CT Colonography Algorithm Helps Radiologists Recognize Polyps

July 14, 2014

Using an algorithm with CT colonography allows radiologists to recognize previously identified polyps.

An algorithm for automatic endoluminal coregistration used at initial CT colonography allows radiologists to predict endoluminal polyp location at subsequent CT colonographies, helping them recognize previously identified polyps, according to an article published in the journal Radiology.

Polyps over 1 cm are generally removed, but smaller polyps represent a treatment dilemma, particularly among patients who may have certain risks associated with colonoscopy screening or who refuse colonoscopies, the authors wrote. Currently, radiologists compare polyps between initial and subsequent studies, but this can be challenging without markers.

In this study, researchers from the U.K. and the U.S. evaluated the accuracy of a method of automatic coregistration of endoluminal surfaces to determine if they could match polyps between initial and subsequent colonographies.

Twenty-six asymptomatic patients (16 mine) who underwent colonography screening between 2004 and 2011 participated in the study. The mean age was 61 at time of the initial study and 64 at follow-up. There were 35 polyps, with a mean size of 6.1 mm at initial study and 7.7 mm at follow-up.

The researchers used a registration algorithm designed to coregister 

the coordinates of endoluminal colonic surfaces on images from prone and supine CT colonographic acquisitions to match polyps in sequential studies. The algorithm was tested by longitudinal and consistency methods.

The polyp coordinates from the initial supine acquisition were used to identify polyp location on images from the initial prone acquisition, and for follow-up prone and follow-up supine acquisitions.

[[{"type":"media","view_mode":"media_crop","fid":"25782","attributes":{"alt":"Manual identification of polyp location. Image courtesy of Radiology. ©RSNA, 2014. ","class":"media-image media-image-right","id":"media_crop_6989809318911","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"2418","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"border-style: solid; line-height: 1.538em; height: 181px; width: 299px; float: right;","title":"Manual identification of polyp location. Image courtesy of Radiology. ©RSNA, 2014. ","typeof":"foaf:Image"}}]]

According to the findings both methods resulted in coregistration being achieved for all 35 polyps. “Mean standard deviation Euclidean registration error for the longitudinal method was 17.4 mm ± 12.1 and for the consistency method, 26.9 mm ± 20.8,” the authors wrote. There was no significant difference between these results and the registration error when prone and supine acquisitions in the same study were compared.

“The use of CT colonography for colon polyp surveillance has already been incorporated into management strategy at some centers, and is used for patients who have small polyps that pose no immediate risk and/or who wish to avoid colonoscopy,” the authors wrote. “Expansion of colorectal cancer screening programs in many countries and increasing population age and frailty suggest that surveillance with CT colonography will become an increasingly popular strategy for patients with low-risk polyps.