CHICAGO— As CT colonography (CTC) gains wider acceptance as an effective and less-invasive exam for colon cancer screening, advanced visualization tool development is keeping pace by offering applications that will allow interpreting clinicians the ability to efficiently and effectively manage the virtual colonoscopy workflow.
Dipti K. Lenhart, MD, of New York University Medical Center in New York City discussed the initial results of the application of one of these software tools, panoramic 3D, this week at the 94th annual meeting of the Radiological Society of North America.
“We assessed the evaluation times and accuracy of unidirectional panoramic 3D interpretation to traditional 2D and bi-directional 3D endoluminal techniques,” she said.
Using a team of three CTC readers, the researchers asked the image interpreters to report on a pre-selected data set using the three imaging techniques. The researchers selected 81 CTC exams that were performed on a 64-slice CT system after standard bowel preparation. Lenhart reported that the data set consisted of 50 studies with no polyps, 18 with at least one polyp of 6 to 9 mm in size, and 13 with at least one polyp ? 10 mm. As a reference standard, the patients underwent same-day colonoscopy with segmental unblinding, she said.
“The panoramic view unfolds haustra allowing visualization of both sides of the folds during a single flythrough,” Lenhart said. “This ‘flattening’ of the folds allows for the visualization of polyps in one pass; unlike traditional endoluminal 3D which requires a bi-directional technique to visualize all polyps. Compared to traditional 3D interpretation, a panoramic 3D interpretation allows a single flythrough to be performed with similar sensitivity and in less time.”
Interpretations of the CTC exams were performed on a Siemens Healthcare Leonardo workstation. The research team used the statistical method of mixed model analysis of variance (ANOVA) to compare the three techniques with respect to evaluation time and logistic regression for correlated data was used to compare techniques.
“Sensitivity was assessed on a per-polyp and per-patient basis, whereas specificity was valuated per-patient only,” Lenhart noted.
Among the three readers, 2D interpretation had a mean time of 8.85 minutes; endoluminal 3D clocked in at 14.74 minutes; and panoramic 3D took 12.19 minutes.
Sensitivity of each interpretation technique per polyp and per patient was 68.4 percent and 76.7 percent for primary 2D, 78.9 percent and 93.3 percent for endoluminal 3D; and 78.9 percent and 86.7 percent for panoramic 3D, Lenhart said. Specificity for the three techniques was 98 percent for 2D and panoramic 3D, with endoluminal 3D reporting 90.2 percent.
Lenhart said her study was limited by a small sample size, the use of only three readers, and that none of the readers had familiarity with the panoramic 3D application prior to the study. However, she believes the results are promising.
“Primary 2D interpretations are quicker than 3D, but sensitivity for 2D is significantly lower,” she noted. “Panoramic and traditional 3D techniques have similar sensitivity and specificity, with panoramic interpretation requiring significantly less time.”