The cornerstone of O'Doherty research is the use of computational algorithms, which include reinforcement learning, models that are derived in part from computer science and robotics. By combining these models with data taken using fMRI (functional magnetic resonance imaging), which measures changes in blood oxygenation—a proxy for neural activity—he and his colleagues can probe how the brain learns and makes decisions. This approach allows the researchers to characterize how a particular cognitive function is implemented, instead of merely identifying the location in the brain of such functions, which is typical of more traditional fMRI studies.

In addition, O'Doherty studies the effects of discrete lesions in specific brain regions on decision making. He also uses transcranial magnetic stimulation (TMS) to induce temporary lesions in healthy subjects. These methods allow the researchers to pinpoint the causal roles that specific brain regions may play when making decisions, thereby complementing the fMRI studies.

A deeper understanding of the brain will not only inspire new theories of decision making but also contribute to the development of artificial intelligence. It will enable scientists to learn why some humans are better at making decisions, why those with certain psychiatric disorders or brain lesions are less capable of doing so, and why some people systematically fail to make rational decisions under certain circumstances.

Key Research Areas:

Reward, Decision-Making, Reinforcement Learning, Conditioning, Associative Learning, Computational Neuroscience