Polyp measurement and size categorisation by CT colonography: Effect of observer experience in a multi-centre setting

David Burling, Steve Halligan, Douglas G. Altman, Wendy Atkin, Clive Bartram, Helen Fenlon, Andrea Laghi, Jaap Stoker, Stuart Taylor, Roger Frost, Guido Dessey, Melinda De Villiers, Jasper Florie, Shane Foley, Lesley Honeyfield, Riccardo Iannaccone, Teresa Gallo, Clive Kay, Philippe Lefere, Andrew LoweFilipo Mangiapane, Jesse Marrannes, Emmanuele Neri, Giulia Nieddu, David Nicholson, Alan O'Hare, Sante Ori, Benedetta Politi, Martin Poulus, Daniele Regge, Lisa Renaut, Velauthan Rudralingham, Saverio Signoretta, Paola Vagli, Victor Van der Hulst, Jane Williams-Butt

Research output: Contribution to journalArticlepeer-review


The extent measurement error on CT colonography influences polyp categorisation according to established management guidelines is studied using twenty-eight observers of varying experience to classify polyps seen at CT colonography as either 'medium' maximal diameter 6-9 mm) or 'large' (maximal diameter 10 mm or larger). Comparison was then made with the reference diameter obtained in each patient via colonoscopy. The Bland-Altman method was used to assess agreement between observer measurements and colonoscopy, and differences in measurement and categorisation was assessed using Kruskal-Wallis and Chi-squared test statistics respectively. Observer measurements on average underestimated the diameter of polyps when compared to the reference value, by approximately 2- 3 mm, irrespective of observer experience. Ninety-five percent limits of agreement were relatively wide for all observer groups, and had sufficient span to encompass different size categories for polyps. There were 167 polyp observations and 135 (81%) were correctly categorised. Of the 32 observations that were miscategorised, 5 (16%) were overestimations and 27 (84%) were underestimations (i.e. large polyps misclassified as medium). Caution should be exercised for polyps whose colonographic diameter is below but close to the 1-cm boundary threshold in order to avoid potential miscategorisation of advanced adenomas.

Original languageEnglish
Pages (from-to)1737-1744
Number of pages8
JournalEuropean Radiology
Issue number8
Publication statusPublished - Aug 2006


  • Computerised tomography
  • CT colon
  • Virtual colonoscopy

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology


Dive into the research topics of 'Polyp measurement and size categorisation by CT colonography: Effect of observer experience in a multi-centre setting'. Together they form a unique fingerprint.

Cite this