Using miniscopes and deep learning to compare neurobehavioral representations of psychostimulant and opioid self-administration

Concurrent abuse of psychostimulants and opioids represents a growing public health concern. However, preclinical models of substance use disorder often situate psychostimulants and opioids similarly within a unified self-administration procedure. This approach may fail to capture important differences in neurobehavioral activity related to each drug type. A large volume of <i>in vivo</i> literature suggests that, within canonical reward-related brain regions such as the nucleus accumbens (NAc), multiple reward-seeking behaviors may be represented by distinct neural populations. This comparison is often made between drug and natural rewards, however there is evidence for a similar distinction between psychostimulants and opioids. Here, we review the evidence for distinct neurobehavioral characteristics of psychostimulant versus opioid self-administration and consider the utility of two experimental approaches-miniscope calcium imaging and deep learning-assisted behavior analysis-in further exploring this topic.