A New Frontier for Understanding Brain Injury

By Lt. Cmdr. Peter Walker, Naval Medical Research Center

The Neurotrauma Department at the Naval Medical Research Center (NMRC) has adopted an approach to help develop an understanding of the underlying mechanisms behind Traumatic Brain Injury (TBI).

Dr. Francoise Arnaud and I are exploring the use-spectral-graph methods to understand the relationship among multitudes of parameters gathered during the experimentation on aerial evacuation after blast injury and pre-exposure to stress in a laboratory model. Specifically, we are interested in discovering the strongest determining factors that lead to mortality. Advances in computing power and innovative approaches in the field of machine learning have paved the way for a new generation of scientists called data scientists.

These next generations of data scientists seek to extract knowledge and discover insights from large volumes of data through a combination of data mining and predictive analytics. This discipline has come to embrace the diversity necessary to understand the complexity of problems facing a variety of scientific disciplines.

One area of interest among data scientists to explore is the use of graph analytics. Network/Graph models attempt to identify relationships among different variables in a dataset. Edges that connect two nodes in a network suggest that there is a relationship among those variables. The distance between those nodes indicates the strength of the relationship among those variables. More importantly, these relationships can be defined quantitatively; the edge weight can be used to indicate a measure of similarity/ dissimilarity between two nodes.

Specifically, Arnaud and I want to apply this form of analytical technique to understand the “small world” properties that might exist within this ongoing work on TBI. Most nodes in the graph can be connected through a series of “jumps” from one node to the next. Through fairly simple computations, the connectedness of a graph can be computed through metrics such as clustering coefficients and path lengths. Properties of smallworld networks have been witnessed in a number of different real-world settings including brain activation and the analysis of social network sites.

Many are familiar with the game “six degrees of separation,” which suggests that any singular individual can be connected to another individual in six or less “jumps” in the network. Such properties are useful to the analyst, as well as the scientist, by illustrating the connectedness of the dataset as a whole while helping to illustrate the relationships among the multitudes of variables.

In the Neurotrauma Department, we are hoping that, through the appropriate analysis and merging of multiple datasets we can develop a deeper understanding behind brain injury and function of the brain. We believe that analytic techniques such as small worlds and data scientists in general, can help to unravel complex cause and effect relationships that might exist within these large datasets. For example, our group is currently exploring the possibility of an interaction between altitude exposure, physiological stress, and survivability.

We believe that a small world model will help us to understand how these different variables directly (or indirectly) affect survivability. In the end, such approaches may not only yield better insights into the mechanisms of injury, but also provide a promise for better diagnosis and treatment for our wounded warriors.

Follow us on Twitter for military science and technology updates!

Disclaimer: Re-published content may have been edited for length and clarity. The appearance of hyperlinks does not constitute endorsement by the Department of Defense. For other than authorized activities, such as, military exchanges and Morale, Welfare and Recreation sites, the Department of Defense does not exercise any editorial control over the information you may find at these locations. Such links are provided consistent with the stated purpose of this DoD website.