Monday, April 21, 2008

CNS Virtual Poster Session

Below are abstracts from the two TB West posters presented at Cognitive Neuroscience Society this year. You can get a copy of the posters from here:

If you would like your CNS poster included in this virtual poster session, feel free to email the title, authors, and abstract, along with a link to your poster. Thanks!


An fMRI Investigation of the Functional Specificity of Sentence Processing Networks: A Comparison of Sentences and Melodies

Corianne Rogalsky & Gregory Hickok
UC Irvine

A number of recent studies have identified portions of the anterior temporal lobe (ATL) that respond preferentially to structured sentence-level stimuli (versus word-lists, for example). It is unclear, however, whether this response to sentences reflects syntactic computations, semantic integration operations, or more general hierarchical structure-building. The present study directly compares the neural systems associated with sentence and melodic structure processing to investigate the specificity of this ATL activity. We implemented a mixed-design fMRI paradigm to compare activity in the ATL whiles subjects listened to blocks of jabberwocky sentences, scrambled jabberwocky sentences, simple novel melodies, and scrambled novel melodies. In order to separate activations associated with hierarchical structure processing from activations resulting from general temporal processing, stimuli were presented at three different rates within each block. Regions with a greater BOLD response to sentences than to scrambled sentences, and regions with a greater response to melodies than to scrambled melodies were identified. In agreement with previous research, inferior frontal areas and ATL sub-regions, bilaterally, were found to prefer sentence-level structure. Similar regions were found to prefer hierarchical structure in general: these areas were more active for melodies than scrambled melodies. Further analysis indicates that inferior frontal, not anterior temporal, regions are more active for sentences than melodies, once the response to corresponding scrambled conditions are subtracted out. These preliminary analyses suggest that regions that prefer sentence-structure in the ATL also are recruited during more general hierarchical-structure building. Supported by NIH DC03681.

Modulation of brain regions involved in overt picture naming by parametric variation in word frequency, word length and reaction time

Stephen M. Wilson [1,2], A. Lisette Isenberg [2], and Gregory S. Hickok [2]

[1] Dept of Neurology
University of California, San Francisco

[2] Dept of Cognitive Sciences
University of California, Irvine


Picture naming is a cognitive task commonly used to study lexical access. However it is a complex operation, entailing not only semantic and phonological stages of lexical access, but also ancillary processes such as visual processing, articulation, self-monitoring and executive functions. Previous neuroimaging studies have demonstrated recruitment of a wide range of brain areas which presumably support these various components of the task. In order to better delineate the functions of regions activated by picture naming, we used fMRI to identify brain areas where BOLD responses to picture naming were modulated by three different parametric variables: word frequency, word length and reaction time, each of which we hypothesized to be associated with different aspects of the task. Twelve subjects were scanned while they named 165 pictures in a rapid event-related design, and digital signal processing was used to extract vocal responses from background scanner noise. Lower frequency words were associated with greater BOLD responses in occipitotemporal cortex bilaterally. These regions are associated with visual and semantic processing. Longer words led to increased BOLD activity in speech motor areas as well as superior temporal cortex. Longer reaction times resulted in greater BOLD activity in areas including inferior frontal regions associated with both cognitive control and linguistic processes, and the pre-SMA. Of particular interest was a region in the left superior temporal sulcus correlated both with word length and reaction time (each independently of the other). We argue that such a pattern suggests a role for this region in retrieval of phonological form.

No comments: