Computerized Detection of Polyps in CT Colonography

Massachusetts General Hospital, Boston MA

Investigators

Linked publications & trials

Abstract

Project Abstract Colon cancer, the second leading cause of cancer deaths for men and women in the United States, can be prevented if precursor colonic polyps are detected and removed. The long-term goal of the proposed project is to develop an image-based diagnostic method that advances the early detection of colon cancer. Computed tomographic colonography (CTC), also known as virtual colonoscopy (VC), is a promising technique for colon cancer screening. However, for CTC to be a practical screening tool, it should be easy for patient to accept, allow for visualization of the entire colonic mucosa, and should have high accuracy in detecting polyps. Current problems with CTC include low patient adherence when rigorous cathartic bowel cleansing is used, difficulties in visualizing the mucosal surface when residual fecal materials present, and variations in diagnostic performance among readers who interpret CTC cases. Laxative-free CTC (lfCTC) is an emerging technique for removing rigorous cathartic cleansing from bowel preparation in CTC examinations. Fecal tagging CTC (ftCTC) is a method of tagging stool and fluid remaining in the colon by a radiopaque oral contrast agent, allowing them to be effectively differentiated from polyps. Electronic cleansing (EC) is also an emerging technique for removal of tagged residual fecal materials in CTC images, virtually cleansing the colon without a prior physical bowel cleansing. Computer-aided detection (CAD) of polyps, which automatically detects polyps in CTC images, has the potential to detect polyps with high sensitivity and specificity, and to improve radiologists'polyp detection performance. The short-term goal of the project is to develop a high-performance CAD scheme for the automated detection of polyps in lfCTC to assist radiologists in detecting polyps quickly and accurately. We hypothesize that the CAD scheme will yield as high detection performance of polyps in lfCTC as that in cathartically cleansed CTC examinations. We also hypothesize that CAD will improve radiologists'performance in the detection of polyps in lfCTC and reduce inter-observer variability in the detection performance. To explore these hypotheses, we propose the following specific aims: Specific Aim 1: To expand the CTC database for development and evaluation of CAD and EC schemes Specific Aim 2: To develop pre-processing methods for CAD and EC schemes Specific Aim 3: To develop a CAD scheme for detection of polyps in lfCTC Specific Aim 4: To develop an advanced EC scheme for visualization of the colonic mucosa in lfCTC Specific Aim 5: Evaluate the benefit of the CAD scheme for lfCTC Successful development of the proposed CAD scheme for lfCTC will substantially advance the clinical implementation of CT-based colon cancer screening, promote the early diagnosis of colon cancer, lead to an increased screening rate, and ultimately reduce the mortality due to colon cancer.