Automated Dielectric Model Generation

Dosimetry studies require detailed knowledge of tissue dielectric properties. Previous studies have relied on slice-by-slice segmentation of organs from magnetic resonance images; a time-consuming process that limits application of the method to a wider variety of anatomic data sets. We propose to overcome this problem using a novel method of determining dielectric properties on a voxel-by-voxel basis. This approach, when combined with traditional segmentation methods applied to a small subset of organs, will provide fully automated, extremely rapid computation and assignment of tissue characteristics for large anatomic data volumes. In Phase I, we will implement and test a prototype version of our dielectric parameter computation algorithms, as well as initiate development of a data visualization environment that is capable of displaying and editing anatomic data, dielectric properties, and the results of thermal analyses. The Creare team assembled for this project has expertise in thermal dosimetry, image segmentation, data visualization, and software development.