Diagnostic Accuracy of 3T Unenhanced MR Imaging for the Detection of Parathyroid Lesions Before Surgery in Hemodialysis Patients With Secondary Hyperparathyroidism Comparison With 4DCT

MR examinations were performed by using the 3 Tesla unit (Siemens Magnetom Verio, Siemens Medical Solutions) with a dedicated eight-channel neurovascular phased array coil. The field of view was adjusted to include the area from the angle of the mandible to the level of the tracheal carina. The protocol included an axial and a coronal T1-weighted fast spin-echo (FSE) sequence, an axial T2-weighted FSE sequence with Dixon fat saturation, and a coronal and sagittal T2-weighted FSE sequence with inversion recovery fat saturation.

4DCT images were scanned with the 128-section dual-source multi-detector row unit (Somatom Definition Flash; Siemens Medical Solutions). All patients were scanned from the angle of the mandible to the carina. The 4DCT protocol included the acquisition of unenhanced images followed by the acquisition of arterial and delayed (venous) phase image series. Unenhanced images were obtained by using the following parameters: a collimation of 64×0.6 mm, 120 kVp for the deliverance of a dose-modulated amperage of 250 mAs, a pitch of 0.7, and a gantry rotation time of 0.28 seconds. The arterial and venous phase scans was performed in dual-energy mode with a dual source of 80 kVp and 140 kVp tube voltages using the following parameters: tube current-time product 230 and 89 mAs, respectively; collimation, 64 × 0.6 mm; rotation speed, 0.28 seconds; and helical pitch, 0.7. The arterial and venous phase scanning began automatically at 10 s and 35 s after the left common carotid artery reached the trigger attenuation threshold (100 HU) using automated scan-triggering software (CARE Bolus CT; Siemens Healthineers). The volume of iodinated contrast material (iohexol, Omnipaque 300; GE Healthcare, SH, China) was 70 mL, and the injection rate was 3 mL/s. The mean dual-source CT dose index volume was 34.5 mGy.

The sensitivity and specificity as well as the positive-predictive and negative-predictive values (PPV, NPV) of each imaging modality for preoperative parathyroid hyperplasia localization to the correct quadrant were compared. For the per-quadrant localization, if a lesion was suspected to be parathyroid hyperplasia by the imaging report and was correlated with the surgical and pathology results, this was interpreted as a true-positive lesion. Otherwise, if the lesion was not confirmed to be parathyroid hyperplasia during surgery, this was considered to be a false-positive case. If a lesion was not reported by the radiologist in the exact quadrant but was present during surgery and histopathologic examination, this was regarded as a false-negative lesion.