Virtual Colonoscopy: A Novel Imaging Modality for Colorectal Cancer David T. Rubin, MD* and Abraham H. Dachman, MD†

Address *Section of Gastroenterology, Department of Medicine, University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637–1470, USA. E-mail: [email protected] †

Department of Radiology, University of Chicago. E-mail: [email protected]

Current Oncology Reports 2001, 3:88–93 Current Science Inc. ISSN 1523–3790 Copyright © 2001 by Current Science Inc.

Advances in computed tomography (CT) technology and computer capabilities have contributed to the development of a new imaging modality for colorectal lesions called CT colonography or virtual colonoscopy. Virtual colonoscopy is a rapid, minimally invasive scan of the cleansed and distended colon. Early work has demonstrated that this test is safe and well tolerated, and that it may be sensitive and specific enough to identify most significant precancerous or cancerous lesions. A number of technical and practical problems remain before virtual colonoscopy can be applied at a population level.

Introduction Despite the safety and effectiveness of current methods of colorectal cancer screening, compliance with screening recommendations is poor for a number of reasons, not the least of which is patient acceptance. Advances in CT technology and software have led to the development of a new colorectal imaging modality that may be more acceptable to patients. This test, called CT colonography (CTC) or virtual colonoscopy, provides multiple-perspective views of the entire colon. This review describes the current state of virtual colonoscopy, with particular emphasis on its use for colorectal cancer screening.

Colorectal Cancer: Need for Increased Screening Colorectal carcinoma is the second leading cause of death in the United States. In 1999, an estimated 129,400 new cases of cancer were reported, and more than 56,000 deaths occurred [1]. Although the incidence of colorectal cancer is increasing, mortality from the disease is decreasing, due in part to improved screening and prevention strategies as well

as improved therapies. Overall, however, compliance with screening remains disappointing. Colorectal cancer is a disease that is often preventable, and in its early stages it is easily treated. Screening strategies are designed to identify precancerous adenomas (polyps) before they transform into adenocarcinoma. Polyps at higher risk for transformation into adenocarcinoma include those that are larger in size and have tubulovillous histology. Therefore, secondary prevention of colorectal cancer involves detection and removal of polyps before they transform into adenocarcinoma [2]. Current methods of screening for colorectal cancer include fecal occult blood testing (FOBT), flexible sigmoidoscopy, double-contrast barium enema, and colonoscopy, but all suffer from some limitations. FOBT and flexible sigmoidoscopy are limited by incomplete views of the colon and low rates of sensitivity and specificity. Both barium enema and colonoscopy offer complete views of the colon, but they are limited by the need for a cathartic bowel cleansing, a small but important risk of perforation, and fear of discomfort by patients. Colonoscopy also requires sedation and is the most expensive screening test, but it offers the advantages of biopsy and polypectomy and is considered the gold standard for colorectal cancer screening. Consensus is growing that colon cancer screening should involve a view of the entire colon. Two recently published studies have supported the use of colonoscopy over flexible sigmoidoscopy for colorectal screening [3•, 4]. National screening recommendations support this approach as well. The 1997 American Gastroenterology Association guidelines offer colonoscopy or barium enema as options for total colonic evaluation in average-risk individuals (those who are asymptomatic and aged 50 years or older) [5]. Most recently, the American College of Gastroenterology proposed colonoscopy every 10 years as the “preferred” examination for screening average-risk patients [6•]. Despite these guidelines, Medicare and most insurers currently only offer reimbursement for flexible sigmoidoscopy and FOBT in average-risk patients. Despite the evidence and these recommendations, many physicians do not offer colorectal cancer screening, and less than 40% of eligible patients have ever undergone fecal occult blood tests [7,8]. A number of reasons are cited for this persistently poor rate of screening. Polyps and early colorectal cancer are asymptomatic and therefore may be forgotten by busy physicians or patients seeking treatment for

Virtual Colonoscopy • Rubin and Dachman

89

Virtual colonoscopy or CT colonography uses spiral CT data to create standard axial and reformatted twodimensional (2D) or three-dimensional (3D) images of the colon. In 1994, Vining and Gelfand [9] piqued the interest of the gastrointestinal community with the first report of this technique, in which they used volumetric CT data to create three-dimensional endoluminal images of the colon (Figs. 1, 2, and 3).

Most centers now use the newer multi-slice spiral CT scanners. This technology permits the rapid collection of volumetric data from the abdomen and pelvis in less than 30 seconds. Patients are now able to hold their breath during the entire scan and therefore minimize motion artifact. In addition, scans are obtained with the patient in supine and prone positions to allow shifting of gravity-dependent material and minimize errors in interpretation [11]. A number of investigators have described their results with differing CT scanning parameters. The goal of these parameters is to obtain a maximum sensitivity rate with the least amount of data collection and a safe amount of radiation exposure. The best image quality appears to be obtained with a collimation of 5 mm, pitch of 1 to 2, and reconstruction intervals of 1 mm to 3 mm. At 70 mA with these settings, the radiation dose administered is substantially less than that required for standard body CT settings, and approximately 20% less than the standard films obtained during a double-contrast barium enema [12].

Technique The technique by which virtual colonoscopy is performed continues to evolve, but the basic approach has become fairly uniform at major centers (Table 1). The technical performance of this imaging modality relies on a combination of careful bowel preparation, dedicated hardware for data collection, and specialized software for data manipulation and observer interpretation. Thorough bowel preparation is required for an adequate view of the gas-distended colon. Because stool or liquid may complicate interpretation of virtual colonoscopy, investigators have attempted to minimize these sources of error with a cathartic bowel preparation similar to that used for conventional colonoscopy (eg, a standard polyethylene glycol electrolyte solution). Next, insufflation of the colon is necessary in order to distinguish collapsed loops of bowel from polyps or masses. A flexible-tip catheter is inserted into the rectum, and room air or carbon dioxide is infused either to a set volume (1.5 to 2 L) or to patient discomfort. Our approach is to use room air and insufflate the colon to minimal discomfort of the patient. Scout films are obtained to assess adequacy of bowel distention before scanning is performed. Despite these rather imprecise methods of bowel distention, no complications have been reported to date, and image quality is quite good due to the high contrast between the air-filled colon and the soft tissue of the colonic wall. Next, glucagon is administered intravenously to limit bowel spasm, which is a particular problem in the sigmoid colon. This minimizes patient discomfort and provides better images, although glucagon administration may also result in ileocecal valve relaxation and unwanted refluxing of air into the small bowel. Only one study has examined the effect of glucagon on image acquisition [10].

Three-dimensional versus two-dimensional images Early virtual colonoscopy techniques attempted to simulate conventional colonoscopy perspective with endoluminal views and navigated “fly throughs.” There are two different techniques for constructing 3D images: volume rendering and surface rendering. Volume rendering allows multiplanar display of extraluminal soft tissues and attenuation data but is computationally more demanding, expensive, and time-consuming than surface rendering. Surface rendering eliminates the extraluminal data and uses only the endoluminal surface data, thereby offering somewhat limited information but requiring less computer resources. Advances in computational speed have made volume rendering more available, but optimal processing parameters are not yet clarified [13,14]. In order to “navigate” through the 3D endoluminal images and “fly through” the colon, a centerline must be calculated. This allows the viewer to traverse the colon while inspecting pathology without focusing on navigation. However, poor data quality can substantially distort 3D images and result in artificial floating “debris.” The difficulties with this approach for primary reading of CTC have supported the primary use of 2D images. Two-dimensional images are the standard with which radiologists are most familiar. Numerous investigators have demonstrated that 2D images (usually standard axial images) are as effective at polyp detection as 3D endoluminal constructions [15,16]. In addition, navigation through 2D images is more rapid than 3D reconstruction and navigation [17•]. Optimal 2D images require lung window settings, but soft-tissue windows may be necessary to assess areas of bowel collapse, bowel wall thickening, and incidental extra-colonic findings. Lipomas are also distinguished more clearly with soft-tissue windows.

other problems. Patients dislike the bowel cleansing necessary for the imaging tests, and there is a great deal of embarrassment and fear of discomfort from the examinations. Recognition of the benefit of colonoscopy for colorectal cancer screening, combined with the currently poor rate of screening compliance, has led researchers to seek a total colonic examination that is minimally invasive and fast and offers increased patient acceptance. Virtual colonoscopy is a novel imaging modality that may address these needs.

Virtual Colonoscopy

90

Gastrointestinal Cancers Figure 1. A represents the supine two-dimensional axial view of the distended colon. A small abnormality is seen (arrow), but it is unclear if this represents a polyp or retained fecal matter. B is a three-dimensional construction of the data simulating the endoluminal view of the colon, showing the abnormality as a polyp (arrow). More proximally in this colon view is a normal fold (also seen on the inset axial perspective).

Figure 2. A, Axial soft-tissue windows suggest a possible mass or retained stool in the rectum (arrow). B, Endoluminal reconstruction demonstrates partially circumferential mass (arrows).

A significant advantage of CTC is that data interpretation occurs after the examination is completed. Most centers now employ user-friendly workstations that allow easy scrolling between cine-loop 2D images and 3D images-on-demand. Several independent companies and academic centers are developing more specialized software for CTC. At the University of Chicago, a workstation has been formatted to provide multiple axial views of the colon simultaneously and scroll through a cine loop of the supine images, followed by a cine loop of the prone images. Endoluminal and multiplanar reconstructions are sparingly used to distinguish polyps and masses from colonic folds or retained stool [18]. Other centers describe a similar approach, and studies have shown better sensitivity and specificity for polyp detection using this method than with 2D or 3D images alone [16]. Various investigators now report reading times of approximately 10 to 15 minutes per examination.

Performance Total colonic examination Total colonic examination has been studied at a number of centers to evaluate sensitivity and specificity for polyp detection. Although most investigators have targeted polyps of 1 cm or larger, some investigators believe that, in order for virtual colonoscopy to be widely applicable, it should be sensitive for polyps of 5 mm or more [19].

Hara et al. [16] performed the first blinded prospective trial of virtual colonoscopy in 1996. Seventy patients underwent supine-only CTC followed by conventional colonoscopy. Sensitivity for lesions greater than or equal to 1 cm was 75%, and specificity was 90%. Smaller lesions resulted in lower rates of sensitivity and specificity. At the University of Chicago, 44 patients were studied with supine and prone scanning, and a sensitivity rate of 83% was achieved for polyps greater than or equal to 1 cm, and 66% for polyps of 6 mm to 9 mm. Small studies from other centers have reported sensitivity for polyps of 1 cm or more of between 75% and 100%, and specificity of 86% to 100% [10,18,20]. Sensitivity for polyps that are 6 mm to 9 mm in size has remained disappointingly low. Most recently, Fenlon et al. [21••] described a blinded prospective trial of 100 patients, comparing virtual to conventional colonoscopy. In their study, 100 patients at high risk for colorectal neoplasia underwent virtual colonoscopy prior to conventional colonoscopy. They identified three of three cancers and reported sensitivity rates of 91% for polyps of 1 cm or more, 82% for polyps of 6 mm to 9 mm, and 55% for polyps less than or equal to 5 mm [21••]. Although the results of these early studies are encouraging, no study has yet been performed on an average-risk screening population. The effectiveness of this technology needs to be examined in average-risk individuals before

Virtual Colonoscopy • Rubin and Dachman

91

colon in 26 of 29 patients, compared with preoperative barium enema, which failed to show the proximal colon in any of the four patients examined [24]. It has been suggested that the unique nature of CTC images may provide information not only about the intralumenal lesions but also about the stage of a colorectal cancer. Morrin et al. [25] studied 34 patients with colorectal cancers and obstructing colorectal lesions and found that CTC correctly staged 13 of 16 cancers and identified 16 of 17 synchronous polyps. Ninety-seven percent of the colonic segments were visualized, compared with only 60% with barium enema. In addition, CTC correctly identified the surgical anastomosis of nine patients, but local tumor recurrence could not be distinguished from surgical changes in one patient.

Figure 3. A supine axial view of the abdomen demonstrating retained fluid (arrows), which may obscure abnormalities and create falsenegative results. A subsequent prone scan would allow shifting of this fluid and improved interpretation.

widespread application can occur. A multicenter trial to assess asymptomatic patients is now underway. CTC versus barium enema Early reports of the efficacy of virtual colonoscopy have compared this imaging modality with barium enema. The reported performance of CTC for polyp detection is better than that previously described for barium enema, although only one specific study has yet reported a direct comparison of these tests [22]. However, several studies have demonstrated that CTC is superior to barium enema for viewing colonic segments proximal to obstructing lesions [23•,24,25]. CTC for incomplete colonoscopy Virtual colonoscopy is the application used in cases of incomplete colonoscopy due to technical difficulties, patient discomfort, or obstructing lesions. Morrin et al. [23•] reported their experience with 40 patients in whom the cecum could not be reached during routine colonoscopy. CTC revealed 96% of all colonic segments and identified the probable reason for the incomplete colonoscopy in 74% of the cases. A similar study by Fenlon et al. [24] evaluated the utility of CTC in 29 patients with occlusive colon carcinoma. These authors found that CTC identified all 29 occlusive carcinomas and found two cancers and 24 polyps in the proximal colon. Postoperative conventional colonoscopy in 12 patients confirmed 16 polyps. Virtual colonoscopy successfully showed the entire proximal

Extra-intestinal findings Because data collection for virtual colonoscopy involves CT scanning of the abdomen and pelvis, a proposed benefit of this examination is the ability to identify clinically significant extra-intestinal abnormalities. In fact, occurences of incidental findings are more likely than with polyps of significant size. Hara et al. [26•] reported the extraintestinal findings in 264 patients who underwent CTC. Half of these patients had an abnormality identified, and 11% had “highly significant” findings (abdominal aortic aneurysm, renal adenocarcinoma, and inguinal hernia with bowel). Investigators remain excited by the possibility of identifying these extra-intestinal lesions, but outcomes and cost-analysis studies of the impact of these “incidentalomas” need to be performed. In addition, the potential for finding such unintended abnormalities should be included in the patient consent process.

Problems and Pitfalls Despite the major advances that have occurred with the practice of CTC, a number of problems and pitfalls remain. Although the learning curve for CTC has not been defined, it is understood that interpretation errors can be minimized with experience and the application of the common principles of barium radiography. In fact, the levels of experience of the investigators may be one explanation for the differences in sensitivity and specificity among the published virtual colonoscopy studies. False-negative results may be caused by retained fluid or fecal material as well as by collapsed portions of bowel. These problems are often addressed but not eliminated by the addition of prone scanning and shifting of gravity-dependent material [11,27]. In addition, the balloon cuff on the tip of the insufflation catheter may obscure a distal rectal lesion. A careful digital rectal examination prior to placement of the rectal catheter has become common practice. Other investigators have proposed using intravenous contrast to better distinguish polyps or masses from stool (Vining D, Presentation

92

Gastrointestinal Cancers

Table 1. Virtual colonoscopy: University of Chicago technique Identification of patient (average risk = asymptomatic, age >50 years) Bowel preparation (polyethylene glycol ) Glucagon, 1 mg intravenous Digital rectal examination Placement of rectal tube; insufflation until minimal patient discomfort Scout film to assess bowel distention Breath hold and supine scanning of patient with multi-slice spiral CT (~ 20 seconds) (helical technique: 5 mm collimation; pitch, 1.5; reconstruction index, 2.5 mm) Reposition patient; additional insufflation Breath hold and prone scanning of patient with multi-slice spiral CT (~ 20 seconds) Patient discharged to colonoscopy (if in study) Radiologist reading 2D axial images of supine scan in cine loop 2D axial amages of prone scan in cine loop Multiplanar reconstructions and 3D problem-solving for unclear lesions Recording and reporting of data

at the Medical University of South Carolina, 1999). A common error in the right colon is to mistake a polyp for the ileocecal valve. In addition, some villous adenomas may appear as stool, due to their surface architecture. Flat adenomas and infiltrating cancers need to be distinguished from collapsed bowel folds, and this may be accomplished by comparison with adjacent folds. False-positive readings occur from misinterpretations of folds or stool. This error may be minimized by recognition of the fact that retained stool appears heterogeneous on CT imaging because of a mixture of air and fecal material. Stool also does not have an obvious attachment to the bowel wall when prone images are obtained [27]. The problem of falsepositive readings is less dangerous (although not less expensive) than that of false-negative readings, because patients with false-positive errors will undergo colonoscopy, and presumably, the error will be corrected.

Unsettled Issues and Future Applications A number of areas require clarification before virtual colonoscopy can reach its potential. Performance in screening populations needs to be evaluated. Even missing a small percentage of clinically significant adenomas may be unacceptable on a population screening level. Researchers are developing methods of automated polyp detection to minimize reading time and potentially increase accuracy [28]. Detailed cost and resource analysis is needed, including cost and outcome analysis for work-up of incidental findings. A published model by Sonnenberg et al. [29•] assumed a hypothetical population of 100,000 people who had screening virtual colonoscopy every 10 years and underwent conventional colonoscopy for abnormalities, with surveillance examinations subsequently. Their model suggests that, in order for virtual colonoscopy to have cost effectiveness similar to that of colonoscopy, it needs to be 54% less expensive or have compliance rates

15% to 20% higher than conventional colonoscopy (even if it is 100% sensitive). CTC remains unproven in patients in high-risk groups, such as those with inherited forms of colorectal cancer. Patients with inflammatory bowel disease have an increased risk of developing colorectal cancer, but it is via dysplasia of (flat) inflamed mucosa, and CTC is unlikely to offer an alternative to conventional colonoscopy surveillance and biopsies. Early work has demonstrated that patients may prefer the minimally invasive CTC to barium enema or conventional colonoscopy [30,31]. Current CTC methods do not eliminate the need for a bowel preparation, which is described by some patients as the worst part of colorectal cancer screening tests. In fact, if the CTC shows positive findings, an additional bowel cleansing will then be required for the subsequent conventional colonoscopy. Institutions will need to offer immediate image interpretation and subsequent colonoscopy for positive findings to eliminate the unpleasantness of requiring a repeat bowel preparation. Physicians and patients alike eagerly await development of a “virtual preparation.” Early studies suggest that contrast agents or magnetic resonance (MR) colonography may hold promise in this area but do not yet offer comparable sensitivity to the cleansed colon [32•].

Conclusions Virtual colonoscopy offers a minimally invasive view of the entire colon that may improve population compliance with colorectal cancer screening. The current approach to this exciting technology appears to be safe, well tolerated, and sensitive for detection of polyps or masses larger than 1 cm. In addition, the current approach may allow preoperative cancer staging and identification of synchronous lesions in incomplete or obstructed colonoscopy. Routine use of CTC for screening purposes is limited by the absence of population studies, unclear cost and resource allocation, an

Virtual Colonoscopy • Rubin and Dachman

undefined learning curve for radiologists, and possibly by unacceptable sensitivity for polyps that are 6 mm to 9 mm. Advances in technology and ongoing research hold much promise for the widespread use of this test in the future.

References and Recommended Reading Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance 1.

Landis SH, Murray T, Bolden S, Wingo PA: Cancer statistics, 1999 [see comments]. CA Cancer J Clin 1999, 49:8–31. 2. Winawer SJ, Zauber AG, Ho MN, et al.: Prevention of colorectal cancer by colonoscopic polypectomy: the National Polyp Study Workgroup [see comments]. N Engl J Med 1993, 329:1977–1981. 3.• Lieberman DA, Weiss DG, Bond JH, et al.: Use of colonoscopy to screen asymptomatic adults for colorectal cancer: Veterans Affairs Cooperative Study Group 380 [see comments]. N Engl J Med 2000, 343:162–168. This paper provides recent data demonstrating that colonoscopy in average-risk individuals will identify adenomas missed by flexible sigmoidoscopy. The authors offer support for total colonic examination in colorectal cancer screening. 4. Imperiale TF, Wagner DR, Lin CY, et al.: Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 2000, 343:169–174. 5. Winawer SJ, Fletcher RH, Miller L, et al.: Colorectal cancer screening: clinical guidelines and rationale [see comments]. Gastroenterology 1997, 112:594–642. 6.• Rex DK, Johnson DA, Lieberman DA, et al.: Colorectal cance prevention 2000: screening recommendations of the American College of Gastroenterology. Am J Gastroenterol 2000, 95:868–877. This publication contains the most recent recommendations for colorectal cancer screening. The authors advocate colonoscopy as the “preferred” screening modality in average-risk individuals. 7. Centers for Disease Control and Prevention: Screening for colorectal cancer: United States, 1997. JAMA 1999, 281:1581–1582. 8. Vernon SW: Participation in colorectal cancer screening: a review [see comments]. J Natl Cancer Inst 1997, 89:1406–1422. 9. Vining D, Gelfand D: Noninvasive colonoscopy using helical CT scanning, 3D reconstruction, and virtual reality. Proceedings of the Society of Gastrointestinal Radiologists, 1994. 10. Yee J, et al.: Colonic distention and colorectal polyp detection with and without glucagon on virtual colonoscopy. Proceedings of the Society of Gastrointestinal Radiologists, 1999. 11. Chen SC, Lu DS, Hecht JR, Kadell BM: CT colonography: value of scanning in both the supine and prone positions. AJR Am J Roentgenol 1999, 172:595–599. 12. Hara AK, Johnson CD, Reed JE, et al.: Reducing data size and radiation dose for CT colonography. AJR Am J Roentgenol 1997, 168:1181–1184. 13. Reed J: Display/navigation: Mayo Clinic experience. Proceedings of First International Symposium on Virtual Colonoscopy, 1998. 14. McFarland EG, Brink JA, Loh J, et al.: Visualization of colorectal polyps with spiral CT colography: evaluation of processing parameters with perspective volume rendering. Radiology 1997, 205:701–707. 15. Royster AP, Fenlon HM, Clarke PD, et al.: CT colonoscopy of colorectal neoplasms: two-dimensional and three- dimensional virtual-reality techniques with colonoscopic correlation. AJR Am J Roentgenol 1997, 169:1237–1242.

16.

93

Hara AK, Johnson CD, Reed JE, et al.: Colorectal polyp detection with CT colography: two- versus three-dimensional techniques. Work in progress [see comments]. Radiology 1996, 200:49–54. 17.• Macari M, Milano A, Lavelle M, et al.: Comparison of timeefficient CT colonography with two- and three-dimensional colonic evaluation for detecting colorectal polyps. AJR Am J Roentgenol 2000, 174:1543-1549. This report supports the combined use of 2D and 3D images for efficient reading in virtual colonoscopy. 18. Dachman AH, Kuniyoshi JK, Boyle CM, et al.: CT colonography with three-dimensional problem solving for detection of colonic polyps. AJR Am J Roentgenol 1998, 171:989–995. 19. Rex D: Virtual colonoscopy. In Digestive Disease Week. San Diego, CA. May 2000. 20. Hara AK, Johnson CD, Reed JE, et al.: Detection of colorectal polyps with CT colography: initial assessment of sensitivity and specificity. Radiology 1997, 205:59–65. 21.•• Fenlon HM, Nunes DP, Schroy PC III, et al.: A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps [see comments]. N Engl J Med 1999, 341:1496–1503. Report from a blinded prospective trial in 100 high-probability polyp patients that shows high sensitivity and specificity for polyps greater than 1 cm. 22. Hara AK, Johnson CD, Reed JE, et al.: Computed tomographic colography (virtual colonoscopy) for polyp detection: early comparison against barium enema [abstract]. Gastroenterology 1997, 112(suppl):A575. 23.• Morrin MM, Kruskal JB, Farrell RJ, et al.: Endoluminal CT colonography after an incomplete endoscopic colonoscopy. AJR Am J Roentgenol 1999, 172:913–918. Demonstrates the utility of virtual colonoscopy (VC) in cases of incomplete colonoscopy, indicating that VC is better than barium enema. 24. Fenlon HM, McAneny DB, Nunes DP, et al.: Occlusive colon carcinoma: virtual colonoscopy in the preoperative evaluation of the proximal colon. Radiology 1999, 210:423–428. 25. Morrin MM, Farrell RJ, Raptopoulos V, et al.: Role of virtual computed tomographic colonography in patients with colorectal cancers and obstructing colorectal lesions. Dis Colon Rectum 2000, 43:303–311. 26.• Hara AK, Johnson CD, MacCarty RL, Welch TJ: Incidental extracolonic findings at CT colonography. Radiology 2000, 215:353–357. Reports the incidental findings seen on virtual colonoscopy. 27. Fletcher JG, Johnson CD, MacCarty RL, et al.: CT colonography: potential pitfalls and problem-solving techniques [see comments]. AJR Am J Roentgenol 1999, 172(5):p. 1271-8. 28. Summers RM, Beaulieu CF, Pusanik LM, et al.: Automated polyp detector for CT colonography: feasibility study. Radiology 2000, 216:284–290. 29.• Sonnenberg A, Delco F, Bauerfeind P: Is virtual colonoscopy a cost-effective option to screen for colorectal cancer? Am J Gastroenterol 1999, 94:2268–2274. Although hypothetical, this paper offers some important insights into application of virtual colonoscopy technology at a population level. 30. Farrell RJ, Morrin MM, Silas A, et al.: Virtual colonoscopy in patients undergoing elective colonoscopy: diagnostic accuracy and patient tolerance [abstract]. Gastroenterology 2000, 118:A258. 31. Hara AK, Johnson CD, Reed JE: Colorectal polyp detection with CT colography (virtual colonoscopy): a blinded prospective study [abstract]. Radiology 1996, 201:A252. 32.• Weishaupt D, Patak MA, Froehlich J, et al.: Faecal tagging to avoid colonic cleansing before MRI colonography. Lancet 1999, 354:835–836. The authors suggest the possible use of “virtual bowel preparation” in the future and explore the possibility of MR colonography.