Adam J. Schwarz1,2, Jaymin Upadhyay,
2,3, Alexandre Coimbra, 2,4, Richard Baumgartner, 2,5,
Julie Anderson, 2,3, James Bishop, 2,3, Ed George,
2,6, Lino Becerra, 2,3, David Borsook, 2,3
1Translational Imaging, Eli Lilly and
Company, Indianapolis, IN, United States; 2Imaging Consortium for
Drug Development, Boston, MA, United States; 3PAIN Group, Brain
Imaging Center, McLean Hospital, Belmont, MA, United States; 4Imaging,
Merck, West Point, PA; 5Biometrics Research, Merck, Rahway, NJ,
United States; 6Anesthesiology and Critical Care, Massechussets
General Hospital, Boston, MA, United States
Graph
theoretic analyses of functional connectivity networks report on topological
properties of the brain and may provide a useful probe of disease or drug
effects. However, verifying node-wise effects over a range of binarization
thresholds is inconvenient and often subjective for large, voxel-scale
networks. We present a straightforward method for calculating graph theoretic
node parameters that are robust to binarization threshold and suitable for
image analysis in the study of functional connectivity. The method is applied
to mapping drug modulation of localized functional network topology by the
opioid analgesic buprenorphine in healthy human subjects.