Software Made For The Military Is Being Used By The NFL To Diagnose Brain Injuries

Ayasdi hopes to identify specific kinds of traumatic brain injury that can eventually be pegged to different treatment options.

Software Made For The Military Is Being Used By The NFL To Diagnose Brain Injuries
[Concept of the human brain: Guru 3D via Shutterstock]

A software platform originally developed with funding from the Defense Department and America’s intelligence agencies is now being used for a novel purpose: Diagnosing traumatic brain injuries. Ayasdi, one of Fast Company’s Most Innovative Companies for big data, is collaborating with the University of California-San Francisco (UCSF), the Icahn School of Medicine at Mount Sinai, GE, and the National Football League to develop novel ways of visualizing neurotrauma. If it works, doctors will have a new way of visualizing what happens when brain injuries occur.

Devi Ramanan, a program director at Ayasdi who is working on traumatic brain injury research, explained to Co.Labs that the company’s main product, a topological data platform which generates 3-D models of extremely complicated data sets, is being used to find more granular diagnoses of injuries. Ayasdi and their partners hope to identify specific kinds of traumatic brain injury that can eventually be pegged to different treatment options. Essentially, their hope is that software platforms will find small differences that a human researcher couldn’t.

There’s precedent for this, and it comes from the world of pro sports. Muthu Alagappan, a former intern at Ayasdi, used the company’s software to claim there are really 13 different basketball positions, rather than five. Alagappan’s analysis won first prize at the MIT Sloan Sports Analytics Conference and earned him consultant contracts with the Miami Heat and the Portland Trail Blazers. Although basketball is far different from neurotrauma, both generate extensive sets of data points which can then be leveraged by software.

Another pro sports league, the NFL, is helping to fund ongoing neurotrauma research at UCSF using Ayasdi’s software. The joint Ayasdi-UCSF team received a $300,000 research grant from GE and the NFL earlier this year as part of the football league’s “Head Health Challenge.” UCSF’s Adam Ferguson and Esther Yuh are using the big data platform to sort through results gained from a new type of brain imaging technology called Diffusion Tensor Imaging (DTI). DTI is useful for medical professionals in detecting mild injuries that wouldn’t necessarily show up in a MRI. Ayasdi’s visualization software is then used to create geometric-like analyses of the results, where anomalies show up as changes in the shapes generated on the screen.

Part of the challenge, said Ramanan and Ayasdi cofounder Gunnar Carlsson, is the massive amount of data generated by DTI. The imaging technology captures the way water diffuses in the brain in 3-D, which creates a large amount of data to parse and analyze. Because the data sets created by the images include more than 100,000 voxels (3-D pixels) mapping the white matter in the brain alone, analyzing them causes logistical challenges for medical researchers. Using Ayasdi, or a similar platform, simplifies what Ramanan calls the “needle in a haystack game” for UCSF’s team. Research at UCSF using Ayasdi’s platform is still ongoing.

But across the country in New York, another study is taking place using the same software as UCSF. The Icahn School of Medicine at Mount Sinai (which also made our Most Innovative Companies in big data list) is using Ayasdi to create visual analysis of combined MRI and imaging data from brain diseases. Although Ayasdi would not go into specifics due to the fact that research is ongoing and still has not been published, the Icahn-Ayasdi partnership is centered on linking brain imaging data to a data set supplied by the University of Pennsylvania, which will then be used for research into schizophrenia and related conditions. The medical school and software company already work together to create precision medicine protocols for Type 2 diabetes.