The focus of this paper is to apply some of the new analytic techniques to positron emission tomography (PET) data in order to test the neuroanatomical plausibility of two very specific models of language processing: the model of Geschwind (1965), and that of Petersen et al. (1988; 1989). Model plausibility was assessed by fitting both models to a word repetition and a synonym generation rCBF PET dataset using methods based on the examination of interregional correlations, and structural equation modeling. / Physiologically-based regions of interest were defined a priori through a review of PET activation literature. The current study demonstrated (1) the feasibility of using a priori defined physiological regions of interest in model evaluation, (2) the advantages of using structural equation modeling over interregional correlations, and (3) that neither model proved to be a particularly good fit to the data.

No description available.

No description available.

No description available.

Thesis (Ph. D.)--University of Washington, 1998. / Vita. Includes bibliographical references (leaves [159]-167).

Thesis (Ph. D.)--University of Wisconsin--Madison, 1983. / Typescript. Vita. eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references (leaves 254-263).

This work presents an analytic three dimensional image reconstruction algorithm that was developed for a proposed volume-imaging PET scanner. The development of the algorithm was motivated by the scanner's ability to collect an order of magnitude more data than current PET systems and the lack of an efficient algorithm that could use the extra data. The algorithm is based on an extension of the Recovery Operator of Orlov[68] and operates by convolution in object space. This method of operation sets it apart from other analytic direct image reconstruction algorithms that rely on Fourier transforms. The algorithm is tested with ideal data and parameters that are appropriate to the new PET scanner. The results of the test show that the algorithm behaves as expected except for a 17% overshoot in the reconstructed value in one area. An explanation of this artifact is suggested, although not verified. Finally, the efficacy of the algorithm is demonstrated by proving that it is functionally equivalent to Fourier transform methods. / Science, Faculty of / Physics and Astronomy, Department of / Graduate

No description available.

No description available.

No description available.