Automated Computer-Aided Polyp Detection for Computed Tomography Colonography (Virtual Colonoscopy)
Posted Feb 11 2010 4:00pm
Description of Invention: This invention describes an automated method for colon registration from supine and prone scans that combines the use of Computed Tomographic Colonography (CTC) and Computer Aided Detection (CAD) software. Currently, in order to detect colonic polyps, patients are scanned twice - once in the supine, and again in the prone positions. This approach improves CTC sensitivity by reducing the extent of non-interpretable collapsed or fluid-filled segments. In order to assist radiologists in interpreting CTC data or evaluating colonic polyp candidates detected by CAD in both scans, it is important to provide not only the locations of suspicious polyps, but also the possible matched pairs (correspondences) of polyps in these scans. To achieve this, the two scans need to be aligned. In this invention, the colon registration problem is formulated as time series matching along the centerline of the colon. Anatomically salient points on the colon are initially distinguished as they can be viewed as landmarks along the central path of the colon. Correlation optimized warping is then applied to the segments defined by the anatomical landmarks to find better global registration based on the local correlation of segments.
When CTC is performed in conjunction with CAD software, screening may become easier on patients, less time-consuming, and more accurate. The effectiveness of the method was verified in experiments in which the polyp location was used as a measure for the registration error. The algorithm was tested on a CTC dataset of 12 patients with 14 polyps. Experimental results showed that by using this method, the estimation error of polyp location could be reduced 60.4% (from 47.2mm to18.7mm on average) compared to a traditional method based on dynamic time warping.
Colon cancer is the second leading cause of cancer-related deaths in the United States, and the method used in this invention will aid in effective early detection of the disease, which will have a significant impact on its prognosis.
Applications: Efficient and robust detection of colon cancer
Huang A, Roy D, Franaszek M, Summers RM. Teniae coli guided navigation and registration for virtual colonoscopy. Visualization, 2005. VIS 05. IEEE, pp. 279-285, 23-28 Oct 2005; doi 10.1109/VISUAL.2005.1532806.
Licensing Status: Available for licensing.
Portfolios: Devices/Instrumentation Devices/Instrumentation - Diagnostics Devices/Instrumentation - Software Cancer Cancer - Diagnostics
For Additional Information Please Contact: Jeffrey James Ph.D. NIH Office of Technology Transfer 100 N. Charles St., 5th Floor, Baltimore, MD 21201 United States Email: firstname.lastname@example.org