Purpose: To compare the diagnostic performance of iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL)-spoiled gradient-recalled echo (SPGR) with that of standard magnetic resonance (MR) arthrography sequences for detecting and grading cartilage lesions within the hip joint during MR arthrography.
Materials and methods: Following institutional review board approval, 67 consecutive hip MR arthrograms were retrospectively reviewed independently by three musculoskeletal radiologists and one musculoskeletal fellow. IDEAL-SPGR images and the two-dimensional images, the latter from the routine MR arthrography protocol, were evaluated at separate sittings to grade each articular surface of the hip joint. By using arthroscopy as the reference standard, the sensitivity and specificity of the two techniques for detecting and grading cartilage lesions were determined. The McNemar test was used to compare diagnostic performance. Interreader agreement was calculated using Fleiss κ values.
Results: For all readers and surfaces combined, the sensitivity and specificity for detecting cartilage lesions was 74% and 77%, respectively, for IDEAL-SPGR and 70% and 84%, respectively, for the routine MR arthrography protocol. IDEAL-SPGR had similar sensitivity (P = .12) to and significantly lower specificity (P < .001) than the routine MR arthrography protocol for depicting cartilage lesions. When analyzing the differences in sensitivity and specificity by reader, the two readers who had experience with IDEAL-SPGR had no significant difference in sensitivity and specificity for detecting cartilage lesions between the two sequences. For all readers and surfaces combined, IDEAL-SPGR had a higher accuracy in correctly grading cartilage lesion (P = .012-.013). Interobserver agreement for detecting cartilage lesions did not differ between the two techniques.
Conclusion: IDEAL-SPGR had similar sensitivity and significantly lower specificity for detecting cartilage lesions and higher accuracy for grading cartilage lesions than did a routine MR arthrography protocol; the lower specificity of IDEAL-SPGR for detecting cartilage lesions was not seen in experienced readers.
© RSNA, 2011.