Purpose or Case Report: This study compares the accuracy of three MRI acquisitions commonly used to assess intra articular pathology in identifying labral, transition zone, and articular cartilage injuries.
Methods & Materials: Two musculoskeletal radiologists evaluated high resolution (HR), 3-dimensional with multiplanar reconstruction (MPR), and radially reformatted (RR) magnetic resonance images (MRI) of 33 patients who later underwent surgical treatment for FAI. Raters reviewed each MRI sequence for the presence of labral tear or injury to the transition zone or true acetabular cartilage as well as the size of the injury. An orthopedic surgeon reviewed intra operative images to assess for intra articular injury. Bland-Altman methods were used to estimate agreement between the percentage of hips affected by injury as measured by MRI relative to direct visualization. Bias or mean difference between MRI and arthroscopic visualization as well as the limit of agreement was also calculated for each of the three formats in assessing size of injury. The 12 o’clock to 2 o’clock positions within the acetabulum were used for comparison, the most common locations for intra-articular pathology in femoroacetabular impingement (FAI).
Results: MPR was most accurate at identifying labral injuries (accuracy: 88%, sensitivity: 92%, specificity: 25%) compared to HR (accuracy: 82%, sensitivity: 87%, specificity: 0%) and RR (accuracy: 83%, sensitivity: 89%, specificity: 0%). MPR (accuracy: 38%, sensitivity: 38%, specificity: 50%) and RR formats (accuracy: 38%, sensitivity: 38%, specificity: 50%) were also more accurate at identifying transition zone injuries compared to HR images (accuracy: 29%, sensitivity: 27%, specificity: 100%). HR was more accurate at identifying articular cartilage injuries (accuracy: 65%, sensitivity: 40%, specificity: 76%) compared to MPR (accuracy: 56%, sensitivity: 30%, specificity: 67%) and RR (accuracy: 61%, sensitivity: 40%, specificity: 70%) formats. MPR had performed best in determining the size of labral injury (bias -0.10), while MPR and RR performed best in identifying transition zone (bias -0.06) and acetabular cartilage (bias -0.31) injuries.
Conclusions: Labral injuries were best identified with the MPR format while transition zone cartilage injuries were best identified with MPR and RR formats, and the HR format was best to identify true articular cartilage pathology. 3D with MPR and RR are a less commonly utilized though beneficial approach in the evaluation of the labrum and transition zone injuries.