Brent A. Field, PhD

I am a neuroscience researcher in Princeton University’s Princeton Neuroscience Institute. My research has empirical, theoretical and technical foci. Much of my recent research has focused on decision making, which is important for understanding of the brain and behavior, but also other statistical disciplines. For instance, much effort in the tech industry and corporate finance has gone into optimizing multi-arm bandit problems that are similar to ones we have investigated. These problems arise whenever one has finite resources to invest in multiple incompletely understood opportunities (but will learn more about their payoffs in the future). My colleagues and I have studied how well people optimize such problems. We have attempted to use peoples’ corresponding brain dynamics to infer how they try to find optimal solutions. Other related work has looked at how the brain’s reward system, which underpins decision-making, interacts with executive function. This body of work has been funded by the National Institute of Health and the Templeton Foundation.

Interest in optimality extends to a completely different domain. I have had a long-standing interest in neural and phenomenological dynamics associated with advanced meditation practice. Towards that end my colleagues and I have compiled a database comprising of around 25 different measures of “adept” Tibetan Buddhist meditators and matched controls. More recently we have extended this to experienced meditators with a Christian background. Although typical descriptions of meditation imply that it is an activity on which the mind must focus its resources, I am interested in the possibility that common forms of meditation, when well-practiced, reduce overall mental load. More specifically, might meditation de-noise neuronal signal processing and optimize performance? Much of this work has been funded by or otherwise has been in conjunction with the Mind and Life Institute.

As a former Microsoft Program Manager I bring a technical perspective to my work. Much effort in recent years has gone into understanding simultaneously-collected functional MRI (fMRI) and electroencephalographic (EEG) data. This has been a challenged area. There remain important opportunities for innovation. For instance, movement of any kind, even very subtle movement, in the scanner’s magnetic field buries signals from the brain in unwanted and statistically misbehaving noise. Towards improving upon this problem, I have been implementing a method of separating the brain-derived signal and noise with a multi-step machine learning algorithm (and lots of resources on Princeton’s high-performance computing cluster).

In recent years I have also studied the application of functional programming to scientific computation. I have interest in developing a new analysis platform that would run faster, be less bug-prone, and be easier to implement than current widely-used tools.

I also co-teach a graduate-level course on human neuroimaging.