Estimating multivariate similarity between neuroimaging datasets with sparse canonical correlation analysis: an application to perfusion imaging.

An increasing number of neuroimaging studies are based on either combining more than one data modality (inter-modal) or combining more than one measurement from the same modality (intra-modal). To date, most intra-modal studies using multivariate statistics have focused on differences between datasets, for instance relying on classifiers to differentiate between effects in the data. However, to fully characterize these effects, multivariate methods able to measure similarities between datasets are needed. One classical technique for estimating the relationship between two datasets is canonical correlation analysis (CCA). However, in the context of high-dimensional data the application of CCA is extremely challenging. A recent extension of CCA, sparse CCA (SCCA), overcomes this limitation, by regularizing the model parameters while yielding a sparse solution. In this work, we modify SCCA with the aim of facilitating its application to high-dimensional neuroimaging data and finding meaningful multivariate image-to-image correspondences in intra-modal studies. In particular, we show how the optimal subset of variables can be estimated independently and we look at the information encoded in more than one set of SCCA transformations. We illustrate our framework using Arterial Spin Labeling data to investigate multivariate similarities between the effects of two antipsychotic drugs on cerebral blood flow.

Amygdala subregions tied to SSRI and placebo response in patients with social anxiety disorder.

The amygdala is a key structure in the pathophysiology of anxiety disorders, and a putative target for anxiolytic treatments. Selective serotonin reuptake inhibitors (SSRIs) and placebo seem to induce anxiolytic effects by attenuating amygdala responsiveness. However, conflicting amygdala findings have also been reported. Moreover, the neural profile of responders and nonresponders is insufficiently characterized and it remains unknown whether SSRIs and placebo engage common or distinct amygdala subregions or different modulatory cortical areas. We examined similarities and differences in the neural response to SSRIs and placebo in patients with social anxiety disorder (SAD). Positron emission tomography (PET) with oxygen-15-labeled water was used to assess regional cerebral blood flow (rCBF) in 72 patients with SAD during an anxiogenic public speaking task, before and after 6-8 weeks of treatment under double-blind conditions. Response rate was determined by the Clinical Global Impression-Improvement scale. Conjunction analysis revealed a common rCBF-attenuation from pre- to post-treatment in responders to SSRIs and placebo in the left basomedial/basolateral and right ventrolateral amygdala. This rCBF pattern correlated with behavioral measures of reduced anxiety and differentiated responders from nonresponders. However, nonanxiolytic treatment effects were also observed in the amygdala. All subgroups, including nonresponders, showed deactivation of the left lateral part of the amygdala. No rCBF differences were found between SSRI responders and placebo responders. This study provides new insights into the brain dynamics underlying anxiety relief by demonstrating common amygdala targets for pharmacologically and psychologically induced anxiety reduction, and by showing that the amygdala is functionally heterogeneous in anxiolysis.

Neuroimaging evidence of altered fronto-cortical and striatal function after prolonged cocaine self-administration in the rat.

Cocaine addiction is often modeled in experimental paradigms where rodents learn to self-administer (SA) the drug. However, the extent to which these models replicate the functional alterations observed in clinical neuroimaging studies of cocaine addiction remains unknown. We used magnetic resonance imaging (MRI) to assess basal and evoked brain function in rats subjected to a prolonged, extended-access cocaine SA scheme. Specifically, we measured basal cerebral blood volume (bCBV), an established correlate of basal metabolism, and assessed the reactivity of the dopaminergic system by mapping the pharmacological MRI (phMRI) response evoked by the dopamine-releaser amphetamine. Cocaine-exposed subjects exhibited reduced bCBV in fronto-cortical areas, nucleus accumbens, ventral hippocampus, and thalamus. The cocaine group also showed an attenuated functional response to amphetamine in ventrostriatal areas, an effect that was significantly correlated with total cocaine intake. An inverse relationship between bCBV in the reticular thalamus and the frontal response elicited by amphetamine was found in control subjects but not in the cocaine group, suggesting that the inhibitory interplay within this attentional circuit may be compromised by the drug. Importantly, histopathological analysis did not reveal significant alterations of the microvascular bed in the brain of cocaine-exposed subjects, suggesting that the imaging findings cannot be merely ascribed to cocaine-induced vascular damage. These results document that chronic, extended-access cocaine SA in the rat produces focal fronto-cortical and striatal alterations that serve as plausible neurobiological substrate for the behavioral expression of compulsive drug intake in laboratory animals.