J Gastroenterol 2002; 37:23–28

CT and 99mTc-WBC vs colonoscopy in the evaluation of inflammationand complications of inflammatory bowel diseases

Martin Charron1, Carlo Di Lorenzo

2, and Samuel Kocoshis

2

1 Division of Nuclear Medicine, Department of Radiology, Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard,Philadelphia, PA 19104, USA2 Department of Gastroenterology, Children’s Hospital of Pittsburgh and University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Introduction

The exact role of computerized tomography (CT) inthe investigation of IBD is controversial. CT is recom-mended by some1 as an important tool in the evaluationof patients with suspected colonic inflammation. CTis also believed to be important in the diagnosis andmanagement of inflammatory bowel diseases (IBD) inpatients presenting with acute symptoms,2 as a manage-ment tool in the evaluation of patients with knownIBD,3 presenting with exacerbation of their disease,2,4

and as an essential diagnostic tool in the acutely illpatient with Crohn’s disease (CD).5,6 Finally, CT hasbeen recommended as an initial imaging study in chil-dren with known CD and a changing pattern of clinicalsymptoms.3

For others,4 the role of CT is limited to the assessmentof mural disease and its effect on luminal diameter,and the differential diagnosis of mesenteric disease. Stillothers have noted that there is a disturbing lack ofcorrelation between radiographic findings and the clini-cal severity and activity of the disease.4,7 For some, CTwas not useful for detecting early CD.5,8 Failure to ad-equately opacify the bowel wall may lead to scans thatare difficult to interpret, or to the overestimation ofbowel wall thickening.5 Abnormal bowel thickness canalso be seen in cirrhosis, lymphoma, infection, diverticu-litis, pancreatitis, edema, hypoalbuminemia, ischemia,Henoch-Schonlein purpura, and bleeding in the bowelwall.8,10 In one recent report, abnormal wall thicknesswas seen in 7% of controls.10 The presence of mesen-teric masses is not a helpful sign for differentiating vari-ous disease processes.9 In one series, 83% of childrenwith bowel thickness greater than 1cm had neoplasticdisease.9 Tomei et al.11 evaluated the correlation of CTfindings with laboratory indices of inflammation in pa-tients with CD. All patients had abnormal CT findings.One-fourth of the patients with abnormal bowel wallthickness had normal laboratory values, indicating that

Background. The goal of this study was to evaluate theaccuracy of computerized tomography (CT) and 99mTc-white blood cell (WBC) scintigraphy versus colono-scopy in assessing inflammatory bowel diseases (IBD)in a large population of children. Methods. In a patientpopulation of 313 consecutive children who had a 99mTc-WBC scan, 106 colonoscopies were done within a me-dian time interval of 8 days of the 99mTc-WBC scan. Onehundred and three CT scans were performed on 84patients. Results. Of the 42 CT scans obtained within ashort time interval after the 99mTc-WBC scan, 21 (50%)were normal. In the 21 children with a positive 99mTc-WBC scan, 62% (13/21) of the CT scans underestimatedthe bowel wall inflammation in at least one segment. Inthe children with a negative 99mTc-WBC study, therewere 17 negative CT examinations and 4 examinationsshowing an abnormal terminal ileum. When CT wascompared with colonoscopy in assessing inflammation,there were five true-negative CT, two true-positive CT,no false-positive, and seven false-negative CT examina-tions. When 99mTc-WBC scintigraphy was comparedwith colonoscopy in assessing inflammation, therewere seven true-positive, two false-negative, five true-negative, and no false-positive 99mTc-WBC studies. The99mTc-WBC scan was positive in five patients with afalse-negative CT examination. Of the total 103 CTscans obtained, 53 (51%) were normal. Four abscesses(3.8%) were demonstrated by CT. Conclusions. 99mTc-WBC scintigraphy is more sensitive than CT for detect-ing inflammation of the bowel wall. The incidence ofcomplications from IBD in this retrospective study wasmuch lower than had been previously reported.

Key words: inflammatory bowel disease, Crohn’s dis-ease, computerized tomography, 99mTc-WBC

Received: February 26, 2001 / Accepted: June 6, 2001Reprint requests to: M. Charron

24 M. Charron et al.: CT and 99mWBC vs colonoscopy in IBD

the disease findings were quiescent, and yet the CTfindings were abnormal. An abnormal layer of submu-cosal attenuation was seen in all patients, all withoutevidence of active disease. Others have also reportedthat patients with quiescent disease may have abnormalwall thickening on CT.12 A prominent submucosal layerof decreased attenuation on CT is not specific,13 and canbe caused by fat deposition or active inflammation.Additionally, CT cannot differentiate thickening causedby edema and cellular infiltration from that caused byfibrosis.14 Gossios et al.15 demonstrated that bowel wallthickening, the most common pretreatment abnormal-ity, may remain abnormal for many years after treat-ment.15 The presence of a “target” or halo sign is notspecific for IBD.5

The goal of this study was to compare, in a largepopulation of children with suspected or known IBD,the accuracy of CT and 99mTc-white blood cell (WBC)scintigraphy versus colonoscopy in assessing bowel in-flammation. A second objective of this study was toevaluate the incidence of complications detected byeach modality.

Patients and methods

Patients

Over a 6-year period, 313 consecutive 99mTc-WBC stud-ies were performed, and the hospital charts of thesechildren were reviewed (Table 1). There were 144 boysand 169 girls (average age, 13 years). Three groups ofchildren were evaluated: (a) 132 children with knownIBD16 who had a 99mTc-WBC scan for assessment of aclinical flare; (b) 130 children who had a 99mTc-WBCscan to exclude IBD; the final diagnoses (based oncolonoscopy and biopsies) in this latter group were CD

in 27, ulcerative colitis (UC) in 9, miscellaneous colitis(MC) in 13 (7 indeterminate colitis [IC], 4 infectiouscolitis, and 2 autoimmune colitis), normal in 39 (normalcolonoscopy), and probably normal (PN) in 42 (childrenwith no evidence of IBD by clinical follow-up and radio-graphic study and no colonoscopy.); and (c) the thirdgroup, who consisted of 51 controls (NL) who had un-dergone 99mTc-WBC scanning for other medical prob-lems (final diagnoses: osteomyelitis, cellulitis, trauma,myocarditis, fever of unknown origin, painful hip, andperitonitis).

Colonoscopy

Total colonoscopy, using an Olympus CF100TL orPCF20 instrument (Olympus Optical, Tokyo, Japan)was performed to assess the extent and activity ofthe mucosal inflammatory changes in eight intestinalsegments, defined below. The endoscopic findingsof inflammatory activity were classified as absent(noninflamed mucosa), mild (granularity, edema, invis-ible vascular pattern), moderate (hyperemia, friability,and all features of mild activity), or severe (ulceration,in addition to features of moderate activity).

One hundred and six colonoscopies were done,within an average time interval of 14 days (median, 8days) of the 99mTc-WBC scan. One hundred and threeCT scans were performed on 84 patients. Forty-two CTscans were done within an average time interval of 8days of the 99mTc-WBC scan. Fourteen patients had thethree examinations (CT, 99mTc-WBC, and colonoscopy)performed at mean time intervals of 11 days, and threeadditional patients had colonoscopy and CT performedwithin a short time interval.

In children with no recent colonoscopy, the findingsof the 99mTc-WBC scan were compared with findings onlong-term clinical follow-up, laboratory values, and withthe gastroenterologist’s initial clinical assessment. Theaverage elapsed time between 99mTc-WBC scintigraphyand the clinical follow-up was 285 days.

Performance of the scans

Labeling of leukocytes with Tc99m has been described.17

At 0.5–1h after injection, imaging was done with a largefield of view (LFOV) gamma camera fitted with a low-energy, all-purpose collimator (SMV, Twinsburg, OH,USA). Anterior and posterior 8-min images of the ab-domen and pelvis were recorded in analog and digitalform. Pelvic outlet views were also obtained to distin-guish bladder activity from rectal activity. Anteriorviews of the abdomen, with the patient standing, wereobtained to separate the liver from the transverse colon.Two to four hours after injection, an 8-min anteriorview of the abdomen was repeated. Single photon emis-

Table 1. Patient data

Number of children 313Male 144Female 169Average age (years) 13Known IBD 132Suspected IBD 130Controls 51Number of CT scans 103Number of colonoscopies 106Number of 99mTc-WBC scans 313Final diagnosis

Crohn’s disease 119Ulcerative colitis 38Controls 51Probably normal 80Indeterminate colitis 25

IBD, Inflammatory bowel diseases; CT, computerized tomography;WBC, white blood cell

M. Charron et al.: CT and 99mWBC vs colonoscopy in IBD 25

sion computed tomography (SPECT) images were ac-quired on a double-headed camera (SMV, Twinsburg,OH, USA) at 2–4 h, with the following acquisitionparameters: 30 s/stop, 60 stops, 120 projections, 1.6 mag,pre-filter Hanning-Nyquist-1.515, Cutoff 1.315,rRampfilter-Nyquist 1.515, cutoff 1.31. SPECT images wereobtained in 97 children. SPECT was not done in thecontrols. Volume rendered images, using the maximumactivity projection (MAP) technique,16 were also de-rived from the SPECT data. All images were reviewedon a workstation with ad-lib manipulation of intensityscale, contrast, color scale, and cine display rotationalspeed. When scan findings were equivocal, delayed im-aging was performed at 6–8h and, rarely, at 24 h. Thiswas occasionally useful to differentiate active CD fromlate accumulation of 99mTc-WBC in the right lowerquadrant. Cinematic representations of MAP imagesof 99mTc-WBC are available for review at: http://www.arad.upmc.edu/users/charron/index.htm.

Analysis of the scans

In each set of scans (0.5–1h and 2–3 h) from the 313patients, the bowel was divided into eight segments(rectum, sigmoid colon, descending colon, transversecolon, ascending colon, cecum, terminal ileum, andsmall bowel), resulting in 5008 bowel segments for scor-ing (8 segments � 2 images � 313 patients � 5008).The inflammatory activity in each segment was gradedsemiquantitatively by comparing the uptake in thebowel with that in the iliac crest bone marrow and liver:grade 0, no activity; grade 1, activity less than that iniliac crest; grade 2, activity similar to that in iliac crest;grade 3, activity greater than that in iliac crest; grade 4,activity equal to that in liver; grade 5, activity greaterthan that in liver; grade 6, activity equal to that inspleen. The grade of uptake on the 99mTc-WBC scanswas then compared with the colonoscopy result. Onenuclear physicianm, who was blinded to the clinical de-tails, interpreted all images. The scintigraphic result wasconsidered a false-negative if there was a difference inthe degree of inflammation between the 99mTc-WBCfindings and the colonoscopy result, i.e., a 99mTc-WBCscan with mild inflammation (grade 1 or 2) and severeinflammation on colonoscopy was considered a false-negative. The scintigraphic result was interpreted as afalse-positive if it revealed uptake of 99mTc-WBC in anarea that was normal on colonoscopy and biopsy. Activ-ity seen at a colostomy or ileostomy site was excludedfrom analysis.

Statistical analysis

The CT findings in each segment were compared withthe 99mTc-WBC scan findings and the endoscopic assess-

ment. Sensitivity, specificity, positive predictive value,negative predictive value, and diagnostic accuracy werecalculated, with histological assessment chosen as thereference method when specified.

Results

Detection of IBD complications by CT

This subgroup consisted of 44 patients with CD, 13 withIC, 4 NL and 20 PN and 3 UC. Of the 103 CT scansobtained in these 84 patients, 53 (51%) were normal.Four abscesses (3.8%) were demonstrated by CT(Table 2). In 1 child with an abscess demonstrated byCT, the 99mTc-WBC scan was obtained within 3 days anddemonstrated a focal abnormality compatible with anabscess. In the 3 other children with an abscess, the timedifference between the CT and the 99mTc-WBC scanwas greater than 1 month. Abnormal mesenteric fatwas seen in 6 patients (5.8%). A stricture was seen in 1patient (0.9%), and no fistulas were demonstrated.Backwash ileitis was not seen in any patients with UC.

CT versus 99mTc-WBC scan

In this subgroup, there were 16 children with CD, 2 withUC, 7 with IC, 14 PN, and 3 NL. Of the 42 CT scansobtained within a short time interval after the 99mTc-WBC scans, 21 (50%) were normal (Table 3). In the 21children with a positive 99mTc-WBC scan (Fig. 1), 62%(13/21) of CT scans underestimated the bowel wall in-flammation in at least one segment (total of 28 bowelsegments underestimated). In the children with a nega-tive 99mTc-WBC scan, there were 17 negative CT exami-

Table 2. Detection of IBD complications by CTa and99mTc-WBC

CT 99mTc-WBC

Abscess 4 1Abnormal mesenteric fat 6 0Stricture 1 0Fistula 0 0

a One hundred and three CT scans were performed in 84 children (44with Crohn’s disease, 13 with indeterminate colitis, 4 controls, and 20probably normal and 3 UC)

Table 3. Detection of inflammation by CT, 99mTc-WBC, andcolonoscopy in 42 children

Positive 99mTc-WBC Negative 99mTc-WBC

Positive CT 8 4Negative CT 13 17

26 M. Charron et al.: CT and 99mWBC vs colonoscopy in IBD

nations and 4 CT examinations showing an abnormalterminal ileum.

CT and 99mTc-WBC scan versus colonoscopy

This subgroup of 14 children consisted of 5 with CD, 3with IC, 1 with UC, and 5 with normal findings. WhenCT was compared with colonoscopy, there were fivetrue-negative CT, two true-positive CT, no false-positive, and seven false-negative CT examinations(Table 4). The seven false-negative CT examinationswere in patients with an abnormal colonoscopy findingwith mild to moderate inflammation (in 24 bowel seg-ments). When Tc-WBC was compared with colono-scopy, there were seven true-positive 99mTc-WBCstudies (Fig. 2), two false-negative, five true-negative,

and no false-positive 99mTc-WBC studies. The 99mTc-WBC scan was positive in 5 patients with a false-nega-tive CT. This translated into 19 segments that the CTdid not disclose as abnormal. The 5 other segments thatwere abnormal on colonoscopy and that had a normalCT finding were also normal on the 99mTc-WBC studies.

In three additional patients, CT and colonoscopy re-sults were available (no recent 99mTc-WBC scans). Onepatient was a child with a CT that showed a thick termi-nal ileum and cecum, and in whom the colonoscopyfindings were normal. The two other children had nor-mal findings on CTs, and the colonoscopy in these chil-dren showed inflammation in nine segments.

99mTc-WBC scan vs colonoscopy

In the 313 children studied with 99mTc-WBC scintigra-phy, there were seven false-negative and eight false-positive 99mTc-WBC studies. When compared with the

Fig. 1. Abnormal uptake on 99mTc-white blood cell (WBC)scintigraphy in the small bowel in a child with Crohn’s disease

Table 4. CT versus 99mTc-WBC versus colonoscopy

CT 99mTc-WBC Colonoscopy

True positive 2 7 9True negative 5 5 5False positive 0 0 0False negative 7 2 0

Fig. 2. Continuous left-sided colitis in a child with ulcerativecolitis (UC)

M. Charron et al.: CT and 99mWBC vs colonoscopy in IBD 27

106 colonoscopies, the sensitivity of the 99mTc-WBCscan was 92%, specificity was 94%; positive predictivevalue was 96%; negative predictive value was 93%; andaccuracy was 94%. In the 151 children (73 with CD, 17with UC, 8 with MC, and 53 children with no evidenceof IBD on follow up :probably Normal [PN]) without arecent colonoscopy, the gastroenterologist’s clinical as-sessment, the laboratory values, and the clinical follow-up (average 350 days) were consistent with the results ofthe 99mTc-WBC scan in most children (145/151). Therewas one false-negative study, in a 14-year-old femalewith MC with exacerbation of symptoms and a normal99mTc-WBC scan. There were five false-positive studies(two obtained during active gastrointestinal hemor-rhage and three unexplained). No false-positive 99mTc-WBC studies were seen in the controls.

The seven false-negative 99mTc-WBC scans were inthree patients with MC (one with indeterminate colitis,one with toxic colitis, and one with autoimmune colitis),two with CD, and two wth UC. Two of the childrenwith a false-negative 99mTc-WBC scan had normal labo-ratory values, a visually normal colonic mucosa duringcolonoscopy, and mild chronic inflammation on biopsy.Finally, the two children with UC and a normal 99mTc-WBC scan had normal laboratory values and mildinflammation on colonoscopy, limited to the rectum. Infour patients (two with CD, one PN, and one with MC)the scan underestimated the amount of inflammationin one segment; other segments were positive in thesepatients.

Uptake of 99mTc-WBC was seen in an abdominal scarin a child who had surgery (9 days before scanning).One child showed activity in a dilated ureter. Uptake of99mTc-WBC was seen at a stricture site, documented byupper gastrointestinal and small-bowel follow throughin 9 children. Uptake of 99mTc-WBC was absent at astricture site documented by upper gastrointestinalfollow through in another child. There was no recentcolonoscopy and biopsy to document the histologicalnature of the strictured areas. Two children with CDhad accumulation of 99mTc-WBC, in a perirectal abscessand perirectal fistula. Four other children had uptake of99mTc-WBC at an anastomotic site; the study could notdifferentiate active CD from physiologic inflammatoryrepair. Seven fistulae seen on upper gastrointestinal andsmall-bowel follow through showed focal uptake on the99mTc-WBC scan.

Discussion

Recent reviews18–22 profess emphatically that the clinicalutility of CT in IBD, has been demonstrated. There isonly one prospective study, in 13 patients with IBD, thatcompares CT with colonoscopy (63% of segments

reached by colonoscopy, for a total of 44 bowel seg-ments).23 CT had a lower sensitivity than the 99mTc-WBCscan for detecting bowel inflammation (68% for CT and85% for 99mTc-WBC). No controls or IBD patients withquiescent disease were included in this study to evaluatethe specificity of the findings on CT or 99mTc-WBCscans. Otherwise, the purported accuracy of CT in theevaluation of IBD stems from a few retrospectivestudies, typically involving approximately 20 patients.These studies14,19,24–26 did not correlate the findings witha gold standard such as endoscopy and clinical indicesof disease activity (Crohn’s disease activity index[CDAI]). No controls were included in these studies toassess the specificity of the CT findings, especially bowelthickness. Moreover, bias in patient selection favoredpatients with the most severe form of IBD.14,24,25 Thestudy of patients with the most severe form of IBD withCT can explain the high incidence of complicationsnoted in those reports. In those series, abscesses werereported in 25%–59% of patients,26,27 abnormal wallthickening in 82%–100%,26 mesenteric abnormalities in39%–72%,26 fistulae in 33%,14 abnormalities in the peri-rectal or perianal region in 40%–82%,3 abnormal termi-nal ileum in 80%,3 and backwash ileitis in 15%–20% ofpatients with UC.2 This high incidence of complicationsor signs of advanced disease is discordant with our expe-rience. Over half of the 103 CT scans we reviewed werenormal. Only 4% of our patients had abscesses, whileother groups have reported rates as high as 25%–59%.26,27 The incidence of abnormal mesenteric fat,stricture, and fistulae in this series was also significantlyless than that in the above-mentioned series. In ouropinion, the difference in reported rates of complica-tions detected by CT is owing to the different popula-tion studied. It should not be inferred from the studiescited above that these high incidences of complicationsare to be expected in patients with IBD; rather, theywere probably owing to a large bias in patient selection.Unfortunately, most recent review2,4–6,20,22,28 of IBD ada-mantly quote these high incidences of complications inIBD, and, accordingly, cite as references these smallretrospective studies that lack gold standards. We be-lieve that our series does not have such a bias, and ismore representative of the true incidence of complica-tions in children with IBD. One might also be temptedto conclude that CT is overutilized as a screening proce-dure. Every step must be taken to minimize radiationexposure in children.28–32

In this report we have documented, in a large numberof patients, the accuracy of 99mTc-WBC scan and thelimitations of CT for the evaluation of children withknown or suspected IBD. We found that the 99mTc-WBC scan was unlikely to miss significant inflammationwhen used for screening children for IBD, and CT hadan unacceptable sensitivity for detecting inflammation

28 M. Charron et al.: CT and 99mWBC vs colonoscopy in IBD

of the bowel wall. There were no false-positive 99mTc-WBC scan studies in the controls, and this confirms thehigh negative predictive value of 99mTc-WBC scintigra-phy. The incidence of complications detected by CT inthis series was substantially less than that noted inprevious reports.

References

1. Jacobs JE, Birnbaum BA. CT of inflammatory disease of thecolon. Semin Ultrasound CT MR 1995;16:91–101.

2. Scotiniotis I, Rubesin SE, Ginsberg GG. Imaging modalities ininflammatory bowel disease. Gastroenterol Clin North Am 1999;28:391–421, ix.

3. Jabra AA, Fishman EK, Taylor GA. CT findings in inflammatorybowel disease in children. AJR Am J Roentgenol 1994;162:975–9.

4. Wills JS, Lobis IF, Denstman FJ. Crohn disease: state of the art.Radiology 1997;202:597–610.

5. Carroll K. Crohn’s disease: new imaging techniques. BaillieresClin Gastroenterol 1998;12:35–72.

6. Gore RM, Balthazar EJ, Ghahremani GG, Miller FH. CT fea-tures of ulcerative colitis and Crohn’s disease. AJR Am JRoentgenol 1996;167:3–15.

7. Bartram CI, Halligan S. Radiological investigation of chronicinflammatory bowel disease in childhood. Baillieres ClinGastroenterol 1994;8:101–19.

8. Jabra AA, Fishman EK, Taylor GA. CT findings in inflammatorybowel disease in children. AJR Am J Roentgenol 1994;162:975–9.

9. Siegel MJ, Evans SJ, Balfe DM. Small bowel disease in children:diagnosis with CT. Radiology 1988;169:127–30.

10. Karahan OI, Dodd GD 3rd, Chintapalli KN, Rhim H, Chopra S.Gastrointestinal wall thickening in patients with cirrhosis: fre-quency and patterns at contrast-enhanced CT. Radiology 2000;215:103–7.

11. Tomei E, Diacinti D, Marini M, Boirivant M, Paoluzi P. Com-puted tomography of bowel wall in patients with Crohn’s disease:relationship of inflammatory activity to biological indices. Ital JGastroenterol 1996;28:487–92.

12. Gore RM. CT of inflammatory bowel disease. Radiol Clin NorthAm 1989; 27:717–29.

13. Jones B, Fishman EK, Hamilton SR, Rubesin SE, Bayless TM,Cameron JC, Siegelman SS. Submucosal accumulation of fatin inflammatory bowel disease: CT/pathologic correlation. JComput Assist Tomogr 1986;10:759–63.

14. Goldberg HI, Gore RM, Margulis AR, Moss AA, Baker EL.Computed tomography in the evaluation of Crohn disease. AJRAm J Roentgenol 1983;140:277–82.

15. Gossios KJ, Tsianos EV. Crohn’s disease: CT findings after treat-ment. Abdom Imaging 1997;22:160–3.

16. Charron M, delRosario JF, Kocoshis S. Pediatric inflammatorybowel disease: assessment with scintigraphy with 99mTc whiteblood cells. Radiology 1999;212:507–13.

17. Charron M, delRosario JF, Kocoshis S. Assessment of terminalileal and colonic inflammation in Crohn’s disease with Tc99m-WBC. Acta Paediatr 1999;88:193–8.

18. Gore RM, Balthazar EJ, Ghahremani GG, Miller FH. CT fea-tures of ulcerative colitis and Crohn’s disease [see comments].AJR Am J Roentgenol 1996;167:3–15.

19. Jabra AA, Fishman EK, Taylor GA. Crohn’s disease in the pedi-atric patient: CT evaluation. Radiology 1991;179:495–8.

20. Jacobs JE, Birnbaum BA. CT of inflammatory disease of thecolon. Semin Ultrasound CT MR 1995;16:91–101.

21. Caroline DF, Friedman AC. The radiology of inflammatorybowel disease. Med Clin North Am 1994;78:1353–85.

22. Horton KM, Corl FM, Fishman EK. CT evaluation of the colon:inflammatory disease(1) [In Process Citation]. Radiographics2000;20:399–418.

23. Kolkman JJ, Falke TH, Roos JC, Van Dijk DH, Bannink IM,Den Hollander W, et al. Computed tomography and granulocytescintigraphy in active inflammatory bowel disease. Comparisonwith endoscopy and operative findings. Dig Dis Sci 1996;41:641–50.

24. Gore RM, Marn CS, Kirby DF, Vogelzang RL, Neiman HL. CTfindings in ulcerative, granulomatous, and indeterminate colitis.AJR Am J Roentgenol 1984;143:279–84.

25. Gore RM. Cross-sectional imaging of inflammatory bowel dis-ease. Radiol Clin North Am 1987;25:115–31.

26. Gore RM, Cohen MI, Vogelzang RL, Neiman HL, Tsang TK.Value of computed tomography in the detection of complicationsof Crohn’s disease. Dig Dis Sci 1985;30:701–9.

27. Lambiase RE, Cronan JJ, Dorfman GS, Paolella LP, Haas RA.Percutaneous drainage of abscesses in patients with Crohn’s dis-ease. AJR Am J Roentgenol 1988;150:1043–5.

28. Balthazar EJ. CT of the gastrointestinal tract: principles and inter-pretation. AJR Am J Roentgenol 1991;156:23–32.

29. Paterson A, Frush DP, Donnelly LF. Helical CT of the body: aresettings adjusted for pediatric patients? AJR Am J Roentgenol2001;176:297–301.

30. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks ofradiation-induced fatal cancer from pediatric CT. AJR Am JRoentgenol 2001;176:289–96.

31. Diederich S, Lenzen H. Radiation exposure associated withimaging of the chest: comparison of different radiographic andcomputed tomography techniques. Cancer 2000;89(Suppl 11):2457–60.

32. Dixon AK, Dendy P. Spiral CT: how much does radiation dosematter? Lancet 1998;352:1082–3.