Whole-Tumor Histogram Analysis of Synthetic MRI for the Differentiation of Benign and Malignant Soft-Tissue Tumors

Motivation: Soft-tissue tumors (STTs) display considerable heterogeneity, making accurate differentiation between benign and malignant forms challenging. This study pioneers the application of synthetic MRI (SyMRI) to address these diagnostic complexities in STTs. Goal(s): To evaluate the diagnostic effectiveness of SyMRI-based histogram analysis for distinguishing benign from malignant STTs by constructing models from histogram features derived from SyMRI data. Approach: A cohort of 72 STTs patients underwent SyMRI scanning, enabling extraction of histogram features analyzed via logistic regression to develop diagnostic models. Results: SyMRI-based models, particularly the Combined_Model integrating PD, T1, and T2 map features, demonstrated high diagnostic accuracy, achieving an AUC of 0.946. Impact: This inaugural application of SyMRI for STTs presents a non-invasive diagnostic alternative with high accuracy, supporting clinical decision-making and personalized medicine in STT diagnosis.