Whole-tumor histogram analysis of synthetic MRI for the differentiation of benign and malignant soft-tissue tumors: a preliminary study.

OBJECTIVES

To determine the diagnostic potential of histogram features derived from synthetic magnetic resonance imaging (SyMRI) for the differentiation of benign and malignant soft-tissue tumors (STTs).

MATERIALS AND METHODS

Seventy-two patients with STTs of the extremities or trunk (29 benign, 43 malignant) were prospectively enrolled. Quantitative histogram features, extracted from T1, T2, and proton density (PD) images, were obtained by delineating the three-dimensional volume of interest (VOI). Patients were divided into a benign group and a malignant group according to their pathological results. Logistic regression was used to construct diagnostic models. Receiver operating characteristic (ROC) analysis was performed to assess diagnostic performance.

RESULTS

Several histogram features derived from SyMRI, such as PD_Energy, T1_Skewness, and T2_Kurtosis, were significantly elevated in patients with malignant STTs (p < 0.05). All the models constructed on the basis of SyMRI histogram features performed comparably to or better than the Clinical_Model (a model based on conventional MRI features). The Combined_Model (a model integrating histogram features from the PD, T1, and T2 maps) has the highest AUC (0.946).

CONCLUSION

Whole-tumor histogram analysis via SyMRI can distinguish benign from malignant STTs well. The Combined model, incorporating features from the PD, T1, and T2 models, is effective. Future efforts should focus on external validation across multiple institutions and prospective prognostic validation studies.

KEY POINTS

Question Differentiating between benign and malignant STTs is crucial for effective treatment, but current imaging methods struggle to reliably distinguish these tumor types. Findings Whole-tumor histogram analysis using SyMRI identified significant differences in quantitative features between benign and malignant STTs with high diagnostic accuracy. Clinical relevance This non-invasive imaging approach enhances diagnostic precision, reducing the need for biopsies and assisting clinicians in developing tailored treatment strategies for soft-tissue tumor patients.