Home Comparison of 2D and 3D Views for Evaluation of Flat Lesions in CT Colonography
Post
Cancel

Comparison of 2D and 3D Views for Evaluation of Flat Lesions in CT Colonography

Posted Dec 31, 2009 Updated Apr 7, 2023
By Antonella Lostumbo MD , Christian Wanamaker MD PhD , Joy Tsai MD , Kenji Suzuki PhD , Abraham H. Dachman MD
8 min read

Rationale and Objectives

Flat lesions in the colon may result in false-negative computed tomography colonography interpretations. It is unknown whether flat lesions are better measured on two-dimensional (2D) or three-dimensional (3D) images and which settings are optimal for enhanced reproducibility and decreased variability. We evaluated these factors to determine whether 2D or 3D is best for flat lesion measurements.

Methods and Materials

Eighty-eight lesions in 66 patients from a previously published clinical trial were analyzed. Lesions were viewed with four methods including 2D at three window/level settings and 3D endoluminal view. Lesions in either supine or prone were counted as one dataset. Long axis and height were measured. Criteria of “height” (≤3 mm high) or “ratio” (height ≤half the long axis) were applied. A subset of lesions was subject to inter- and intra-observer variability analysis.

Results

With the “height” criterion, more datasets were classified as flat in 2D flat ( n = 76), 2D soft tissue ( n = 82), and 3D ( n = 73) views than in the 2D lung ( n = 49) view. If long axis is used as the key metric, endoluminal 3D (12.1%) views significantly showed the least inter-observer variability compared to lung (18.9%) or soft tissue (20.2%) views. Intra-observer variability was low overall for all methods.

Conclusion

When characterizing lesions as flat, a consistent viewing method should be used. To minimize inter-observer variability (such as when following a patient over time), it is best to use the ratio criterion for flat lesion definition incorporating the single longest dimension on 3D views as the key metric.

Flat lesions of the colon are a potentially important source of false negative computed tomography colonography (CTC) interpretations . Several different definitions of flat lesions have been proposed including height <3 mm, a definition recommended in a consensus opinion and a height less than one-half the width (as seen on two-dimensional [2D] views, or long axis as seen on three-dimensional [3D] views). In many CTC investigations and clinical reports, endoscopists and radiologists may classify a lesion as “flat” based on subjective visual impression without defining the term in their methods or without measuring the lesion .

There are no clinical data to indicate whether flat lesions are better measured on 2D views or the 3D endoluminal views. On 2D views, the optimal window and level settings to visualize or to measure a flat lesion have not been determined.

Get Radiology Tree app to read full this article<

Materials and methods

Study Population

Get Radiology Tree app to read full this article<

Observers

Get Radiology Tree app to read full this article<

Lesion Assessment

Get Radiology Tree app to read full this article<

Figure 1, Conspicuity scale demonstrating eight different lesions in the four views. The top row is two dimensional in lung (W = 1500, L = -200) and in three-dimensional in the bottom row. In each case, the same scale applies as follows: 0 = lesion was not visible at all and could not be measured; 1 = lesion was barely visible and it was hard to visualize its borders; 2 = lesion was somewhat visible and paging through the slices on computed tomography or rotating the viewing angle may have helped in visualizing the lesions' borders, especially if the lesion was located on a fold; 3 = lesion was relatively visible, but had a few minor limitations; 4 = lesion was highly visible with very discrete borders.

Get Radiology Tree app to read full this article<

Lesion Conspicuity and Measurement

Get Radiology Tree app to read full this article<

Statistical Experimental Design

Get Radiology Tree app to read full this article<

Intra-observer Variability

Get Radiology Tree app to read full this article<

Inter-observer Variability

Get Radiology Tree app to read full this article<

Results

Definition

Get Radiology Tree app to read full this article<

Figure 2, Flow chart of lesion analysis. Of the 88 lesions in 36 patients that met one of the two criteria for flat lesions in at least one viewing methods, 65 were adenomas, 19 were hyperplastic polyps, 3 were other histology, and 1 was normal mucosa.

Table 1

OC, Long Axis, and Height Measurements in the Two-dimensional Lung View Comparing 68 Adenomas (121 Datasets) versus 26 Non-adenomas (39 Datasets)

Adenoma ( n = 68) Non-adenoma ( n = 26) Mean (mm) Standard Deviation Mean (mm) Standard Deviation OC 8.83 3.08 7.48 1.78 Long axis 9.09 2.93 8.26 2.46 Height 3.81 1.32 3.28 0.65

OC, optical colonoscopy; CTC, computed tomography colonography.

One dataset was not visualized in the two-dimensional lung view. Ninety-four lesions (161 datasets) in 73 patients seen on CTC had flat or sessile type morphologies that were measured to determine if they fit a proposed flat lesion definition. Non-adenomas included those classified as normal, hyperplastic, or other. A lesion in either supine or prone view is counted as one dataset. OC measurements ranged from 6–18 mm in size.

Figure 3, Illustrations of the measurements of height and long axis of the lesions. The first row shows three flat lesions in two-dimensional lung tissue, soft tissue, and flat views measured on an axial plane (from left to right). The second row shows flat lesions in the three-dimensional (3D) endoluminal view. After optimizing the magnification and cutting plane, manual rotation of the 3D image was used to determine the proper perspective to measure maximal height. In difficult cases, trial and error was used to find the maximal height.

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Table 2

The Number of Datasets that Fit Two Different Definitions of a Flat Lesion

2D Lung 2D Soft Tissue 2D Flat 3D_n_ = 141n = 134n = 135n = 141 Fits less than or equal to 3 mm height definition (% of total) 49 (34.8%) 82 (61.2%) 76 (56.3%) 73 (51.8%) Fits height less than half the longest axis definition (% of total) 116 (82.3%) 110 (82.1%) 121 (89.6%) 119 (84.4%)

2D, two-dimensional; 3D, three-dimensional.

Seven and six datasets could not be visualized in the 2D soft tissue and flat views, respectively.

Get Radiology Tree app to read full this article<

Conspicuity

Get Radiology Tree app to read full this article<

Comparison of Viewing Methods

Get Radiology Tree app to read full this article<

Figure 4, Inter-observer long axis and height measurement average % differences versus view. In the long axis measurement, a post-hoc two-tailed t -test showed a statistically significant difference between the 3D endoluminal view (12%) compared to the 2D lung (19%; P = .016) or to 2D soft tissue (20%; P = .0099) viewing methods.

Table 3

Statistical Analysis of Height Measurements Taken by Observer #1 Compared to Observer #2 in the 2D Lung, 2D Soft Tissue, and 3D Endoluminal Views

Measurement Parameter Mean Percentage Difference (%) ∗ Mean Difference (mm) † Standard Deviation of Difference 95% Bland-Altman Limits of Agreement ‡ 2D Lung 24.1 0.26 1.07 -1.84, 2.36 2D soft tissue 21.8 0.26 0.84 -1.39, 1.91 3D 24.1 0.47 1.36 -2.21, 3.14

2D, two-dimensional; 3D, three-dimensional.

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Table 4

Statistical Analysis of Long Axis Measurements Taken by Observer #1 Compared to Observer #2 in the 2D Lung, 2D Soft Tissue, and 3D Endoluminal Views

Measurement Parameter Mean Percentage Difference (%) ∗ Mean Difference (mm) † Standard Deviation of Difference 95% Bland-Altman Limits of Agreement ‡ 2D Lung 18.9 0.94 2.32 -3.61,5.48 2D soft tissue 20.2 0.92 2.22 -3.43, 5.28 3D 12.1 0.46 2.11 -3.67, 4.59

2D, two-dimensional; 3D, three-dimensional.

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Figure 5, Bland-Altman plot of the difference between polyp long axis measurements of Observer #1 and Observer #2 versus mean long axis measurement of lesions in the 3D endoluminal view.

Get Radiology Tree app to read full this article<

Lesion Measurements

Get Radiology Tree app to read full this article<

Figure 6, Intra-observer measurement average % differences versus view. There was a statistically significant difference (analysis of variance, P = .02) in the average percent difference in intra-observer height measurements among the two-dimensional (2D) lung (8.9%), 2D soft tissue (11.5%), 2D flat (12.4%), and three-dimensional (3D) endoluminal (18.6%) viewing methods. A post-hoc two-tailed t -test showed a significant difference between the 2D lung and 3D endoluminal viewing methods ( P = .025).

Get Radiology Tree app to read full this article<

Discussion

Definition

Get Radiology Tree app to read full this article<

Comparison of Viewing Methods

Get Radiology Tree app to read full this article<

Conspicuity and Measurement

Get Radiology Tree app to read full this article<

Lesion Follow-up

Get Radiology Tree app to read full this article<

Source of Variability

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Acknowledgments

Get Radiology Tree app to read full this article<

Appendix 1

Get Radiology Tree app to read full this article<

Get Radiology Tree app to read full this article<

Figure 1A, Bland-Altman plot of the difference between polyp height measurements of Observer #1 and Observer #2 versus mean height measurement of lesions in the three-dimensional endoluminal view.

Figure 2A, Bland-Altman plot of the difference between polyp long axis measurements of Observer #1 and Observer #2 versus mean long axis measurement of lesions in the two-dimensional lung view.

Figure 3A, Bland-Altman plot of the difference between polyp height measurements of Observer #1 and Observer #2 versus mean height measurement of lesions in the two-dimensional lung view.

Figure 4A, Bland-Altman plot of the difference between polyp long axis measurements of Observer #1 and Observer #2 versus mean long axis measurement of lesions in the two-dimensional soft tissue view.

Figure 5A, Bland-Altman plot of the difference between polyp height measurements of Observer #1 and Observer #2 versus mean height measurement of lesions in the two-dimensional soft tissue view.

Get Radiology Tree app to read full this article<

References

  • 1. Fidler J.L., Johnson C.D., MacCarty R.L., et. al.: Detection of flat lesions in the colon with CT colonography. Abdom Imaging 2002; 27: pp. 292-300.

  • 2. Park S.H., Ha H.K., Kim A.Y., et. al.: Flat polyps of the colon: detection with 16-MDCT colonography—preliminary results. AJR Am J Roentgenol 2006; 186: pp. 1611-1617.

  • 3. Sawada T., Hojo K., Moriya Y.: Colonoscopic management of focal and early colorectal carcinoma. Balliere Clin Gastroenterol 1989; 3: pp. 627-645.

  • 4. Kudo S., Kashida H., Tamura T., et. al.: Colonoscopic diagnosis and management of nonpolypoid early colorectal cancer. World J Surg 2000; 24: pp. 1081-1090.

  • 5. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon. Gastrointest Endosc 2003; 58: pp. S3-S43.

  • 6. Soetikno R., Friedland S., Kaltenbach T., et. al.: Gastroenterology 2006; 130: pp. 566-576.

  • 7. Pickhardt P.J., Nugent P.A., Choi J.R., et. al.: Flat colorectal lesions in asymptomatic adults: implications for screening with CT virtual colonoscopy. AJR Am J Roentgenol 2004; 183: pp. 1343-1347.

  • 8. Endoscopic Classification Review Group: Update on the Paris classification of superficial neoplastic lesions in the digestive tract. Endoscopy 2005; 37: pp. 570-578.

  • 9. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002. Gastrointest Endosc 2003; 58: pp. S3-S43.

  • 10. Zalis M.E., Barish M.A., Choi J.R., et. al.: Working Group on Virtual Colonoscopy. CT colonography reporting and data system: a consensus proposal. Radiology 2005; 236: pp. 3-9.

  • 11. Kim D.H., Pickhardt P.J., Taylor A.J., et. al.: CT colonography versus colonoscopy for the detection of advanced neoplasia. N Engl J Med 2007; 357: pp. 1403-1412.

  • 12. Rockey D., Paulson E., Niedzwiecki D., et. al.: Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: prospective comparison. Lancet 2005; 365: pp. 305-311.

  • 13. Doshi T., Rusinak D., Halvorsen R.A., et. al.: CT colonography: false-negative interpretations. Radiology 2007; 244: pp. 165-173.

  • 14. Park S.H., Lee S.S., Choi E.K., et. al.: Flat colorectal neoplasms: definition, importance, and visualization on CT colonography. Zalis ME, Barish MA, Choi JR, et al. Working Group on Virtual Colonoscopy 2007; 188: pp. 953-959.

  • 15. Pickhardt P.J., Choi J.R., Hwang I., et. al.: Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003; 349: pp. 191-200.

  • 16. Young B.M., Fletcher J.G., Paulsen S.R., et. al.: Polyp measurement with CT colonography: multiple-reader, multiple-workstation comparison. AJR Am J Roentgenol 2007; pp. 122-129.

  • 17. Summers R.M., Frentz S.M., Liu J., et. al.: Conspicuity of colorectal polyps at CT colonography: visual assessment, CAD performance, and the important role of polyp height. Acad Radiol 2009; 16: pp. 4-14.

Academic Radiology, Volume 17, Issue 1
Journals General Radiology
This post is licensed under CC BY 4.0 by the author.
Recently Updated
Contents