2010 Postdoctoral Fellowship Trainees

Sponsored by National Institutes of Health

View 2010 Predoctoral and Postdoctoral Fellowship Trainees


Name: Anastasia Flevaris
Sponsor: Steven Hillyard
Project Title: Spatio-temporal analysis of neural mechanisms underlying hierarchical perception
Email address:

In 2009, I completed my Ph.D. in the lab of Dr. Lynn Robertson at UC Berkeley. My dissertation work investigated the extent to which selective attention to spatial frequency (SF) mediates perception of the whole (“global”) versus the parts (“local”) during object perception in general as well as in the context of face perception.

My training grant project will use a combined ERP/fMRI approach to determine the precise timing, locations, and functional attributes of the neural mechanisms involved in processing global versus local objects and to elucidate the role of early attentional selection of visual information in different spatial frequency bands. The role of spatial frequency in object-based attention and its corresponding neural circuitry will be tested through a series of ERP and fMRI experiments to (1) examine the time course (i.e., “early” versus “late”) of the effects associating spatial frequency channel selection with  perceptual organization of and attention to different object levels, (2) isolate the neural systems involved in the selection of spatial frequencies for multi-tiered object perception through source localization and fMRI activation, and (3) determine the functional connectivity between visual regions to characterize the neural circuits involved



Name: Elan Liss Ohayon
Sponsor: Terrence Sejnowski
Project Title: A Neurocomputational Platform for Modeling the Anatomical and Dynamical Foundations of Social Cognition
Email address:

Elan Ohayon completed his PhD at the University of Toronto studying dynamics in random and spatial neural networks. He has previously applied these techniques to the study of epilepsy and autonomous activity demonstrating how autonomous transitions can take place in seizures and embodied networks. At INC he is collaborating with the Computational  Neurobiology Laboratory ( and the Laboratory for Cognitive Neuroscience ( to understand the anatomical principles underlying social cognition and changes seen in Williams Syndrome. He is also interested in bridging the gap between our understanding of brain function and phenomenological experience.

What are the fundamental structural and plastic mechanisms that allow the brain to maintain the persistent and interactive neural activity required for social cognition? The aim of this project is to explain the relation between: (a) network topology (b) activity dynamics and (c) social cognition. Large-scale, spatial, recurrent neural network models are being developed in order to identify the basic architectural principles that enable networks to remain interactive in the world.



Name: David Peterson
Sponsor: Terrence Sejnowski
Project Title: Learning perception/action hierarchies
Email address:

I received Bachelor's degrees in Electrical Engineering and Business Administration at the University of Colorado, Boulder.  In the process of getting my PhD in Computer Science at Colorado State University, I completed a program of coursework and research combining computer science, psychology, and neuroscience. In the first part of my postdoctoral training here at UCSD's INC, I have been using theoretical and experimental methods to investigate the role of dopamine in striatal synaptic plasticity and rewarded learning in humans.

The goal of this project is to develop and test a biologically-realistic computational model of cortico-basal ganglia circuits that are implicated in learning perception/action hierarchies. It will instantiate learning dynamics by incorporating extant data about dopamine-mediated plasticity at corticostriatal synapses. I will use the model to simulate the influence of different striatal dopamine abnormalities found in Parkinson's disease and dystonia to generate hypotheses about those patients' abilities to learn perception/action hierarchies. I will test the hypotheses experimentally by measuring perception/action hierarchy formation using sequence learning experiments with patients and age-matched controls. After adapting the model based on the experimental results, I hope to thereafter use the model as a neurobiologically-mechanistic platform for developing future lines of basic and clinical research.



Name: Ned T Sahin
Sponsor: Eric Halgren
Project Title: Neural Circuits for Reading, Inflecting and Producing Words: Spatiotemporal Mapping with Human Intracranial Electrophysiology and fMRI
Email address:

Ned T. Sahin, PhD, completed his B.A. at Williams College (Biology;  Neuroscience), M.S. at MIT (Brain and Cognitive Sciences Dept.), and Ph.D. at Harvard (Psychology Dept. – Cognition, Brain and Behavior), and won the Richard J. Hernnstein prize for that year’s best dissertation across all Harvard departments.  During grad school he did research at Massachusetts General Hospital (Martinos Center) and learned fMRI and human intracranial electrophysiology (ICE); and then began a post-doctoral fellowship at UC San Diego in the Multi-Modal Imaging Lab.  He has used the ICE method, in which high-fidelity recordings are made directly from brain cells (in patients about to undergo surgery for epilepsy), to uncover temporal and physiological dynamics in the human language system (c.f. Sahin, et al, Science 326, 445-9 (2009)).  His more recent research has uncovered transient functional networks that connect distant brain centers via synchronized neural firing – in order to divvy up, process, then bind together the many types of information required to read and produce words and sentences.

The present INC-funded project aims to close a gap in language research. Language is uniquely human and central to human nature, and ICE uniquely allows the resolution to find the natural kinds of neural processing, however the neuroscience of language is almost never studied in its natural setting: conversation. Instead, most previous studies including my own involve single-word or short utterances, in oversimplified and odd contexts unlike natural discourse. It is like trying to understand the rules of baseball by watching a player warming up in a batting cage. The proposed project will apply the extreme resolution of ICE (in collaboration with Prof. Eric Halgren), and the classifying ability of modern machine-learning algorithms (in collaboration with Prof. Terry Sejnowski), to uncover dynamics of word production during natural, extemporaneous speech. Timing and evoked-potential dynamics are hypothesized to be very different in organic speech compared to many laboratory language experiments. If consistent signatures are identified for typical grammatical constructions, or even for individual words, the trained algorithms will be used on new brain data to “read backward” from the neural signal and reconstruct what the patient was saying at the time.




Faculty Spotlight

Tzyy-Ping Jung
Elevated to IEEE Fellow for contributions to blind source separation for biomedical applications.

...more info

Staff Spotlight

INC's First Annual Hike N' Lunch Event
INC Staff at the Torrey Pines Glider Port.

...see more