MR grading system of osteochondritis dissecans lesions: Comparison with arthroscopy

Chia Hui Chen, Yi Sheng Liu, Pei Hsi Chou, Chin Chiang Hsieh, Chien Kuo Wang

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Objective: To assess the diagnostic performance of combined three-dimensional (3D) gradient-echo (GRE) T1-weighted and routine MR imaging protocol for the evaluation of osteochondritis dissecans (OCD). Materials and methods: This prospective study was approved by our institutional review board and all patients gave informed consent. Three-dimensional GRE MR sequence was added to the routine protocol performed on 40 consecutive patients (35 men, 5 women; age range, 12-57 years; mean age, 20 years) with 17 juvenile and 24 adult OCD lesions (27 in knees; 14 in elbows) which were confirmed by arthroscopy. Two independent musculoskeletal radiologists reviewed all MR images. The OCD lesions were classified into five stages by assessing the signal intensity of fragment-bone interface and the integrity of articular cartilage on MR images. Stage-IV and -V lesions were considered as unstable. The sensitivity, specificity, accuracy, and interobserver agreement (κ statistics) were calculated. Results: The sensitivity, specificity, and accuracy for detection of OCD instability were 100% (11 of 11), 100% (6 of 6), and 100% (17 of 17) in juvenile lesions; and 93% (14 of 15), 100% (9 of 9), and 96% (23 of 24) in adult lesions. The overall accuracy of MR findings in determining the staging was 90% (37 of 41) for reader 1 and 83% (34 of 41) for reader 2. Agreement between readers was substantial with a κ value of 0.75 for MR staging of OCD lesions. Conclusions: Three-dimensional GRE T1-weighted MR imaging combined with the routine sequences demonstrates excellent diagnostic capabilities in detecting unstable OCD lesions.

Original languageEnglish
Pages (from-to)518-525
Number of pages8
JournalEuropean Journal of Radiology
Volume82
Issue number3
DOIs
Publication statusPublished - 2013 Mar 1

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging

Cite this