The specific aim of this project is development of a suite of novel methods to facilitate the use of CT colonography for colorectal cancer screening. Colorectal cancer, the second leading cause of cancer-related death, can be largely prevented by early identification and removal of precursor lesions. Unfortunately, compliance with colorectal cancer screening remains poor. CT colonography is a fast, non-invasive, and safe test that has the potential to significantly increase screening compliance. However, in order for CT colonography to become a primary screening modality, several current limitations must be overcome. This project attempts to overcome those limitations by developing: algorithms for automated polyp detection to assist radiologists in accurately evaluating large numbers of studies from a screening population with low disease prevalence, methods for digital removal of tagged stool to permit testing to be performed without the need for complete evacuation, enhanced workflow tools and improved visualization methods to permit fast and accurate review of the large volumes of data produced by each examination, and provide more detailed quantification of lesions to assist radiologists in identifying specific lesions with the greatest malignant potential. The Phase II effort will include evaluation of our methods by comparison to radiologist reports and the results of optical endoscopic examinations. We will also work with leading radiologists and a gasteroenterologist to optimize the clinical utility of our work.
