Meeting Banner
Abstract #1710

Robust, Unbiased General Linear Model Estimation of PhMRI Data in the Presence of Variance in the Temporal Response Profile

Gautam Vithal Pendse1,2, Adam James Schwarz2,3, Richard Baumgartner2,4, Alexandre Coimbra2,5, David Borsook1,2, Lino Becerra1,2

1Imaging and Analysis Group (IMAG), McLean Hospital, Harvard Medical School, Belmont, MA, USA; 2Imaging Consortium for Drug Development (ICD), Belmont, MA, USA; 3Translational Imaging Group, Lilly Research Laboratories, USA; 4Biometrics Research, Merck Research Laboratories, USA; 5Imaging Department, Merck Research Laboratories, USA

PhMRI can elucidate brain circuits underlying pharmacological action and provide a translatable, pharmacodynamic biomarker of CNS activity for novel compounds in the early phases of drug development. As applications increasingly shift to compounds whose direct effects on brain activity may be more subtle including new chemical entities and novel target profiles there is increasing demand on accurate and robust estimation of the phMRI signal. We present a simple method to generate parsimonious design matrices that accurately estimate, within a GLM framework, the phMRI response in the presence of confounding signals and variability in temporal profile.