Multirater agreement on arthroscopic image quality

G.J.M. Tuijthof, M. Abbink, I.N. Sierevelt, C.N. Van Dijk

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Scopus)

Abstract

In arthroscopy (minimally invasive orthopaedic surgery), the view is frequently disturbed. To optimize the view, quantification of the arthroscopic image quality is important. Thereto, disturbances were categorized as bleeding, air bubbles, turbidity (synovial fluid), loose fibrous tissue, and attached fibrous tissue, which cover the arthroscopic image area. The goal is to determine the percentages of disturbance coverage for which the view is acceptable.

Thirty-two short films of the five disturbances were selected from arthroscopic knee procedures. The films showed disturbances covering different percentages of the image area. Thirty-nine orthopaedic surgeons were asked to judge whether or not the view of each film was acceptable. Multiple-choice questions on irrigation and disturbances were asked.

A clear transition from acceptable to unacceptable view was found for bleeding (5 per cent of the covered area was acceptable; 25 per cent was not acceptable), and air bubbles (10 per cent was acceptable; 20 per cent was not acceptable). Loose fibrous tissue showed a gradual transition where 25 per cent was still accepted by a third of the surgeons. Turbidity and attached fibrous tissue were tolerated up to 50 per cent by half of the surgeons. Surgeons using a mechanical pump tolerated a lower percentage of synovial fluid (p<0.05). The most intolerable disturbance was bleeding. The results were consistent and will be used for computerized detection of disturbances.

Original languageEnglish
Pages (from-to)179-187
Number of pages9
JournalProceedings of the Institution of Mechanical Engineers. Part H: Journal of engineering in medicine
Volume223
Issue number2
Early online date18 Dec 2008
DOIs
Publication statusPublished - 1 Feb 2009
Externally publishedYes

Keywords

  • Arthroscopy
  • Image quality
  • Irrigation
  • Knee joint
  • Video analysis

Fingerprint

Dive into the research topics of 'Multirater agreement on arthroscopic image quality'. Together they form a unique fingerprint.

Cite this