Earlier this month, the American Journal of Neuroradiology (AJNR) announced nominees for the AJNR Lucien Levy best research article award. Among the shortlisted papers was a June 2022 study of mild traumatic brain injury (mTBI), also known as concussion, led by scientists at NYU Langone Health and the Center for Advanced Imaging Innovation and Research.
In the study, the authors conduct a diffusion analysis of brain MRI data from three groups: college football players with a sports-related concussion, college football players without a concussion, and athletes who played a non-contact sport. The football players who suffered a concussion were removed from play until their symptoms subsided. Meanwhile, the players without concussions remained exposed to repeated head impacts throughout the season. The athletes who did not play contact sports served as healthy controls.
The MRI data come from the Federal Interagency Traumatic Brain Injury Research registry, or FITBIR, a database created by the National Institutes of Health and the U.S. Department of Defense in order to help advance the understanding of mTBI in the civilian population and military personnel. To maximize consistency, the researchers focused on athletes playing a single sport (college football) and on data acquired with a single MRI scanner model.
“We didn’t want to lose information,” said Sohae Chung, PhD, assistant professor of radiology at NYU Langone and lead author of the study. “If we used multi-scanner data, we would have to harmonize it.”
Diffusion MRI is the leading noninvasive means of studying brain microstructure in vivo and relies on probing water, which is present throughout the body. As water molecules diffuse through the brain, various tissues restrict the natural random motion of these molecules in characteristic ways. MRI signal can detect this behavior, and theoretical models relate the acquired data to specific “tissue compartments,” which allow scientists to deduce certain features of the tissue microenvironment.
“We are the first to do a multicompartment diffusion analysis” of FITBIR data, said Dr. Chung, whose group uses imaging techniques to study concussion.
Researchers analyzed data from concussed football players at four time points: first, within 48 hours after concussion; second, when symptoms subsided and the players were being cleared to return to the field; third, a week after the return to play; and last, six months after the concussion. The scientists ran the same analysis on data gathered at corresponding times from the football athletes who remained in play (subject to repeated head impacts) and from the control group.
“We were curious whether athletes may be different and have different outcomes” than people who are not exposed to head impacts, said Dr. Chung.
The team found significant differences between the white matter of players who had concussions and that of healthy controls at every time point. Although these differences diminished over time, when the concussed athletes were returning to play, their brains still differed from those of the athletes in the control group. What is more, the differences in the brains of concussed football players persisted even six months after their mTBI, especially in the corpus callosum. This finding surprised the researchers.
“After six months … I expected they would be okay,” said Dr. Chung. “But they still have some changes and aren’t quite recovered. Maybe they need more follow-ups or should take greater care.”
And what about the football players who did not have concussions?
“In the repeated head impact group, we see a lesser degree of changes compared to the concussed group, but it shows a very similar pattern,” said Dr. Chung.
The finding adds to mounting evidence that repeated head impacts, like concussions, may cause adverse changes that predispose players to neurodegenerative conditions down the line.
The analysis shows that the players who were withdrawn from action to recover from an mTBI have fewer changes in white matter as they return to the field than their fellow athletes who have not had a concussion and therefore continued to play. “They keep playing, and they keep hitting,” said Dr. Chung.
In some sense, the study suggests that even though concussion is bad, repeated head impacts may be worse because they do not produce evident symptoms at the time. “If people don’t get a concussion, they continue to play and practice,” said Dr. Chung. “And they don’t know what happened to their brain.”
The multicompartment diffusion analysis has also uncovered clues to how concussion and repeated head impact correlate to slightly different types of changes. Indicators of axonal injury appear similar in concussed athletes and non-athletes, said Dr. Chung, but the non-concussed players exposed to head impacts show higher indicators of extra-axonal injury. “Axonal injury is actually more serious, harder to recover from than extra-axonal injury,” said Dr. Chung.
The framework for this type of analysis was developed by NYU Langone researchers about a decade ago and is known as the standard model of diffusion. “If you want to look at microsctructure changes in the intra-axonal and extra-axonal compartment, you need to distinguish these areas in the neuron,” said Dr. Chung. Thanks to the standard model, “we can see changes in these compartments and understand the meaning of these changes.”
Additional authors of this research were Junbo Chen, Tianhao Li and Yao Wang, PhD, at the Department of Electrical and Computer Engineering, NYU Tandon School of Engineering; and Yvonne Lui, MD, at NYU Langone’s Department of Radiology and the Center for Advanced Imaging Innovation and Research.
The research was supported by funding from the National Institute of Biomedical Imaging and Bioengineering award P41 EB017183; the National Institute of Neurological Disorders and Stroke awards R01 NS039135, R01 NS119767, R21 NS090349, R56 NS119767; the U.S. Department of Defense award PT190013; and the Leon Lowenstein Foundation.
Related Resource
MATLAB software for robust standard-model parameter estimation from diffusion MRI data