Progress Reports

Progress Report 1/14

Problem Statement

Concussions are a chronic injury that are widely not understood. While many studies have looked into impacts, the main pitfall of these studies is around understanding localized internal effects of impact.

Overall Goal

Probable: To create a finite element analysis (FEA) of the brain that shows internal motion, deformation, and trauma that results from impact.
Possible: Create custom models from MRI data to analyze personalized brain impact
Reach: Take real-time impact data from athletes and determine whether a player sustained a concussion using player personalized model tied to an app used by trainers

Progress

Before starting any work in actual software, our group gleaned through many papers concerning finite element models of the brain and how they went about their model: segmented vs. non-segmented model, software choice, etc.
We have keyed in on a specific finite element model software, ANSYS, which Vanderbilt provides for free. It is possible that we would need to buy an extra package within ANSYS to supplement our work. In practicing and familiarizing ourselves with ANSYS, we built a very simple FEA model of human head with layered spheres representing the material properties of bone, cerebrospinal fluid, and brain tissue. The model shows an object impacting a cylinder representing the human head, simulating the events leading up to a concussion or brain trauma. Model only shows shear stress on the layers during impact at this point in the process. In addition, we have found a way to integrate CT scans and create models of the brain from that, but one of the neurosurgeons in our group will be giving us high-resolution MRI scans of the brain to use that would be much more accurate.

Action Items/Goals

Investigate biomechanical properties of the brain, skull, cerebrospinal fluid, etc. in order to determine useful parameters for the model
Achieved by searching the literature for well-supported data for each of the parameters

Our current preliminary model only depicts the brain, skull, and cerebrospinal fluid so we need to investigate and create a way to add a football helmet to the current ANSYS model
Achieved by searching the literature for existing models and for research on the specifications of helmets so that we can use those specifications in constructing a useful model

Progress Report 1/28

Since our last senior design meeting with our advisers was on 1/14, we have still maintained the same goals that we had in Progress Report 1. The advisers are going to send us high-resolution MRI data that we can use for our initial ANSYS finite element model. Until then, we cannot make any more additions to our model. In the meantime, we are gleaning through research studies and attempting to contact the investigators so we can examine their models to further enhance our model and understand what the state of the technology is.

Progress Report 2/13

In the last two weeks, our group has found a suitable and malleable finite element model courtesy of the Global Human Body Models Consortium (GHBMC). Since we have recently acquired the model, it is important to first familiarize ourselves with navigating through the model and changing parameters. In the coming days, we will be having a Skype call with an author of the model to provide us with an orientation and introduction to it.
Furthermore, we have narrowed down our design objectives to be as follows: in our model, we need to change the biomechanical properties of brain tissue so that it is modeled as a non-linear rather than linear material (brain tissue is generally considered to be visco/hyperelastic). To accomplish this, group members have been peering through scientific literature to come up with useful biomechanical data of brain tissue such as: Young’s modulus, rigidity modulus, stress-strain curves, viscoelastic curves, toughness, and ductility. We have also discovered several current models of viscoelastic brain deformation that incorporate numerous variables in order to determine the local and directional strain on axonal tissue. These models will be able to provide us with the numerical calculations necessary for determining location-specific brain deformation. In addition, we want to make our model closely representative of the human brain in that we want to build in a function to our model which can account for the additive effects of multiple hits over time. To accomplish this, we need to research the Arrhenius Damage Integral and further consult with Dr. Michael Miga, as he a wealth of knowledge in this field of finite element analysis.
As for dealing with the model itself, we need to convert it from linear to non-linear to more accurately represent the brain. In the GHBMC model, there are a ton of equations to plug in values and we need to construct a generalized Maxwell equation. Moreover, the model is viscoelastically modeled in its current state and we must update it to become hyperelastic, according to the advice of our advisers. For this segment of the project, we plan on consulting a biomechanics professor at Vanderbilt, namely Dr. Merryman, to pick his brain for insight on the brain’s biomechanical properties.
Simply put, our design project will consider two parameters: location and frequency of impact. After we have improved the model to a more indicative representation of the brain, we will then study the effect of location of stress on the head area and the frequency of these impacts on the brain. The data we are looking to study is the maximum strain on brain tissue as well as the plastic deformation caused by impact. Unfortunately, since concussions are currently diagnosed solely on symptoms of the patient and not based on clinical indications of structural changes in the brain, it is out of the scope of this project to correlate structural changes of brain tissue with the severity of a concussion.

Here are our action items before our next meeting after the spring break:
Material properties of brain tissue (hyperelastic vs viscoelastic material)
Biomechanics equations: time-dependent equations for non-linear elastic materials
Find average force and acceleration (linear and rotation) for 3 levels of play (high-school, college, NFL), for standard and concussive impacts
Determine the number of impacts that commonly occur within a successive play based on position
Lineman gets #hits per game/month at an average of #g force
Quarterback experiences #hits per game/month at an average of #g force
Identify most common impact location (ex. Frontal boss) for various positions (ex. quarterback vs linemen)
Locations of greatest focal deformation within the brain

Progress Report 3/1

From our last meeting with our advisers we have developed a plan of action for the next several months and an updated list of goals for our final product. We had previously been divided into pairs and assigned roles to accumulate research on the following areas:

Material properties of brain tissue (hyperelastic vs viscoelastic material)
• Linear Elastic (Isotropic or Orthotropic)
• Viscoelastic (Rate Dependent)
• Hyperelastic (Rate Independent)
• Rigid (Impactor

Biomechanics equations: time-dependent equations for non-linear elastic materials
• Hyperelastic model—Saint Venant-Kirchoff, Mooney Rivlin, Ogden (Polynomial)
• Viscoelastic model—Maxwell, Kelvin-Voigt (Standard linear solid)

Progress Report 3/11

Moved to isotropic model.

Find average force and acceleration (linear and rotation) for 3 levels of play (high-school, college, NFL), for standard and concussive impacts
• NFL averaged concussive impacts to be above a threshold of 98g of linear acceleration
• Average rotational acceleration is found to be about 6000 (rad/s^s)

Linemen gets # hits per game/month at an average of #g force
• Collegiate level: 30-50 impacts/game
• High School level: 20-30 impacts/game
• NFL: 40+ impacts/game
• G-force ranges from 10-100, but averages at 30 g’s
• Linemen experience the most impacts per game at often some of the highest average g forces

Quarterback experiences # hits per game
• Quarterbacks saw impacts at a rate of 5/practice and 10/game
• Impacts for most positions were 2.4 times greater in games than in practices

Identify most common impact location (ex. Frontal boss) for various positions (ex. quarterback vs linemen)

Locations of greatest focal deformation within the brain

Our next course of action is compiling information on mathematical properties and equations for implementation into the model, building a user-friendly interface for the model to be operated on, and segmenting the model elements into a 3D slicer.

 

Progress Report 4/3

Developed a probabilistic model of the material properties of the brain by data collection of impact forces and concussion diagnosis.

We’ve also updated the material properties of the model from isoelastic to viscoelastic to more accurately model the brain tissue.

Instead of an application, we are currently working on an instruction manual on our website which will provide a way for an open source download of the model. The model will output the probability of concussion based on the impact data.