Surface curvature estimation is a common component of CT colonography computer-aided polyp detection algorithms.
A commonly used method to compute such curvatures employs convolution kernels. We have observed situations where
the kernel method produces inaccurate results that could lead to undesirable false negative and false positive polyp
diagnoses. In this paper, we numerically examine this method of curvature estimation. We propose optimal choices for
smoothing parameters intrinsic to the method. The proposed smoothing parameters achieve more accurate and reliable
curvatures compared to those reported in the literature. Our results include responses of the system with respect to
Gaussian smoothing and Gaussian noise, results on the accuracy of the curvature estimation as a function of the distance
from the true surface, and examples of specific topologies of the colonic surface for which the kernel method yields
Low radiation dose requirements create relatively noisy images that contribute to high numbers of false positive detections in CAD for CT colonography. Presumably image denoising techniques such as non-linear, edge-preserving smoothing filters can improve automatic colonic polyp detection in CT colonography by reducing overall per patient false positive rates. Here, we have evaluated multiple edge-preserving smoothing filters to determine whether this is so. Prone and supine scans from 81 asymptomatic, average-risk adults with adenomatous polyps were studied with and without smoothing. FROC curves were generated to analyze CAD results. A single, clinically relevant operating point was compared between the best smoothing filter results and the unsmoothed data. Improvement in performance was observed, but the differences were not found to be statistically significant for average dose CT colonography.