Colorectal cancer (CRC) is the second leading cause of cancer death in the US. Early detection and removal of polyps could reduce the mortality of this disease by up to 90 percent. Current screening methods are either nonspecific, invasive, or hindered by poor patient compliance. Computed tomography colonography (CTC) is a promising technique for polyp detection, with some reports of sensitivity over 90 percent for 10 mm or larger lesions. If adequately validated and disseminated CTC could have a major impact on morbidity and mortality from CRC. Unfortunately, the search for small polyps in a long colon represented by several hundred CT images is tedious and time consuming, requiring up to an hour of costly physician time for accurate interpretation. Without substantially increasing the efficiency of interpretation, CTC has little chance of being cost-effective. Therefore, our primary goal is to minimize the time required for CTC interpretation while maximizing diagnostic accuracy. We will accomplish this by developing and validating computer aided diagnostic (CAD) methods for polyp detection, and making these methods part of the interpretive process.
Our Specific Aims are as follows: (1) Computer Aided Detection Algorithm Development: We will develop, integrate, and validate three different approaches to CAD for colorectal polyp detection. (2) Optimization of CTC Interpretation Efficiency: Using a sensitive CAD algorithm and reader interface development, we will experimentally determine the most efficient and accurate means of augmenting 2D and 3D displays with CAD methods for CTC interpretation. (3) Optimization of Image Acquisition - Multidetector Row Helical CT: We will empirically determine the best acquisition methods for depiction of the colon surface using multi-detector row helical CT, optimizing the tradeoff between dose, scan time, and depiction of the colon surface. In a prospective trial, we will compare these state-of-the-art CTC acquisition and interpretation methods (human visualization with/without CAD) with the current clinical gold standard of fiberoptic colonoscopy. Upon completion of this work, we will have successfully developed new methods that allow radiologists to interpret CTC studies efficiently and accurately, thereby enabling the widespread application of CTC. Successful development and deployment of optimized CTC has the potential to improve patient compliance with screening recommendations which, in turn, can increase detection of potential cancers with a concomitant reduction in morbidity and mortality from the disease.
Showing the most recent 10 out of 24 publications