1. 1/10/13 OHBM
1/3https://ww4.aiev olution.com/hbm1301/index.cf m?do=abs.v iewAbs&subView=1&abs=1229
DTI Detection of Symptomatic and Asymptomatic Injury Due to
Repetitive Head Blows
Abstract Submission No:
1306
Authors:
Il Yong Chun1
, Allan Diaz2
, Xiaodong Li2
, Yun Jang Jin2
, Larry Leverenz2
, Eric Nauman2
, Thomas Talavage1
Institutions:
1
Purdue University, West Lafayette, United States, 2
Purdue University, West Lafayette, IN
Introduction:
Diffusion tensor imaging (DTI) has been used to detect white matter (WM) abnormalities in mild traumatic
brain injury (mTBI) patients, including those who have histories of concussion from sport (e.g., [2]). Recent
work using fMRI and neurocognitive testing [1,4,5] has confirmed the postulation [3] that blows that do
not result in clinical diagnosis of concussion (typically requiring overt deficits) can still produce changes in
brain function. Further, incidence rates of cognitive impairment do not appear to be related to history of
concussions [2], suggesting that accumulated sub-concussive blows may be a critical factor in eventual
development of cognitive deficits. Detection of pre-symptomatic changes could be critical to prevention or
treatment of neurodegenerative diseases, such as chronic traumatic encephalopathy (CTE), associated
with repeated blows to the head.
Methods:
Subjects: Three populations were evaluated using DTI (25 angles) at 3T. (1) 8 male high-school American
football players who experienced hits to the head throughout a 12-week schedule were scanned prior to,
during and after the season. The mean±s.d. of hits accumulated by each player was 849±570 (range:
223-1783). (2) A control population of 34 (20M/14F) high school-aged non-contact sport athletes, of which
11 (6M/5F) were scanned twice for test-retest reliability. (3) Two females (ages 19, 22) with known
cognitive deficits from recent concussions were scanned once.
Statistical Analysis: FA values were extracted from DTI images after eddy-current correction. FA images
were co-registered using non-linear registration with a FMRIB58 FA template, and aligned into the
1×1×1mm MNI152 standard space by affine transformation, and a white-matter-mask was applied. Due to
limited control size (23 individuals, 34 total measurements), the mean and standard deviation of the
general population cannot be well-characterized. Therefore, a bootstrapped z-score (zBS) was used to
characterize deviant and non-deviant measurements: zBS=(ζ-Mean(μ))/(STD(μ)), where ζ is the subject's
FA measurement at the voxel; μ∈RB
where μb is the mean of a set of uniformly sampled data with
replacement for b = 1,…,B; and B is the number of bootstrapped samples (here B=1000). We have
assumed that the resulting errors in FA values due to misregistration are smaller than FA changes due to
injury or aggregate head blow history.
Results:
The category average histograms in Fig 1 illustrate that Control, Player and Injured populations exhibit
characteristic zBS distributions. Symptomatic injuries (I1 and I2) are associated with a marked leftward
shift. Asymptomatic players who experience repeated head blows exhibit a flatter distribution and with
greater deviance on the right side of the histogram.
Counts of extrema (|zBS|>20; Fig 2) suggest zBS distributions exhibit injury-specific bias. Controls are
biased slightly toward high FA values (+ column), but exhibit fewer extrema than Player and Injured
populations. Players exhibit a marked bias for high FA values (+ column) at all assessments, indicating
chronic change in WM health, likely accumulated by years of contact-sport participation. Clinically-
diagnosed injuries (I1 and I2) exhibit marked bias for low FA values (- column). Subject I2 is not depicted
in Fig 2, due to a history of 10+ concussions and (Fig 1) a dramatic bias toward lower FA values (24.2% of
WM voxels with zBS<-20).

2. 1/10/13 OHBM
2/3https://ww4.aiev olution.com/hbm1301/index.cf m?do=abs.v iewAbs&subView=1&abs=1229
Conclusions:
The zBS distributional analysis detects changes in FA in individuals with a history of concussive and/or sub-
concussive head trauma. The analysis suggests that different WM injuries are associated with the

3. presence or absence of concussion symptoms in these cases. These findings suggest the proposed DTI
analysis can be an effective means of detecting pre-symptomatic white matter damage associated with
repetitive blows to the head.
Disorders of the Nervous System:
Traumatic Brain Injury
[1] Breedlove, Evan L. et al. (2012), "Biomechanical Correlates of Symptomatic and Asymptomatic
Neurophysiological Impairment in High School Football", Journal of Biomechanics.
[2] Hart, John (2013) et al., “Neuroimaging of Cognitive Dysfunction and Depression in Aging Retired
National Football League Players”, Journal of the American Medical Association Neurology.
[3] McKee, Ann C. et al. (2009) "Chronic Traumatic Encephalopathy in Athletes: Progressive Tauopathy
following Repetitive Head Injury", Journal of Neuropathology and Experimental Neurology.
[4] Poole, Victoria N. et al. (2012) “Asymptomatic high school contact sport athletes exhibit metabolic
changes that correlate with head collision history,” Neuroscience 2012.
[5] Talavage, Thomas M. et al. (2010), "Functionally-Detected Cognitive Impairment in High School Football
Players Without Clinically-Diagnosed Concussion", Journal of Neurotrauma.