July 10, 2008 — Dr. Dong Song, Research Assistant Professor of Biomedical Engineering, has been awarded a 1-year grant by the USC Zumberge Fund to initiate a research project entitled "Nonlinear Dynamic Modeling of Hippocampal System Function during Learned Behavior and Memory Formation."

Brain regions underlying cognitive function are massively parallel, multiple-input/multiple-output (MIMO) systems.  In any given brain region, information is represented in the ensemble firing of populations of neurons, i.e., spatio-temporal pattern of neural "spikes".  Brain regions process the information by transforming the incoming spatio-temporal patterns into outgoing spatio-temporal patterns.
This input-output transformation is highly nonlinear and dynamic, due to the complex underlying biological mechanisms.  More intriguingly, brain regions also learn from the incoming spatio-temporal patterns by altering their input-output transformational properties. This phenomenon, termed "plasticity" in neuroscience, can be viewed as a non-stationarity in systems engineering. While recent advances in multi-electrode technology have made it possible to record the simultaneous activities of populations of neurons from behaving animal, understanding such complex systems behaviors still remains one of the most challenging tasks in neuroscience.  The primary objective of Dr. Song's research is to model the MIMO, non-stationary, nonlinear dynamic transformation performed by hippocampus, a brain region that has long been known to be responsible for the formation of long-term memories, using recently developed, cutting-edge systems analytic methodologies.  Dr. Song's approach is to represent the input-output transformational property of hippocampus as Volterra kernels in a physiologically plausible model structure.  The emergence and formation of such properties will then be tracked using time-variant Volterra kernels and finally explained by a learning rule.  Dr. Song believes that this kind of modeling work constitutes the first critical step of addressing the two fundamental questions in neuroscience: (1) how the brain processes information, and (2) how the brain learns from experience.