Jeff VanOss

Project Title: Automatic atlas based registration and segmentation of bones from fused CT/nuclear imaging data sets

Longitudinal preclinical imaging studies generate large quantities of data. Analyzing this data is a time consuming process, usually done by hand, sometimes with the aid of software. Relying on human interpretation can lead to different and sometimes conflicting results. My project uses an articulated skeletal atlas of a mouse. This atlas is registered to the ct data, then the registered atlas is applied to the fused nuclear imaging data set to extract uptake for individual bones. The entire process is automated, saving the researcher time and providing consistent results.

Page last modified February 28, 2012