Virtual magnetic resonance colonography

Abstract

Colorectal cancer screening has vast potential. Beyond considerations for cost and diagnostic accuracy, the effectiveness of any colorectal screening strategy will be dependent on the degree of patient acceptance. Magnetic resonance (MR) colonography has been shown to be accurate regarding the detection of clinically relevant colonic polyps exceeding 10 mm in size, with reported sensitivity and specificity values exceeding 95%. To further increase patient acceptance, strategies for fecal tagging have recently been developed. By modulating the signal of fecal material to be identical to the signal characteristics of the enema applied to distend the colon, fecal tagging in conjunction with MR colonography obviates the need for bowel cleansing. The review will describe the techniques underlying MR colonography and describe early clinical experience with fecal tagging techniques.

Full Text

The Full Text of this article is available as a PDF (205.4 KB).

Figure 1.

For filling of the colon with 1000–2000 ml of a water based enema, the patient is placed in the prone position. The enema is administered using 100–150 cm of hydrostatic pressure.

Figure 2.

Colonic filling is monitored with a non-slice select 2D acquisition collecting one image every three seconds. The last image on the right demonstrates that enema has reached the caecum.

Figure 3.

The high contrast between the contrast filled colonic lumen and surrounding structures (on the left) is the basis for subsequent virtual colonic viewing (on the right).

Figure 4.

A single coronal section of a 3D dataset demonstrates a small filling defect in the sigmoid colon. Virtual endoscopic viewing confirms the presence of a polyp.

Figure 5.

Dark lumen MR colonography focuses on the colonic wall. After filling of the colonic lumen with warm tap water a 3D gradient echo dataset is collected over 20 seconds in apnea 75 seconds after the intravenous administration of paramagnetic contrast. At this time there is avid enhancement of the colonic wall as well as of colorectal masses arising from it. In this case a large 15 mm polyp is identified arising from the sigmoid colon. The polyp is seen to clearly on the coronal source image (top left), the sagittal reformation (top right), the axial reformation (bottom left), and the virtual endoscopic view (bottom right).

Figure 6.

Single coronal section of a dark lumen MR colonography 3D dataset, which was acquired over 20 seconds in apnea 75 seconds after the intravenous administration of paramagnetic contrast. A small 7 mm polyp is identified arising from the sigmoid colon. The polyp is seen clearly on the coronal source image (left) as well as on the virtual endoscopic view (right).

Figure 7.

Maximum intensity projection (MIP) display of a 3D MR colonography dataset collected after gadolinium based fecal tagging (left). After filling the colon with a gadolinium containing enema, the tagged stool is no longer seen because its signal intensity is similar to that of the applied rectal gadolinium/water enema. The intraluminal signal intensity is sufficiently homogeneous to permit virtual colonoscopic viewing (right).

Figure 8.

Images of two volunteers undergoing MR colonography based on the rectal administration of a barium enema in conjunction with the intravenous administration of paramagnetic contrast. One volunteer did (right) and one volunteer did not (left) ingest barium for fecal tagging. In both cases the colonic wall is bright after the after application of paramagnetic contrast. Fecal material in the volunteer who had not received barium for fecal tagging is bright and can thus not be differentiated from the brightly enhancing wall. The presence of colonic polyps cannot be excluded with certainty. The colonic lumen of the volunteer who had ingested the barium tagging agent on the other hand is homogenously dark, thereby permitting a full analysis of the colon.

Figure 9.

Coronal source image (right) and virtual endoscopic rendering thereof (left) of 3D gradient echo dataset collected after the administration of a barium enema and the intravenous injection of gadobenate dimeglumine. A contrast enhancing colonic carcinoma close to the right colonic flexure is identified both on the coronal section as well as on the virtual endoscopic view.

Figure 10.

Axial reformatted image (right) and virtual endoscopic rendering (left) of a 3D gradient echo dataset collected after the administration of a barium enema and the intravenous injection of gadobenate dimeglumine. A contrast enhancing 10 mm polyp is identified in the sigmoid colon on both the single axial section as well as the virtual endoscopic rendering.