Chronic pain troubles over 40,000,000 Americans, reducing their quality of life and presenting a significant burden on the American medical system. When non-invasive treatments fail, many chronic pain patients turn to spinal cord stimulation (SCS). In SCS, a device which delivers small electric shocks is implanted in the patient's spinal cord. Fortunately, SCS helps about 50% of patients! Why, though, does SCS help some patients but not others? Why does SCS help at all? We seek to answer these questions by treating the network of neurons constituting the spinal cord as a system of microcircuits, wherein each microcircuit consists of several interacting populations of neurons. These microcircuits reflect the heterogeneity experimentally observed in the spinal cord, and may be divided into classes according to their network structures, neuron types, and expected behaviors. To understand how these microcircuits respond to increasing stimulation intensities, we model each microcircuit as a system of differential equations with noisy input corresponding to SCS stimulation intensity, wherein the model variables are the population averages of the membrane voltages of the neurons. Through such modeling, we find various means by which we can induce the microcircuits to transmit painful signals as they might for patients experiencing chronic pain. Moreover, we show that increasing SCS stimulation intensity from zero can reduce, increase, or non-monotonically affect painful signals according to microcircuit class and within-class variation in model parameters. This suggests that we can identify which sets of microcircuits collectively transmit less pain in response to increasing stimulation intensities, and hence for which SCS could alleviate chronic pain. In this talk, we illustrate such results, explain the model, and highlight some of the existing theory behind the mechanisms of SCS. No prior knowledge of neuroscience is assumed! Speaker(s): Alexander Ginsberg (University of Michigan)