Early diagnosis of colon cancer with computer-aided multi-energy CT colonography

Massachusetts General Hospital, Boston MA

Investigators

Linked publications, trials & patents

Abstract

DESCRIPTION (provided by applicant): Colon cancer, the second leading cause of cancer deaths for men and women in the United States, is preventable if precursor colonic lesions are detected and removed. Computed tomographic colonography (CTC), also known as virtual colonoscopy (VC), could substantially increase the capacity, safety, and patient compliance of colorectal examinations. CTC has been endorsed as a viable option for colorectal cancer screening by the recent guidelines of the American Cancer Society, U.S. Multi-Society Task Force, and the American College of Radiology. Laxative-free fecal-tagging CTC assisted by computer-aided detection (CAD) is an emerging CTC technique for eliminating of laxative agents or diarrhea-inducing high-osmolar contrast agents from bowel cleansing. A recent large-scale multi-center clinical trial suggests that the combination of laxative-free preparation with ingestion of oral contrast agent and CAD is feasible in making CTC examinations easy for patients to tolerate while detecting polyps e10 mm in size at sensitivity comparable to that of optical colonoscopy (OC). However, non-polypoid flat lesions and small polyps (6-9 mm in size) represent a significant source of false-negative studies. Thus the goal of this project is to develop a novel CAD scheme based on meCTC, denoted as a multi-energy CAD (meCAD) scheme, which overcomes the limitations of the CAD schemes based on single-energy CTC. We hypothesize that the meCAD scheme will (1) yield a high performance in the detection of both polypoid and non-polypoid colorectal lesions comparable to OC, and (2) improve radiologists' performance in the detection of clinically significant colorectal lesions in laxative free meCTC images at an ultra-low-dose. Our specific aims are (1) Establish a clinical ultra-low-dose meCTC image database for development and evaluation of a meCAD scheme; (2) Develop a novel meCAD scheme that effectively detects colorectal lesions in ultra-low-dose laxative-free meCTC images; (3) Evaluate clinical benefit of the meCAD scheme. Successful development and validation of the meCAD scheme will provide radiologists with an accurate and reliable non-invasive CTC screening scheme for early detection and diagnosis of colorectal lesions to prevent the occurrence of colorectal cancer.