Differentiating between malignant and benign lesions on the basis of MR images depends on the experience of the radiologist. For non-experts, we aimed to develop a simplified systematic MRI approach that uses depth, size and heterogeneity on T2 weighted MR images (T2WI) to differentiate between malignant and benign lesions, and evaluated its diagnostic accuracy.
MR images of 266 patients with histologically proven soft-tissue tumours of the extremities (102 malignant, 164 benign) were analysed according to depth (superficial or deep), size (<50, ≥50 mm) and signal intensity (homogeneous or heterogeneous) on T2WI, to determine the ability of each to predict benign and malignant tumours. These three parameters were categorised into systematic combinations of different orders of application, and each combination was assessed for its ability to differentiate between benign and malignant lesions.
Univariate analysis showed that depth, size and heterogeneity on T2WI differed significantly between benign and malignant masses (p<0.0001 each). Multiple logistic regression analysis, however, showed that depth was not helpful in distinguishing benign from malignant lesions. The systematic combination of signal intensity, size and depth, in that order, was superior to other combinations, resulting in higher diagnostic values for malignancy, with a sensitivity of 64%, a specificity of 85%, a positive predictive value of 32%, a negative predictive value of 59% and an accuracy of 77%.
A simplified systematic imaging approach, in the order signal intensity, size and depth, would be a reference to distinguish between benign and malignant soft-tissue tumours for non-experts.