“We’re dealing with a very soft material inside a very hard structure. A hit to the head causes pressure waves and shear waves to move through the brain, like a rap on a bowl of Jell-O,” Ramesh says.

“These stresses and strains can cause injuries that cascade through the tissues and create larger damage and cognitive problems,” Daphalapurkar adds.

Here the analogy with Jell-O ends. The brain is not a monolithic structure in which all the physical features and tissues behave identically. It is instead a complex mesh of different structures—neurons and axons, meninges and synapses, dendrites and glia—each responding in its own way to physical forces. In technical terms, the brain is anisotropic. Like a piece of wood, it is stronger and more resilient in certain directions than in others. The engineer’s job is to anticipate and account for these varying behaviors in the computer models. This is no easy task.

The team has focused its attention on the axons, the long, thin threads of cells that carry electrical impulses from neuron to nearby neuron. Filled with fatty lipids, axons appear white to the naked eye and make up the majority of the brain’s inner structure, known appropriately as the white matter.

Surprisingly, however, while strain on the axons has been shown to cause functional damage to neurons, it has never been used to calculate injury in computational models of traumatic brain injury.

The sloshing of the brain due to head trauma stretches the axons, causing damage and brain malfunction,” Ramesh says. “We’re the first to incorporate axonal damage in computer models of head trauma.”

The team is trying to decipher why some physical movements in the brain cause severe damage and others do not. More specifically, they are trying to determine where the fine line between minor and serious injury is drawn.

“In earlier computer models, brain tissue was often treated as this homogeneous material, ignoring the various internal structures that behave very differently from one another,” Ramesh says.

In addition to treating the brain as an anisotropic material, Ramesh has gone deeper still by adding rotational acceleration to his algorithms. There are two types of acceleration: linear, which moves in a straight line, and rotational, which spins around an axis like a figure skater. The automobile industry uses a set of metrics known as the head injury criterion to predict brain injury, but it ignores rotational acceleration and therefore falls short of the nuance Ramesh is after.

“In traumatic brain injury, anatomy really matters. Our models allow us to see in real time how brain deformations vary from one person to another, and from force to force,” Daphalapurkar explains.