Surface shape and curvature scales
Image and Vision Computing
On digital distance transforms in three dimensions
Computer Vision and Image Understanding
Scale-Space Theory in Computer Vision
Scale-Space Theory in Computer Vision
Machine Learning
Virtual endoscopic visualization of the colon by shape-scale signatures
IEEE Transactions on Information Technology in Biomedicine
MICCAI'11 Proceedings of the Third international conference on Abdominal Imaging: computational and Clinical Applications
Hi-index | 0.00 |
Dual-energy computed tomographic colonography (DE-CTC) provides detailed information about the chemical composition of colon that can be used to improve the accuracy of computer-aided detection (CAD). We investigated how to calculate a thick target region for volumetric detection of lesions in DE-CTC. After automated extraction of the region of colonic lumen, the target region is calculated by use of a distance-based scheme, where the image scale of the shape features that are used for the detection of lesion candidates is adapted to the thickness of the target region. False-positive (FP) detections are reduced by use of a random-forest classifier. The detection accuracy of the CAD scheme was evaluated at 5 thicknesses of the target region by use of a leave-one-patient-out evaluation with 23 clinical minimal-preparation DE-CTC cases including 27 lesions ≥6 mm in size. The results indicate that the optimal choice of thickness depends on the size and morphology of the target lesion. At optimal thickness, the per-patient sensitivity was 100% at 5 FP detections per patient on average, where the per-lesion sensitivity was 100% (94%) for lesions ≥10 mm (6 --- 9 mm) in size. The results compare favorably with those of our previous approach.