Learning and predicting cognitive states from neuroimaging data collected in a natural environment

Francois Meyer
University of Colorado, Boulder
Electrical Engineering

The goal of traditional functional magnetic resonance imaging is to identify regions of the brain that are maximally correlated with a simple cognitive or sensory stimulus. This type of experimental paradigm is dictated by the dogma of functional specialization. Very recently, fMRI has been used to infer subjective experience and brain states of subjects immersed in natural environments. Such environments provide the subject with an experience similar to a real life environment with a wide variety of stimuli that are not controlled. Conventional methods of analysis of neuroimaging data fail to unravel the complex activity that natural environments elicit. We will describe a new method that yields a representation of the cognitive states experienced during an fMRI experiment. We will present experiments conducted with the datasets of the
2007 Pittsburgh Brain Activity Interpretation Competition (http://www.ebc.pitt.edu/2007/competition.html).

Audio (MP3 File, Podcast Ready)

Back to Workshop IV: Image Processing for Random Shapes: Applications to Brain Mapping, Geophysics and Astrophysics