Detection of TBI-Related Anomalies in Single-Subject DTI Scalar Images
John M. Ollinger 1 , PIng-Hong Yeh 1 , David Joy 1 , Terrence R Oakes 1 , and Gerard Riedy 1
NICoE, Walter Reed National Military Medical
Center, Bethesda, Maryland, United States
John M Ollinger, Ping-Hong-Yeh, David Joy, Terrence R.
Oakes, Gerard Riedy Although DTI parameters such as FA,
AD, RD, and MD have demonstrated differences between
patients with TBI and controls in group studies, they
have not yet been applied in the clinic. The method
proposed here uses tensor normalization, matched numbers
of gradient vectors, and the Box-Cox transformation to
derive t-statistics and Hotelling T2 maps which are
thresholded to detect anomalous regions. The specificity
and sensitivity are characterized by the probability of
replication and the total number of detected regions as
proxies. The probability of replication is shown be vary
between 0.4 and .7 depending on the parameters chosen.
This abstract and the presentation materials are available to members only;
a login is required.