Have an account? Please log in.
Text size: Small font Default font Larger font
Radiology Daily
Radiology Daily PracticalReviews.com Radiology Daily

Superiority of 2-D versus 3-D Image Analysis Not Known for Colonoscopy

September 7, 2009
Written by: , Filed in: Gastrointestinal Imaging
  • Comments

On both virtual and optical colonoscopy being used to perform screening examinations for colorectal cancer, hidden regions within the colon can be problematic when the images are being evaluated.

For virtual colonoscopy, radiologists who are flying-through 3-dimensional (3-D) images will find hidden regions where they can’t visualize mucosal surfaces.

Nonetheless, with 3-D image analysis, an advantage is that the radiologist can “paint” these hidden areas and manually review them to determine the size of each particular patch.

The approach taken to measuring a polyp depends on whether the image is 2-dimensional (2-D) or 3-D. For 2-D images, the long axis of the polyp must be measured, including the base but not the stalk. With 3-D images, the polyp is measured along the top, along the longest axis. If some material is adhering to that polyp, then the polyp’s size may be artificially increased on the 3-D image.

Therefore, virtual 3-D is prone to overestimating the polyp’s size, and 2-D images tend to underestimate it. Most radiologists believe that 3-D measurements are somewhat more accurate.

Gold Standard for Measurements

When a gastroenterologist is viewing the inside of the colon, he or she is usually making an estimate of a polyp’s size based on some structure with a known size.

As far as the location of the polyp, the gastroenterologist knows exactly where it is when the polyp has been bookmarked on the images. The software can even identify how far the polyp is from the anal verge.

These measurements are not exactly given in terms of colonoscope lengths because the colon is telescoped over the inserted colonoscope. Instead, measurements correspond to the actual distance at the centerline to the polyp, which is useful for at least helping the colonoscopist know where in space the fragment is found.

Selecting 2-D versus 3-D
As more studies are done comparing 2-D and 3-D image analysis, we will eventually know which approach is superior. Compared with 3-D images, 2-D image analysis is faster and simpler and most radiologists know how to read axial CT scans. Less computer power is needed with 2-D images. High-resolution powerful 3-D image analysis requires more computer power.

Nonetheless, 3-D images give radiologists the colonoscopy perspective, so they see the same basic perspective as what gastroenterologists see.

Also, 3-D images show more detail and more sensitivity, particularly for polyps near the folds, and they allow surface tracking and more accurate measurements. Stool false-positive results can be a problem with 3-D images.

Moreover, 3-D images take longer to review, and there can be some reader fatigue when there are several cases to evaluate.

Take-Home Pearl
In the evaluation of virtual colonoscopy images, 3-D images allow hidden regions to be assessed, make more accurate measurements of polyp size, and are more sensitive for polyps near folds.

Assessment of Hidden Regions, Measurements, and Selection Factors in the Use of 2-D and 3-D Image Analysis for Virtual Colonoscopy.
Richard S. Breiman, MD

This review is an abstract of an audio presentation from Practical Reviews in Radiology. If you do not have access to this presentation and would like to purchase a copy, please call 1-800-633-4743, email service@oakstonepub.com, or write Oakstone Medical Publishing, 100 Corporate Parkway, Suite 600, Birmingham, Alabama 35242.

  • Comments

Would you like to keep current with radiological news and information?

Post Your Comments and Responses

Comments are closed.