Abstract The processing of rewards and losses are crucial for learning to adapt to an ever changing environment. Dysregulated reward processes are prevalent in mental health and substance use disorders. While many human brain measures related to reward have been based on activity in individual brain regions, recent studies indicate that many affective and motivational processes are encoded in distributed systems that span multiple regions. Consequently, decoding these processes using individual regions yields small effect sizes and limited reliability, whereas predictive models based on distributed patterns yield much larger effect sizes and excellent reliability. To create such a predictive model for the processes of rewards and losses, from now on termed the Brain Reward Signature (BRS), we trained a LASSO-PCR model to predict the signed magnitude of monetary rewards and losses on the Monetary Incentive Delay task (MID; N = 39) and achieved a high significant decoding performance (92% for decoding rewards versus losses). We subsequently demonstrate the generalizability of our signature on another version of the MID in a different sample (92% decoding accuracy for rewards versus losses; N = 12) and on a gambling task from a large sample (73% decoding accuracy for rewards versus losses, N = 1084) from the Human Connectome Project. Lastly, we also provided preliminary evidence for specificity to rewarding outcomes by illustrating that the signature map generates estimates that significantly differ between rewarding and negative feedback (92% decoding accuracy) but do not differ for conditions that differ in disgust rather than reward in a novel Disgust-Delay Task (N = 39). We thus created a BRS that can be used to make specific, generalizable and reproducible predictions about brain responses to rewards and losses.

A multivariate brain signature for reward