ABSTRACT
PURPOSE: The purpose of this study is to determine the reliability of standard magnetic resonance imaging (MRI) evaluation of AIIS morphology compared with three-dimensional (3D) computerized tomography (CT) (reference standard). METHODS: Sixty hips in 30 patients met the inclusion criteria. The images were reviewed and classified by two fellowship-trained orthopedic surgeons. A second imaging viewing session was conducted in the same manner for validation of AIIS evaluation. The agreement and accuracy indices between the two raters were calculated for each imaging modality (inter-rater agreement) as well the agreement across the imaging modality for each rater (intermethod agreement). RESULTS: The inter-rater agreement for the morphological evaluation of the AIIS for the first session according to 3D CT was .553 (P < .05) and by means of MRI was .0163 (P < .05). The inter-rater agreement for the second session by means of 3D CT was .449 (P < .05) and according to MRI was 0 (P < .05). The inter-method agreement for rater 1 for the first session was .04 (P < 0.05), while the agreement for rater 2 was .016 (P < .05). The intermethod agreement for rater 1 on the second session was .35 (P < 0.05), while that of rater 2 was .297(P < .05). The overall accuracy of MRI compared to 3D CT for rater 1 for the first session was .531, .490, and .959 for types I, II, III respectively and .551, .531, and .980 for the second session for types I, II, and III respectively. The overall accuracy of MRI compared to 3D CT for rater 2 for the first session was .551, .469, and .918 for types I, II, III respectively and .633, .592, and .918 for the second session for types I, II, and III, respectively. CONCLUSION: MRI evaluations and subsequent classifications of AIIS morphology demonstrated a poor to slight correlation compared with that of the reference standard of 3D CT. LEVEL OF EVIDENCE: Level II, retrospective diagnostic study.
