Courtesy of Dr Filippo Queirazza, this image shows weight maps from a feature selection step in multivariate analysis of fMRI data acquired from participants with depression performing a probabilistic reversal-learning task. After the fMRI scanning session, participants engaged in an online self-help programme based on cognitive behavioural therapy (CBT). Approximately half of the participants were classified as responders to treatment, based upon reduction in clinical depression inventory score upon completing the intervention. Multivariate classification of pretreatment fMRI contrast images produced the results shown above, identifying right amygdala (A) and right striatum (B) as the most discriminative features to classify individual response to treatment. Thus, pretreatment activation in those brain regions while acquiring and processing feedback information during probabilistic learning offers significant classification of participants who subsequently respond or do not respond to CBT. The results suggest this fMRI task activation holds the potential to be adopted as a predictive biomarker of response to CBT in depression.


The image is taken from a recent study published in Science Advances:

Queirazza F, Fouragnan E, Steele JD, Cavanagh J, Philiastides MG. Neural correlates of weighted reward prediction error during reinforcement learning classify response to cognitive behavioral therapy in depression. Sci Adv 2019; 5(7): eaav4962.