Abstracts - 2006
Classifying Neural Correlates of a Cognitive Action and its Preparation with Machine Learning Methods
Yongwook Bryce Kim & Polina Golland
Since its development in the early 1990s, Functional Magnetic Resonance Imaging (fMRI) has played a tremendous role in visualizing human brain activity for the study of mechanisms of human brains and clinical practice. Acquired from a regular magnetic resonance machine with special parameter settings, this non-invasive imaging method measures changes in blood flow, which in turn are an indication of neural activity. fMRI produces four-dimensional time-series images (three-dimensional in space and one-dimensional in time) with relatively low temporal resolution and high spatial resolution.
This new imaging technique has generated a huge volume of new high dimensional data and hence, the need for new image analysis methods. In the recent literature, interesting discoveries in human cognitive states have been found using techniques from machine learning, especially multivariate pattern analysis and non-linear pattern classification methods. [1-4]
We explore the relationship between the brain activation in preparation for carrying out an action and the activation during the action itself. In particular, we are interested in developing statistical techniques for image analysis and understanding that can learn to classify and track the cognitive processes using fMRI images.
For example, deception is a complicated cognitive process. Success of deception may depend on processes carried out in the preparation period before the actual deception and on the context to which it is applied. The anatomical correspondence between the brain regions supporting preparation of a lie and its different types is almost entirely unknown. We are currently working on classifying neural correlates of deception and its preparation with machine learning methods.
Professor John Gabrieli Group, Department of Brain and Cognitive Sciences, MIT
This research is supported under the National Alliance for Medical Image Analysis, the Morphometry Biomedical Informatics Research Network, the Neuroimaging Analysis Center, the NIH NINDS R01 grant on Computational Modeling of Shape Distributions, the Engineering Research Center for Computer-Integrated Surgical Systems and Technology.
 J.V. Haxby et al. Distributed and overlapping representations of faces and objects in ventral temporal cortex, <em>Science</em> 293, 2425 (2001)
 T.M. Mitchell et al. Learning to decode cognitive states from brain images, <em>Machine Learning</em> 5, 145 (2004)
 S.M Polyn et al. Category-specific cortical activity precedes retrieval during memory search, <em>Science</em> 310, 1963 (2005)
 Davatzikos et al., Classifying spatial patterns of brain activity with machine learning methods, <em>NeuroImage</em> 28, 663 (2005)