CT Colonography
 

1. CAD for CT Colonography Colon cancer is the second leading cause of cancer deaths among men and women in the United States.  Most colon cancers can be prevented if precursor colonic polyps are detected and removed.  CT colonography has been proposed as a promising technique for colon cancer screening.  However, in order for CT colonography to be deemed a practical screening tool, a large number of images must be interpreted in a prompt and accurate fashion.  To this end, we developed a CAD scheme for the automated detection of polyps with CT colonography.  This scheme provides radiologists with the location of suspected polyps, thereby allowing for quick and accurate detection of suspicious lesions.  In this scheme, first, the colonic wall was extracted from an isotropic volume data set obtained by interpolation of CT colonography images along the axial direction.  Polyp candidates were detected by computation of 3D geometric features that characterized polyps, followed by extraction of polyps with hysteresis thresholding and fuzzy clustering by use of these geometric features.  False-positive detections were reduced by first extracting of 3D texture features from polyp candidates, then by applying of discriminant analysis to the candidates based on the geometric and texture features.  This scheme was applied to 71 CT colonography cases, 14 of which contained a total of 21 colonoscopy-confirmed polyps.  In a by-patient analysis, the sensitivity of our scheme was 100%, with an average false-positive rate of 2.0 per patient.  In a by-polyp analysis, the scheme detected 90% (19 of 21) of the polyps at the same false-positive rate.  This computerized scheme is potentially useful in reducing interpretation time and may enhance the diagnostic accuracy of radiologists in the detection of polyps in CT colonography.  (549, 684, 685, 686, 772, 776, 788, 789, 824, 825, 826, 827, 837, 838, 839, 840, 841)