Skip to content (access key 's')
Logo of Technion
Logo of CS Department
Logo of CS4People

The Taub Faculty of Computer Science Events and Talks

3D Virtual Colonoscopy with Computer-Aided Polyp Detection
event speaker icon
Arie Kaufman
event date icon
Sunday, 06.01.2008, 11:30
event location icon
EE Meyer Building 1061
This talk describes 3D virtual colonoscopy (VC), a combination of computed tomography (CT) scanning and volume visualization technology, and is poised to become the procedure of choice in lieu of the conventional optical colonoscopy for mass screening for colon cancer. The talk further discusses a novel pipeline of computer-aided detection (CAD) of colonic polyps – the precursor of colorectal cancer – complementing VC. In VC, the patient abdomen is imaged by a helical CT scanner during a 40-second single-breath-hold. A 3D model of the colon is then reconstructed from the CT scan by automatically segmenting the colon out of the rest of the abdomen and employing an "electronic cleansing" algorithm for computer-based removal of the residual material. The visualization software, running on a PC, allows the physician to interactively navigate through the colon using volume rendering. An intuitive user interface with customized tools supports 3D measurements, volume-rendered "virtual biopsy" to inspect suspicious regions, and "painting" to help in visualizing 100% of the colon surface. Unlike conventional optical colonoscopy, VC is patient friendly, fast, non- invasive, more accurate, cost-effective procedure for mass screening of colon polyps.

Our CAD pipeline automatically detects polyps by integrating volume rendering and conformal colon flattening with texture and shape analysis. The colon is first digitally cleansed, segmented, and extracted from the CT dataset of the abdomen. The colon surface is then mapped to a 2D rectangle using conformal mapping, thereby converting the problem from 2D to 3D. This flattened image is rendered using a direct volume rendering of the 3D colon dataset with a translucent transfer function generating an electronic biopsy of the entire colon. Suspicious polyps are detected by applying a clustering method on the 2D flattened image. The false positives (FPs) are reduced by analyzing shape and texture features. Compared with shape-based methods, ours is much faster and much more efficient as it avoids computing shape parameters for the whole colon wall. We tested our method and found it to be very sensitive to adenomatous polyps with a low rate of FPs. The CAD results are seamlessly integrated into a VC system, providing the radiologists with visual cues and likelihood indicators of areas likely to contain polyps.