Assessment of peroneal tendon lesions using 2-dimensional and 3-dimensional isotropic magnetic resonance imaging with surgical correlation.

Abstract

Accurate diagnoses of peroneal pathologies remains a challenge due to limitations of conventional 2D (dimensional) imaging, which can impact long-term patient outcomes. This study evaluates MRI accuracy and inter-reader reliability of peroneal compartment pathology for 2D and 3D MRI. A consecutive series of patients who underwent peroneal compartment surgery with preoperative 1.5- or 3.0-Tesla ankle MRIs from 2009 to 2024 included 32 scans (22 with 2D, 10 with 2D+3D) from 31 patients (ages 17-74 years, all genders). Three musculoskeletal readers blinded to surgical findings independently analyzed MRI scans for common peroneal tenosynovitis, peroneus brevis and peroneus longus tenosynovitis, tendinopathy, and tears. Inter-reader reliability and diagnostic performance measures were calculated. Using majority vote, overall accuracy, sensitivity, and specificity for peroneal tendons using 2D MRI were 80%, 81%, and 79%, respectively. Using 3D MRI sequences, whether in isolation or combination with 2D MRI, accuracy, sensitivity, and specificity increased to 85%, 88%, and 83%, respectively. The inter-reader reliability for peroneus brevis lesions was 0.45-0.75 for 2D, 0.25-0.35 for 3D, and 0.31-0.54 for combined 2D+3D, while for peroneus longus lesions, it was 0.45-0.90 for 2D, 0.20-0.71 for 3D, and 0.64-0.81 for combined 2D+3D scans. The inter-reader reliability for tenosynovitis ranged from 0.62-0.64 for 2D, 0.25-0.37 for 3D, and 0.57-0.66 for combined 2D+3D scans. The addition of 3D MRI to 2D high-resolution ankle MRI protocol or 3D MRI alone increases accuracy of peroneal compartment lesion detection with minor decrease in inter-reader reliability for peroneal brevis tendon assessment. Larger studies may help validate our findings.