Fundamental elements for successful performance of CT colonography (virtual colonoscopy)

Abstract

There are many factors affecting the successful performance of CT colonography (CTC). Adequate colonic cleansing and distention, the optimal CT technique and interpretation with using the newest CTC software by a trained reader will help ensure high accuracy for lesion detection. Fecal and fluid tagging may improve the diagnostic accuracy and allow for reduced bowel preparation. Automated carbon dioxide insufflation is more efficient and may be safer for colonic distention as compared to manual room air insufflation. CT scanning should use thin collimation of < or =3 mm with a reconstruction interval of < or =1.5 mm and a low radiation dose. There is not any one correct method for the interpretation of CTC; therefore, readers should be well-versed with both the primary 3D and 2D reviews. Polyps detected at CTC should be measured accurately and reported following the "polyp size-based" patient management system. The time-intensive nature of CTC and the limited resources for training radiologists appear to be the major barriers for implementing CTC in Korea.

Fig. 1

A 68-year-old male with fecal residue that is tagged with orally-administered barium. A. 3D endoluminal view shows an 8-mm polypoid structure (arrowhead) on a haustral fold of the sigmoid colon. B. 2D transverse image using a wide-window setting (width: 1500 HU, level: -400 HU) shows very high attenuation of the olypoid structure (white arrowhead), which clearly demonstrates it is tagged stool. Another piece of tagged stool is noted in the sigmoid colon (black arrowhead).

Fig. 2

A 42-year-old male with fecal residue that is tagged with orally-administered gastrografin. A. 3D endoluminal view shows a 10-mm polypoid structure (arrowhead) in the descending colon. B. 2D transverse image with using a wide-window setting (width: 1500 HU, level: -400 HU) shows very high attenuation of the polypoid structure (arrowhead), and this consistent with tagged stool.

Fig. 3

A 65-year-old male with a 6-mm tubulovillous adenoma in the sigmoid colon. A. Optical colonoscopy shows a sessile polyp (arrowhead) in the sigmoid colon. B. 3D supine endoluminal view shows the corresponding polyp (arrowhead) in a well-distended segment of the sigmoid colon. C. 3D endoluminal view of the same area as figure B, with the patient in a prone position, shows the same lesion (arrowhead) obscured by suboptimal distention.

Fig. 4

A 68-year-old female with a pseudopolyp in the transverse colon. A. 3D prone endoluminal view shows focal bulbous thickening (arrowhead) of a haustral fold in the suboptimally-distended transverse colon, which resembles a sessile polyp. B. 3D endoluminal view of the same area as figure A, with the patient in a supine position, shows the disappearance of the pseudopolyp (arrowhead) when the colon is well-distended.

Fig. 5

A 54-year-old female with a large amount of tagged fecal residue in the sigmoid colon. A. The 3D endoluminal view shows many polypoid and mass-like structures in the sigmoid colon. Examining each polypoid or mass-like structure to distinguish a true polyp/mass from fecal residue is tiresome. Additionally, lesions buried under fecal material are not detected at all during the 3D fly-through. B. Using a wide-window setting (width: 1500 HU, level: -400 HU) for the 2D transverse image, all the pseudolesions (i.e. tagged fecal residue) are easily recognized.

Fig. 6

A 53-year-old male with an adenocarcinoma and a tubular adenoma in the sigmoid colon. A. The locations of the adenocarcinoma (arrowheads) and the tubular adenoma (arrow) in the sigmoid colon are clearly demonstrated on the raysum image. B. The surgical specimen shows a 3.5-cm ulcerofungating adenocarcinoma (arrowheads) and an 8-mm pedunculated tubular adenoma (arrow) in the sigmoid colon. C. The 3D "virtual dissection" image shows the morphologic distortion of the cancer mass (arrowheads) and the polyp (arrow). D, E. In contrast to the virtual dissection image, image distortion is not noted on the volume-rendered image along the centerline of the colon without flattening (D) and on the 3D endoluminal view (E). Both lesions (arrowheads and arrow) show similar morphologies to those of the surgical specimen.

Fig. 7

A 53-year-old male with a tubular adenoma in the sigmoid colon. A. 3D endoluminal view shows three different measurements ("I, II and III") of a sessile polyp with a narrow neck. Which of the three measurements is correct? "I (8.8 mm)" is correct. In "II (16.2 mm)", the measurement cursor (red arrowhead) is placed slightly off the lesion margin and leads to an erroneous measurement. In "III (16.3 mm)", the placement of the measurement cursor (green arrowhead) is clearly incorrect, resulting in an erroneous measurement. B. Correlative measurement diagrams for "I, II and III" of figure A. If the measurement cursor is placed off the lesion boundary (II and III), then the first pixel on the colonic wall beyond the polyp in the viewing direction (arrows) is chosen for measurement (i.e. the measurement cursor cannot be placed in an "empty" luminal space, rather, it is placed on the colonic wall after progressing some distance in the viewing direction until it hits the colonic wall), causing overestimation of lesion size. Even slight misplacement of the cursor (e.g. a single pixel), which may be imperceptible by the naked eye, causes this phenomenon and this can lead to significant overestimation of lesion size as demonstrated in measurement "II".