According to Brent Doiron, Ph.D., professor of mathematics, collections of neurons organize their activity to perform a variety of computational, cognitive, and behavioral tasks. His research group is interested in the “neuro-mechanics” that underlie neural computation. He uses a combination of statistical mechanics, nonlinear systems theory, and information theory to study this broad field.
Of particular interest is the genesis and impact of neuronal variability on sensory processing and computation. Neurons do not have reliable patterns of behavior, and this variability is a consequence of the architecture of neural circuits, the nonlinearities inherent in how electrical signals are conducted, and the adaptability (or plasticity) of nerve cell endings called synapses. Doiron works with a range of experimentalists to understand how various stimuli and neural states allow neuronal networks to have both variability and also synchronized activity. Understanding the specific mechanisms that mediate this shaping, and the consequences these mechanisms have for how information is encoded by large numbers of nerve cells, are central challenges for sensory neuroscience.
His research team recently developed a computational model that provides deep predictions linking brain circuits to brain activity. The new model accurately explains experimental findings in the brains of living animals, according to scientists whose work is described in Nature Neuroscience. The new model provides a much richer understanding of how noisy, coordinated activity between neurons emerges in neural circuits. The model could be used in the future to discover neural “signatures” that predict brain activity associated with learning or disease, according to Doiron.
In 2017, Doiron received one of 13 prestigious Vannevar Bush Faculty Fellowships for the Department of Defense. These awards are given to faculty who conduct basic research in core science and engineering disciplines, such as neuroscience, which underpin future DoD technologies. In the spring of this same year, Doiron also was a co-recipient, together with professor Marlene Cohen of the Neuroscience department at the University of Pittsburgh, of an award from the Simons Foundation. This foundation is one of the country’s top private philanthropies supporting neuroscience.
Doiron received his B.Sc. and Ph.D. from the University of Ottawa. He conducted postdoctoral research at New York University before joining the University of Pittsburgh in 2007.
“Towards a circuit-based theory of cortical variability and its impact on neuronal processing”
Neurons are an unpredictable lot, varying their response from moment to moment within any given experiment. They also generate staggering amounts of information. Coping with such large-scale data and characterizing variability across brain regions will be essential to constructing models of how the brain processes sensory signals, and ultimately for developing an overarching theory of cognition. Brent Doiron and Matt Smith will examine the functional consequences of variability on specific brain computations and the behaviors that they generate.
This image depicts the formation of discrete and densely connected networks of neurons that function as physical memory traces. Orange lines indicate strong connections between groups of these excitatory neurons (triangles). Inhibitory neurons (dots) connect and regulate the activity of these groups (blue lines) preventing distinct memory traces from being activated simultaneously.