tissue and fluid separating the skull from the brain. Since this layer is very thin and nonhomogenous, a very high resolution is required to model adequately the transfer of energy. Our recent numerical simulations show that our model is capable of handling this issue. Visit for research results.

Igor Szczyrba researches traumatic brain injuries with a focus on computer modeling in developing a diagnostic tool. He's professor of Mathematical Sciences at UNC.