Automated polyp detection at CT colonography: Feasibility assessment in a human population

PURPOSE: To test the feasibility of and improve a computer algorithm to aut omatically detect colonic polyps in real human computed tomographic (Ci) co lonographic data sets. MATERIALS AND METHODS: Twenty patients with known polyps underwent CT colon ography in the supine position. CT colonographic data were processed by usi ng a shape-based algorithm that depicts masses that protrude into the lumen . We studied nine shape criteria and three isosurface threshold settings. R esults were compared with those of conventional colonoscopy performed the s ame day. RESULTS: There were 50 polyps (28 were greater than or equal to 10 mm in si ze; 12, 5-9 mm; 10, <5 mm). The sensitivity with optimal settings for detec ting polyps 10 mm or greater was 64% (18 of 28). Sensitivity improved to 71 % (10 of 14) for polyps 10 mm or greater in well-distended colonic segments . Performance decreased for polyps less than 10 mm, poorly distended coloni c segments, and other shape algorithms. There was a mean of six false-posit ive lesion sites per colon. These sites were reduced 39% to 3.5 per colon b y sampling CT attenuation at the lesion site and discarding sites having at tenuation less than a threshold. CONCLUSION: Automated detection of colonic polyps, especially clinically im portant large polyps, is feasible. Colonic distention is an important deter minant of sensitivity. Further increases in sensitivity may be achieved by adding prone CT colonography.