What semi-quantitative scoring instrument for knee OA MRI should you use?

D. J. Hunter, S. Zaim, T. J. Mosher

Research output: Contribution to journalEditorialpeer-review

3 Scopus citations

Abstract

Multiple techniques have been used to assess synovial morphology and change on MRI in OA. Broadly speaking these methods are divided into quantitative and semi-quantitative methods. 1. Quantitative measurements use computer-aided image processing to assess joint quantification (cartilage morphometry, bone volume, bone marrow lesion volume, meniscal position and volume, synovial volume, etc). In contrast to quantitative measures, semi-quantitative image analysis is typically much more observer dependent and generates grades or scales rather than truly continuous output. Multiple methods for semi-quantitative multi-feature assessment of the knee using conventional MRI acquisitions exist. 2-5. These instruments provide for broad assessment of the whole joint and derive from knowledge from reading as to what joint features are morphologically abnormal. They are labour- and expertise-intensive compared to more automated methodologies. As a consequence of their reader dependence, precision and reliability results have not been as favourable for these instruments as their quantitative cousins. These instruments are generally based on past perceptions of what should be considered an important feature and therefore can bias future research. This said they do provide an important tool especially when quantitative methodologies are lacking or have their own inherent limitations.

Original languageEnglish (US)
Pages (from-to)1363-1364
Number of pages2
JournalOsteoarthritis and Cartilage
Volume18
Issue number11
DOIs
StatePublished - Nov 2010

All Science Journal Classification (ASJC) codes

  • Rheumatology
  • Biomedical Engineering
  • Orthopedics and Sports Medicine

Fingerprint

Dive into the research topics of 'What semi-quantitative scoring instrument for knee OA MRI should you use?'. Together they form a unique fingerprint.

Cite this