The diagnostic value of 4D MRI at 3T for the localization of parathyroid adenomas

Purpose: The aim of this study was to assess the feasibility of four-dimensional magnetic resonance imaging (4D MRI) at 3 T for the localization of parathyroid adenomas. Materials and Methods: Preoperative 4D MRI scans, encompassing dynamic contrast-enhanced (DCE) sequences and non-contrast enhanced (non-CE) sequences, including a T2-weighted multipoint Dixon (T2-mDixon) sequence, with in-phase, out-phase, and water-only images, were evaluated retrospectively in 41 patients with surgically proven parathyroid lesions. Two readers who were blinded to the surgical findings independently reviewed the images in two sessions (non-CE sequences alone and non-CE + DCE sequences). The MRI localization of the suspected adenoma in each session and the consensus interpretation of the MRI images, were compared with the surgical results and interobserver agreement was assessed. Results: By interpreting the non-CE sequences alone, reader 1 correctly localized 34 parathyroid lesions (sensitivity 81.0%, positive predictive value (PPV) 87.2%), and reader 2 correctly localized 34 parathyroid lesions (sensitivity 81.0%, PPV 91.9%). With the addition of DCE sequences, reader 1 correctly identified 35 parathyroid lesions (sensitivity 83.3%, PPV 87.5%), while reader 2 correctly identified 36 parathyroid lesions (sensitivity 85.7%, PPV 92.3%). Overall, MRI detected 38 parathyroid lesions (sensitivity 90.5%, PPV 95.0%). Interobserver agreement was slightly superior in non-CE + DCE sequences compared to non-CE sequences alone (kappa = 0.796 vs. kappa = 0.738). Conclusion: 4D MRI with DCE sequencing is a reliable method for the localization of parathyroid adenomas.