| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Departments of Diagnostic Radiology (T.C.L., S.C.G., S.G.R., J.F.D.) and Gastroenterology and Hepatology (G.H.), University Hospital Essen, Hufelandstrasse 55, D-45122 Essen, Germany. Received May 4, 2001; revision requested June 8; revision received August 22; accepted October 10. Address correspondence to T.C.L. (e-mail: thomas.lauenstein@uni-essen.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Twenty-four patients suspected of having colonic lesions because of rectal bleeding, positive fecal occult blood test results, or altered bowel habits underwent MR colonography and subsequent conventional colonoscopy. A 200-mL dose of a barium sulfate–containing contrast agent was ingested with each of four low-fiber meals, beginning 36 hours before the examination. For MR colonography, the colon was filled with tap water. Gadobenate dimeglumine was injected intravenously. Images were acquired 75 seconds after gadobenate dimeglumine administration by using only a T1-weighted three-dimensional gradient-echo sequence. Images were reviewed by two radiologists blinded to conventional colonoscopic data. By using colonoscopy as the reference standard, sensitivity and specificity of MR colonography were determined for detecting colorectal masses.
RESULTS: On the basis of MR colonography, 15 polyps of 5–20 mm and 10 carcinomas were detected and later confirmed with conventional colonoscopy. Conventional colonoscopy depicted three additional lesions less than 8 mm in diameter. Thus, sensitivity of MR colonography was 89.3% (25 of 28) for lesions and 91.7% (22 of 24) for patients.
CONCLUSION: Barium-tagged MR colonography obviates bowel cleansing and depicts all lesions exceeding 8 mm in diameter.
© RSNA, 2002
Index terms: Barium, 75.12143 • Colon, MR, 75.12117 • Colon neoplasms, 75.311, 75.32 • Colonoscopy, 75.1289 • Magnetic resonance (MR), contrast media, 75.12143 • Magnetic resonance (MR), three-dimensional, 75.12143, 12149
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Conventional colonoscopy, known to be highly accurate regarding detection of colorectal lesions (4), is applied for colonic screening (5). It offers the advantage of simultaneous biopsy or polyp removal. Disadvantages include need for bowel cleansing and dietary restrictions prior to examination; procedural discomfort during examination, frequently requiring patient sedation and analgesia; inability to reach the right side of the colon in some patients; and risk of perforation (6,7). These have led to poor patient participation in screening programs, even if access to colonoscopy was free (8–10). The effect of colonoscopic screening on incidence of colorectal cancer has thus been limited.
Virtual colonography, based on acquisition of three-dimensional (3D) computed tomographic (CT) or morphologic magnetic resonance (MR) imaging data sets, is less painful than colonoscopy and does not require administration of analgesics or sedatives (11). Recent studies (12–15) have shown both CT and MR colonography (16–18) to be accurate regarding detection of polyps exceeding 10 mm in diameter. Since CT examinations of the abdomen expose patients to a considerable dose of ionizing radiation (19), the future of CT colonography as a screening method is uncertain, although it is less costly and provided with higher availability, as compared with MR imaging. MR colonography, on the other hand, is not associated with radiation. In addition, contrast agents used for MR colonography are characterized by a more favorable safety profile than are CT contrast agents, as they lack nephrotoxicity and are associated with far fewer anaphylactoid reactions (20,21).
Virtual colonography mandates bowel cleansing in a manner similar to that of colonoscopy. Since more than half of patients undergoing bowel preparation complain about symptoms ranging from "feeling unwell" to "inability to sleep" (22), patient acceptance is negatively affected. To assure high patient acceptance of virtual MR colonography, bowel cleansing needs to be eliminated. This can be achieved with fecal tagging, a concept based on altering the signal intensity of stool by adding contrast-modifying substances to regular meals. The feasibility of one approach to fecal tagging has already been demonstrated: On the basis of brightly tagged stool and a bright rectal enema, polyps are visualized as dark filling defects (23). Despite promising results, the high costs of the applied gadolinium-based tagging agents have prevented this strategy from reaching clinical relevance.
The purpose of this study was to assess a strategy for fecal tagging with barium sulfate, which is dark on images obtained with T1-weighted sequences, as an inexpensive tagging agent in conjunction with MR colonography in patients suspected of having colorectal lesions. The preliminary evaluation of this approach in a volunteer study (24) demonstrated excellent delineation of the colonic wall against a colonic lumen containing dark barium-tagged stool, dark barium enema for colonic distention, and signalless air after intravenous administration of paramagnetic contrast material.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study was conducted in accordance with all guidelines set forth by the approving institutional review board. Informed consent was obtained prior to each examination. Exclusion criteria were based on contraindications to MR imaging, such as the presence of a pacemaker or metallic implants in the central nervous system, or claustrophobia.
Fecal Tagging
For fecal tagging, a highly concentrated barium sulfate–containing contrast agent (1 mg/mL barium sulfate, Micropaque; Guerbet, Sulzbach, Germany) was administered in a dose of 200 mL with each of four principal meals, beginning 36 hours before MR colonography. Patients were instructed to avoid intake of all fiber-rich food and nourishment with a high concentration of manganese, such as chocolate or fruit (25), during this period. Otherwise, all subjects were free to choose their diet, and there were no restrictions on fluid intake.
MR Colonography
MR examinations were performed with a 1.5-T system (Magnetom Sonata; Siemens Medical Systems, Erlangen, Germany) equipped with high-performance gradient systems characterized by a maximum gradient amplitude of 40 mT/m and a slew rate of 200 mT/m/msec. To cover the entire colon, a combination of two large flexible surface coils was used for signal reception. To minimize bowel peristalsis and reduce colonic spasm, 20 mg of scopolamine (Buscopan; Boehringer Ingelheim, Ingelheim, Germany) was administered intravenously. Because of a medical history of glaucoma, one patient received 1 mg glucagon hydrochloride (Glucagen; Novo Nordisk, Mainz, Germany) instead of scopolamine. After placement of a rectal tube (E-Z-Em, Westbury, NY), the colon was filled with 1,500–2,500 mL of tap water with the patient in the prone position, by using hydrostatic pressure (1.0–1.5-m water column). As a precaution to avoid damage to the imaging table, towels were placed under the patient in case the enema could not be retained by the patient.
The filling process was monitored by using a two-dimensional true fast imaging with steady-state precession (FISP) sequence (3.2/1.6 [repetition time msec/echo time msec]; flip angle, 70°; matrix, 256 x 256; field of view, 400 mm; section thickness, 4 mm), with acquisition of one image every 3 seconds in the coronal plane. The filling process was halted when true FISP images demonstrated the water as having reached the cecum. Once complete filling and adequate distention of the colon were demonstrated, paramagnetic contrast material (gadobenate dimeglumine, Multihance; Bracco, Milan, Italy) was intravenously administered by using an automatic injector (Spectris; Medrad, Volkach, Germany). Injection parameters included a dosage of 0.2 mmol per kilogram of body weight gadobenate dimeglumine and a flow rate of 3 mL/sec, followed by rapid injection of 20 mL of normal saline at the same flow rate. After a delay of 75 seconds, a T1-weighted 3D gradient-echo data set was acquired over 22 seconds in a single breath hold, by using the following parameters: 1.64/0.60; flip angle, 15°; field of view, 45 cm; matrix, 256 x 230; and section thickness, 3.14 mm (64 sections). Zero interpolation was applied in all three planes, rendering an effective section thickness of 1.57 mm and a matrix of 460 x 512. The most superior level for 3D acquisition was chosen from the initial true FISP sequence, permitting coverage of the entire liver. The examination was performed with the patient in only the prone position. After acquisition of the 3D data set, the enema bag was placed on the floor for facilitated emptying of the colon, and the patient was removed from the scanner. All MR examinations were completed within 20 minutes, including time needed to place an intravenous catheter. MR colonography was tolerated well by all patients. Neither the oral ingestion of barium nor the MR examination itself, including placement of the rectal tube, adversely affected patient comfort.
Image Analysis
The 3D data sets were later processed by using commercially available software and hardware (Virtuoso; Siemens Medical Systems). MR colonography was interpreted in multiplanar reformation mode, by scrolling through the 3D data set in all three orthogonal planes, as well as on the basis of virtual endoscopic renderings. Two experienced MR radiologists (S.C.G., S.G.R.) blinded to conventional coloscopic data recorded the location and size of all detected endoluminal lesions in a prospective consensus reading. On the basis of lesion size and morphology, detected masses were characterized as either polyp (
2 cm, homogeneous morphology) or carcinoma (>2 cm, inhomogeneous morphology). Suspicion of concomitant hepatic metastasis with the same 3D data set was used as an additional factor.
For quantitative assessment, signal intensity differences between the colonic lumen and the colonic wall, as well as colorectal mass lesions, were determined. To this end, regions of interest were placed by one radiologist in the lumen (ellipse, mean diameter, 6.5 mm) and adjacent normal wall (ellipse, mean diameter, 1.5 mm) of the ascending, transverse, and descending colon. Furthermore, regions of interest were measured in all identified colonic lesions (ellipse; mean diameter, 2.0 mm). Image noise, defined as the SD of signal intensities, was measured in a region of interest (ellipse, mean diameter, 20.0 mm) placed in the recorded field of view on either the left or the right side of the body outside the abdomen. Based on these measurements, contrast-to-noise ratios were calculated in the usual manner: CNR = [SI (colonic wall/colonic lesion) - SI (lumen)]/noise, where CNR is contrast-to-noise ratio and SI is signal intensity.
Conventional Colonoscopy
Between 3 and 7 days following MR colonography, patients underwent conventional colonoscopy with commercially available equipment (model CFQ 140; Olympus Optical Europe, Hamburg, Germany). The attending gastroenterologist (G.H.) was unaware of the MR findings. When necessary, sedatives (midazolam hydrochloride, Dormicum; Roche, Reinach, Switzerland) were administered, or analgesia (pethidin, Dolantin; Aventis Pharma, Bad Soden, Germany) was provided to the patient. The location and size of colorectal masses were recorded. All polyps were removed at polypectomy, and biopsy was performed on suspected cancers. Colorectal masses were investigated with histopathologic examination. There were no incomplete conventional colonoscopic examinations.
Other Diagnostic Methods
In patients with colorectal carcinomas or hepatic metastases, an additional CT examination of the liver was performed before surgery, with use of a multisection CT scanner (Somatom Volume Zoom; Siemens). Examination parameters included intravenous contrast material administration of 100 mL of iopromid (Ultravist 300; Schering, Berlin, Germany) at 3 mL/sec. Data were acquired after a 60-second delay, by using 4 x 2.5-mm collimation. Images were reconstructed at 2-mm intervals, with 3-mm section overlap.
Data Analysis
Conventional colonoscopic findings were considered the standard of reference. The accuracy of MR colonography was assessed by calculating point estimates for sensitivity and specificity.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In contrast with previous approaches of MR or CT colonography, polyp detection with the barium-based technique does not merely rely on identification of an endoluminal filling defect in a cleansed contrast material–distended colon; rather, it is based on enhancement of colorectal mass lesions after intravenous administration of paramagnetic contrast material. Enhancement of colorectal lesions after intravenous administration of contrast material has been described in conjunction with both CT (26) and MR colonography (27). In both cases, lesion enhancement was seen as a means to help differentiate residual fecal material, which can simulate polyps, from colonic masses (28,29). Our described approach extends this strategy to eliminating the need for colonic cleansing altogether. The presence of fecal material does not impair the ability to detect enhancing colorectal mass lesions, as long as the signal intensity properties of stool allow for such differentiation.
The MR signal intensity properties of stool are heterogeneous. While the amount of contained water most strongly influences its appearance on T2-weighted images, T1-shortening manganese, iron, and, to a lesser degree, fat determine the degree of brightness of the otherwise dark stool on T1-weighted images (25,30,31). The signal-modulating effects of commercially available intravenous paramagnetic contrast agents are most evident on heavily T1-weighted images. For maximal contrast between brightly enhancing colorectal masses and surrounding fecal material on T1-weighted images, the latter needs to be rendered homogeneously dark. This can be accomplished by reducing manganese intake and administering a signal intensity–reducing agent in conjunction with all main meals over a prolonged period prior to MR examination. The agent should be well tolerated, not absorbable, readily available, and inexpensive.
Barium sulfate fulfills the properties of an ideal tagging agent. It is inexpensive, widely available in different flavors from a variety of vendors, and totally harmless when administered orally, almost without incidence of anaphylactoid reaction (32). It remains in widespread use as an oral contrast agent for esophageal, gastric, and small-bowel radiography. The agent is not absorbed and mixes well with stool. When ingested orally at doses of 200 mL with each principal meal, it renders stool homogeneously dark (at least it did for a majority of patients in the current study). Two of 24 patients exhibited bright stool, thereby severely limiting our ability to assess the colon for enhancing masses. In one of these patients, two small polyps were missed at MR colonography. In these cases, a subtraction of native and contrast material–enhanced images could be useful as proposed for CT colonography (33). Beyond noncompliance with instructions governing oral intake of barium and restricted intake of foods rich in manganese (25), another cause for the obvious failure of fecal tagging in the two patients may relate to the length of the preparation period. The agent was administered with all main meals over 36 hours preceding the MR examination. This period may have been too short; mean bowel transit times can extend well beyond 48 hours (34). In a recent study in which fecal tagging in conjunction with CT colonography was investigated, Callstrom et al (35) demonstrated that a window of 48 hours for ingestion of tagging substances is optimal.
Furthermore, it is possible that not enough barium was ingested. In this initial study, we chose to administer a standard dose for all patients, regardless of body weight and food intake. Thus, larger patients with greater food volumes may not have received a sufficient amount of barium.
Finally, it is possible that too much food was ingested outside principal meals, thereby causing signal intensity heterogeneity. Clearly, there is ample room for optimizing the fecal tagging regimen. A prolonged food intake–adjusted regimen taking into account small meals is required. Use of a more concentrated formulation of barium, possibly already mixed into prepackaged meals, should be considered. The colon is distended with a tap water enema. Water is ideal as a distention agent, as it is inexpensive and homogeneously dark on T1-weighted images. Furthermore, it mixes well with fecal material in the colon. However, MR colonography also offers new perspectives regarding optimization of bowel distention. A modified strategy could be based on application of gas like CO2 (36). The gas is signalless and would thus easily permit delineation of the contrast-enhanced colonic wall and masses.
Reflecting concomitant administration of a spasmolytic agent immediately prior to colonic filling, precautionary placement of towels was not needed in any of the patients examined in this study. The high-spatial-resolution 3D data set is collected only after real-time true FISP images have documented that the enema has reached the cecum. In contrast with previous MR colonographic techniques that require acquisition of two 3D data sets to overcome problems associated with residual colonic air (16,17), the proposed strategy requires collection of only a single prone data set. Residual air inside the colon, which is dark on all MR images, can be delineated from enhancing colonic wall in the same manner as the dark water enema and dark barium-tagged fecal material. Hence, there is no need for a second image set to be obtained with the patient supine. This shortens the examination considerably and further reduces discomfort to the patient.
The colonic wall with adherent mass lesions is enhanced with gadobenate dimeglumine administered at a dose of 0.2 mmol/kg with a flow rate of 3 mL/sec. A delay of 75 seconds between completion of contrast material administration, which never exceeds 15 seconds, and data acquisition has been determined empirically (Papanicolaou N, personal communication, 2000). On the basis of this timing regimen, enhancement of the colonic wall is characterized by contrast-to-noise ratio values exceeding 20, sufficiently high to permit virtual endoscopic rendering of the colon. At least theoretically, the effect of fecal tagging in conjunction with colonic wall enhancement could be increased by performing image subtraction. The 3D data set is acquired with breath holding and requires merely 20 seconds, reflecting use of high-performance gradient systems. For image subtraction, an additional nonenhanced 3D image set is collected and subsequently subtracted from the contrast-enhanced data set (37). For this technique to work, the depths of inspiration for the two successive breath-hold acquisitions would need to be synchronized. Furthermore, the 3D data sets would need to be collected in immediate succession to avoid misregistration due to peristaltic motion. To our knowledge, the value of subtractions for the detection of colorectal masses still needs to be established.
The value of virtual colonography has been well established. The most relevant advantages over conventional colonoscopy relate to lack of procedural pain and discomfort. Angtuaco et al (38) demonstrated that more than 60% of potential patients preferred virtual colonography over conventional colonoscopy when both methods were offered. Noninvasiveness and lack of sedation were the main reasons for preference of virtual colonography. Another advantage of virtual colonography includes the fact that image interpretation is not limited to the endoscopic perspective. Because 3D data sets are acquired, colonic morphology can be displayed from any desired vantage point. The described MR colonographic strategy really focuses on imaging the colonic wall itself. Beyond detection of colorectal masses, diagnostic workup for other abnormal conditions affecting the colonic wall, such as inflammatory disease, is likely to benefit from this approach. In assessment of the small bowel, the value of bowel wall enhancement in evaluation of patients with Crohn disease is well recognized (39). In addition, other organ systems contained within the imaging volume, such as the liver or the abdominal aorta, can be assessed. Also in this regard, intravenous administration of paramagnetic contrast material is most helpful. Hepatic metastasis, as well as an aortic aneurysm, could be readily identified as such.
Beyond its preliminary nature, this study was limited by a severe selection bias toward patients with a high likelihood of harboring colorectal masses. Diagnostic accuracy of the technique could hence be overestimated when compared with a patient collective including persons with average risk. Clearly, the impact of this method must be confirmed in larger studies with patients with normal risk. Other issues, in particular those related to cost (40) and limited availability of MR imagers, also will need to be considered. Despite these limitations, we believe it warranted to conclude that assessment for colorectal masses on the basis of a strategy combining fecal tagging with barium sulfate and intravenous contrast–enhanced 3D MR imaging is possible with a high degree of accuracy, without prior colonic cleansing.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, G.H., J.F.D.; study concepts, T.C.L., S.G.R.; study design, J.F.D., T.C.L.; literature research, T.C.L.; clinical studies, G.H., S.C.G., T.C.L.; data acquisition, S.C.G., T.C.L.; data analysis/interpretation, S.G.R., S.C.G.; statistical analysis, T.C.L.; manuscript preparation, T.C.L.; manuscript definition of intellectual content, J.F.D., T.C.L.; manuscript editing, T.C.L.; manuscript revision/review, G.H., J.F.D.; manuscript final version approval, G.H., J.F.D., T.C.L.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Yamada, S. Okabe, M. Enomoto, K. Sugihara, N. Yoshino, A. Tetsumura, J. Kumagai, and H. Shibuya Colorectal Carcinoma: In Vitro Evaluation with High-Spatial-Resolution 3D Constructive Interference in Steady-State MR Imaging Radiology, December 19, 2007; (2007) 2462070128. [Abstract] [Full Text]
|
|
|
|
|
|
J. Florie, E. Birnie, R. E. van Gelder, S. Jensch, B. Haberkorn, J. F. Bartelsman, A. van der Sluys Veer, P. Snel, V. P. M. van der Hulst, G. J. Bonsel, et al. MR Colonography with Limited Bowel Preparation: Patient Acceptance Compared with That of Full-Preparation Colonoscopy Radiology, October 1, 2007; 245(1): 150 - 159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B Saar, A Meining, A Beer, M Settles, H Helmberger, E Frimberger, E J Rummeny, and T Rosch Prospective study on bright lumen magnetic resonance colonography in comparison with conventional colonoscopy Br. J. Radiol., April 1, 2007; 80(952): 235 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Florie, S. Jensch, R. A. J. Nievelstein, J. F. Bartelsman, L. C. Baak, R. E. van Gelder, B. Haberkorn, A. van Randen, M. M. van der Ham, P. Snel, et al. MR Colonography with Limited Bowel Preparation Compared with Optical Colonoscopy in Patients at Increased Risk for Colorectal Cancer Radiology, April 1, 2007; 243(1): 122 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Wessling, R. Fischbach, A. Borchert, H. Kugel, T. Allkemper, N. Osada, and W. Heindel Detection of Colorectal Polyps: Comparison of Multi-Detector Row CT and MR Colonography in a Colon Phantom Radiology, October 1, 2006; 241(1): 125 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Adusumilli, H. K. Hussain, E. M. Caoili, W. J. Weadock, J. P. Murray, T. D. Johnson, Q. Chen, and B. Desjardins MRI of sonographically indeterminate adnexal masses. Am. J. Roentgenol., September 1, 2006; 187(3): 732 - 740. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V L Jardine, E Sala, and D J Lomas MR colonography: baseline appearance of the unprepared rectosigmoid Br. J. Radiol., March 1, 2005; 78(927): 202 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Goehde, P. Hunold, F. M. Vogt, W. Ajaj, M. Goyen, C. U. Herborn, M. Forsting, J. F. Debatin, and S. G. Ruehm Full-Body Cardiovascular and Tumor MRI for Early Detection of Disease: Feasibility and Initial Experience in 298 Subjects Am. J. Roentgenol., February 1, 2005; 184(2): 598 - 611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Ajaj, T. C. Lauenstein, G. Pelster, G. Holtmann, S. G. Ruehm, J. F. Debatin, and S. C. Goehde MR Colonography in Patients with Incomplete Conventional Colonoscopy Radiology, February 1, 2005; 234(2): 452 - 459. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. B. Cotton, V. L. Durkalski, B. C. Pineau, Y. Y. Palesch, P. D. Mauldin, B. Hoffman, D. J. Vining, W. C. Small, J. Affronti, D. Rex, et al. Computed Tomographic Colonography (Virtual Colonoscopy): A Multicenter Comparison With Standard Colonoscopy for Detection of Colorectal Neoplasia JAMA, April 14, 2004; 291(14): 1713 - 1719. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Papagrigoriadis, I Arunkumar, A Koreli, and W A Corbett Evaluation of flexible sigmoidoscopy as an investigation for "left sided" colorectal symptoms Postgrad. Med. J., February 1, 2004; 80(940): 104 - 106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. W. F. Geenen, S. M. Hussain, F. Cademartiri, J.-W. Poley, P. D. Siersema, and G. P. Krestin CT and MR Colonography: Scanning Techniques, Postprocessing, and Emphasis on Polyp Detection RadioGraphics, January 1, 2004; 24(1): e18 - e18. [Abstract] [Full Text]
|
|
|
|
 |
|
 J F Debatin and T C Lauenstein Virtual magnetic resonance colonography Gut, June 1, 2003; 52(90004): iv17 - 22. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Preparation-Free Magnetic Resonance Colonography Journal Watch Gastroenterology, May 28, 2002; 2002(528): 3 - 3. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
