BACKGROUND: This study evaluated the imaging features of ganglioneuroma (GN) and assessed the diagnostic value of the enhancement rate (ER) of CT for GN.

MATERIAL AND METHODS: We retrospectively reviewed records of 49 patients with histopathologically confirmed GN who underwent preoperative contrast-enhanced CT or MRI between 2010 and 2018. The independent samples t test and chi-square test were used. Receiver operating characteristic (ROC) curves were generated to analyze the diagnostic sensitivity (SE) and specificity (SP). Positive predictive value (PPV) and negative predictive value (NPV) were calculated.

RESULTS: The CT values were 32.59±3.61 Hounsfield units (HU) for plain scans, 38.87±5.09 HU for the arterial phase, and 54.26±8.14 HU for the venous phase, and the incidence of calcification and cysts was 32.6% and 10.2%, respectively. There was no significant difference in CT results and clinical parameters between mediastinal ganglioneuroma (MGN) and retroperitoneal ganglioneuroma (RGN) (p>0.05). The area under the curves (AUCs) for the arterial enhancement rate (AER), venous enhancement rate (VER), and AER/VER combined index in diagnosing GN were 0.735, 0.980, and 0.990, respectively. The VER of 0.2819 exhibited the SE and SP at 92.9% and 92.9%, respectively, to characterize the GN, whereas the AER of 0.1779 had SE and SP of 52.4% and 90.5%, respectively. The SE and SP for the combined index were 88.1% and 100%, respectively. The GN showed hypointensity on T1WI, hyperintense, or slightly high signal on T2WI with the linear hypointensity, and hyperintense on DWI.

CONCLUSIONS: A hypodense mass was observed for GN on plain scan and presented delayed enhancement on contrast enhancement. VER or AER/VER combination is more accurate than AER for the diagnosis of paravertebral GN.

Keywords: Ganglioneuroma, Magnetic Resonance Imaging, Tomography, Spiral Computed, Adolescent, Child, Child, Preschool, Contrast Media, Diagnosis, Differential, Image Processing, Computer-Assisted, Paraspinal Muscles, ROC Curve, Sensitivity and Specificity, Tomography, X-Ray Computed