Virtual colonoscopy or CT colonography (CTC) has been developed as an alternative colorectal screening method, but the method is controversial due to reports of sensitivity for patients with large ( > or = 10 mm) polyps from 55% to 94% in low prevalence populations.1 This apparent variability of CT colonography in colorectal cancer screening and surveillance is a major obstacle to wider acceptance and utilization. There are additional obstacles that intensify the controversy surrounding CTC concerning patient acceptance, the use of ionizing radiation, observer time and training requirements, difficulties in detecting flat adenomas or colon masses, and extracolonic findings.

Methods for computer-aided detection (CAD) in CTC have been developed to reduce the variation and improve the sensitivity for polyp detection while reducing interpretation time for observers. 2 Computer-based techniques are now available for detection of polyps and masses in CT colonography. 3


Figure 1. Virtual colonoscopy screenshot images from an interactive workstation (Philips Brilliance Workspace). (Click the image for a larger version.)

A recent meta-analysis systematically reviewed the test performance of CT colonography compared to colonoscopy or surgery based on English-language articles published between January 1975 and February 2005 listed in the PubMed, MEDLINE, and EMBASE databases and the Cochrane Controlled Trials Register. 4 Prospective studies of adults undergoing CT colonography after full bowel preparation, with colonoscopy or surgery as the gold standard, were selected. These studies used at least a single-detector row CT scanner with supine and prone positioning, insufflation of the colon with air or carbon dioxide, collimation less than 5 mm, and both two-dimensional and three-dimensional views during scan interpretation. The evaluators of the colonogram were unaware of the findings from use of the gold standard test; 33 studies provided data on 6,393 patients. The sensitivity of CT colonography was variable but improved as polyp size increased to 85% for polyps > 9 mm. Characteristics of the CT scanner, including width of collimation, type of detector, and mode of imaging, explained some of this heterogeneity. Specificity was 97% for polyps > or = 9 mm. 4

Computer-aided detection for CTC automatically detects the locations of suspicious polyps and masses and provides radiologists with a second opinion. 5 CAD has the potential to increase radiologists’ diagnostic performance in the detection of polyps and masses and to decrease variability of the diagnostic accuracy among readers without significantly increasing the reading time. Technical developments have advanced CAD substantially during the past several years, and a fundamental scheme for the detection of polyps has been established. The most recent CAD systems based on this scheme produce a clinically acceptable high sensitivity and a low false-positive rate. However, CAD for CTC is still under active development, and the technology needs to be improved further. 5


The challenges for CAD include:

  1. the methods focus on polyps while relatively little work has been done on colon masses,
  2. effect of CT scanning protocols,
  3. inability to use supine and prone data sets simultaneously,
  4. effect of fecal tagging and digital bowel cleaning,
  5. integration into clinical workflow and training databases.5

Commercially available CAD systems are in a relatively early stage of development, and incorporation of new capabilities developed in the laboratory is delayed due to the need for robust testing and FDA oversight of workstation software.

Computer-aided detection as a second reader has potential for improving polyp detection.6 The ColonCAD prototype was developed and tested on cases representative of the variability and quality in clinical practice. Results of a study with 150 patients provided sensitivity for polyps > or = 6 mm of 90% with a median false-positive rate of 3 per volume. 6


Reduction of the false-positive (FP) rate in CTC-CAD by surface-based measures on the inner colon wall has been achieved 7 with 100% sensitivity and slightly more than 3 FP/data set. Skilled observers can readily resolve most false-positive CAD results by adopting quality and consistency guidelines and applying software visualization tools incorporated in state-of-the-art workstations. Automated detection of the iliocecal valve can eliminate some false positives, 8 while region-based supine-prone correspondence eliminates many others. 9 Primary 3D endoluminal analysis versus primary 2D transverse analysis with computer-assisted reader software are used to visualize and evaluate candidate polyps identified by CAD. 10 The potential gain in using these tools for reading CTC examinations can be estimated by comparing polyp detection sensitivity identified by CAD (81%) with an average sensitivity of 70% for the expert reviewers in a selected test population of difficult cases. 11

Many clinical trials of CT colonography have been reported, but the apparent inconsistencies in results are difficult to resolve. To facilitate clear and consistent communication of CT colonography results, reports should describe the prevalence of polyps and masses in the cohort; the by-patient and by-polyp sensitivity, specificity, and positive and negative predictive values, each stratified by lesion size; the location of the lesion; histologic findings; and a detailed description of the method for performing the examination, interpreting the findings from the examination, and matching the CT and colonoscopic results. 12,13


Several CAD systems for CTC have been introduced as work-in-progress to demonstrate the potential of this technology and to assess its feasibility, prior to regulatory approval by the Food and Drug Administration (FDA). In general, computer-aided diagnosis software requires premarket approval by FDA, a standard that implies there is a body of evidence to show the technology is safe and effective. The requirements for approval of colon CAD will likely follow those used for lung cancer screening and mammography CAD systems. 14 Among the most important obstacles to CAD system evaluation in this process is the need for a test database consisting of CT data sets with gold standard verification of truth. 15 An effort similar to the Lung Imaging Database Consortium 16 to assemble a validated database of test cases is needed for CT colonoscopy to overcome this obstacle to widespread acceptance and use of CAD technology.

Virtual colonoscopy is developing into a practical clinical technique. The steep learning curve and variable accuracy of the technique are being addressed by improvements in patient preparation, scanning technique, reading methods, and CAD. 17 CTC is probably the best test for patients with an incomplete colonoscopy or for those who cannot undergo colonoscopy. Its precise role in screening average-risk patients for colon cancer remains to be defined by ongoing research and clinical trials. 18

Michael W. Vannier, MD, is professor, Department of Radiology, University of Chicago.


  1. van Gelder RE, Florie J, Stoker J. Colorectal cancer screening and surveillance with CT colonography: current controversies and obstacles. Abdom Imaging. 2005;30:5-12.
  2. Perumpillichira JJ, Yoshida H, Sahani DV. Computer-aided detection for virtual colonoscopy. Cancer Imaging. 2005;5:11-6.
  3. Yoshida H, Dachman AH. Computer-aided diagnosis for CT colonography. Semin Ultrasound CT MR. 2004;25:419-31.
  4. Mulhall BP, Veerappan GR, Jackson JL. Meta-analysis: computed tomographic colonography. Ann Intern Med. 2005;142:635-50.
  5. Yoshida H, Dachman AH. CAD techniques, challenges, and controversies in computed tomographic colonography. Abdom Imaging. 2005;30:26-41.
  6. Bogoni L, Cathier P, Dundar M, et al. Computer-aided detection (CAD) for CT colonography: a tool to address a growing need. Br J Radiol. 2005;78(Spec No 1):S57-62.
  7. Wang Z, Liang Z, Li L, et al. Reduction of false positives by internal features for polyp detection in CT-based virtual colonoscopy. Med Phys. 2005;32:3602-16.
  8. Summers RM, Yao J, Johnson CD. CT colonography with computer-aided detection: automated recognition of ileocecal valve to reduce number of false-positive detections. Radiology. 2004;233:266-72.
  9. Nappi J, Okamura A, Frimmel H, Dachman A, Yoshida H. Region-based supine-prone correspondence for the reduction of false-positive CAD polyp candidates in CT colonography. Acad Radiol. 2005;12:695-707.
  10. Taylor SA, Halligan S, Slater A, et al. Polyp detection with CT colonography: primary 3D endoluminal analysis versus primary 2D transverse analysis with computer-assisted reader software. Radiology. 2006 March 16 [Epub ahead of print].
  11. Taylor SA, Halligan S, Burling D, et al. Computer-assisted reader software versus expert reviewers for polyp detection on CT colonography. AJR Am J Roentgenol. 2006;186:696-702.
  12. Dachman AH, Zalis ME. Quality and consistency in CT colonography and research reporting. Radiology. 2004;230:319-23.
  13. Zalis ME, Barish MA, Choi JR, et al, and the Working Group on Virtual Colonoscopy. CT colonography reporting and data system: a consensus proposal. Radiology. 2005;236:3-9.
  14. Freedman M. Improved small volume lung cancer detection with computer-aided detection: database characteristics and imaging of response to breast cancer risk reduction strategies. Ann N Y Acad Sci. 2004;1020:175-89.
  15. Halligan S, Altman DG, Taylor SA, et al. CT colonography in the detection of colorectal polyps and cancer: systematic review, meta-analysis, and proposed minimum data set for study level reporting. Radiology. 2005;237:893-904.
  16. Armato SG 3rd, McLennan G, McNitt-Gray MF, et al, and the Lung Image Database Consortium Research Group. Lung image database consortium: developing a resource for the medical imaging research community. Radiology. 2004;232:739-48.
  17. Heiken JP, Peterson CM, Menias CO. Virtual colonoscopy for colorectal cancer screening: current status. Cancer Imaging. 2005;5(Spec No A):S133-9.
  18. Dachman AH, Yoshida H. Virtual colonoscopy: past, present, and future. Radiol Clin North Am. 2003;41:377-93.