Cheryl Olman

Assistant Professor

University of Minnesota

Departments of Psychology and Radiology

Lab webpage:  http://vision.psych.umn.edu/~olmanlab

BBB Seminar Courses Recent Projects Publications CV Contact Information Japan

My interests are split (roughly) evenly between fMRI methods and vision, specifically how low level vision can be modulated by scene structure and perception.  I did my PhD work in Dan Kersten's lab; my post-doctoral work was done with Kamil Ugurbil at the University of Minnesota's Center for Magnetic Resonance Research and in David Heeger's lab at in the Center for Neural Science at New York University.

The complications of studying natural images.  Like the majority of the vision research community, my research interests fall under the general heading of "how the visual system reconstructs a three-dimensional world from retinal input". The sub-heading that I am particularly interested in is low-level vision: how the V1 population of neurons encodes visual information for use by other parts of the brain.  This necessarily efficient code has been studied for more than half a century, and neural architecture and individual (single unit) responses to isolated image patches are well characterized.  However, single unit responses in a broader spatiotemporal context show plenty of (expected) nonlinearities.  My approach to the problem is to combine computational modeling of V1 neural networks (taking advantage of known single unit data as well as predictions from psychophysics) with functional imaging experiments to study spatial interactions in the V1 population code.  Completed projects along these lines include the contrast paper and the Gabor project.  Current topics of interest: lateral inhibition and surround suppression; separating V1-intrinsic contextual modulation from feedback modulation.


The necessity of studying MR.  BOLD-based fMRI detects changes in neural activity through a complicated network of vascular responses to spatially averaged synaptic activity and energy consumption.  Careful imaging is important - there are plenty of ways to get bad data and artifactual results, while good data requires careful experimental design, analysis, and interpretation.  I am interested in both the mechanisms underlying the BOLD response, as well as the physics of the imaging process; completed projects along these lines include the PSF project, z-shim project,  3T localization project, and MTL high-resolution project (illustrated at left).

Website created 11/25/05.