Developing drugs is a very tedious process that involves rigor testing on animals and on humans. While treatment effect for most of the cases can be relatively easy quantified, brain activity is a challenge.
Measuring qualitative activity of human brain is a very non trivial task. It is only possible to get indirect measurement of what is treated and patients participate in a game that involves certain brain activities. In PyMC Labs I helped to create the model and a packaged solution to estimate the latent property of brain activity. The Bayesian model allowed to compile indirect measurements to inform about drug efficiency.