A Distributed Network of Cerebral Blood Flow Changes Accurately Discriminates Methylphenidate and Atomoxetine: A Gaussian Process Pattern Recognition Approach

Andre Marquand¹, Sara De Simoni, Owen O'Daly, Fernando Zelaya, Mitul Mehta

Methylphenidate and atomoxetine are widely used for the treatment of attention-deficit/hyperactivity disorder, but their differential effects on human brain physiology are poorly understood. We apply a multivariate pattern recognition algorithm (Gaussian process classification) to continuous arterial spin labelling data recorded while subjects were at rest which accurately discriminates methylphenidate from atomoxetine and each drug from placebo. We show a distributed network of brain regions underlies discrimination, with differential effects in putamen, anterior cingulate and temporal poles. Multivariate pattern recognition may be a useful technique for detection of diffuse pharmacological effects.