To assess the diagnostic performance of US and MRI in predicting malignancy of soft tissue masses by using a scoring system.
A total of 120 cases of pathologically confirmed soft tissue masses (71 cases of malignant lesions and 49 cases of benign lesions) were enrolled. All patients underwent ultrasound and MRI examination prior to biopsy or surgical excision. A scoring system based on the parameters of conventional US and MRI to distinguish malignant and benign masses was established by the regression model. The receiver operating characteristic (ROC) analysis was used to evaluate the diagnostic performance of US and MRI.
Multivariate analysis showed that margin, maximum diameter, and vascular density were independent predictors for malignancy found by US, while maximum diameter, margin, and affected peripheral soft tissue were independent predictors for malignancy found by MRI. The mean scores of the benign and malignant groups were 2.8 ± 1.6, 5.1 ± 1.1 on US and 1.3 ± 1.2, 3.5 ± 0.9 on MRI. Based on the cut-off score of 3.5 and 2.5 calculated by ROC analysis, US and MRI had 92% and 87% sensitivity, 72% and 76% specificity, 86% and 89% accuracy, respectively. The combination of these two modalities achieved the sensitivity of 91%, specificity of 82%, and accuracy of 93%.
Both US and MRI can provide valuable information about the differential diagnosis between benign and malignant soft tissue masses. The combination of the two imaging-based scoring systems can increase the diagnostic performance, especially in specificity.