Soft-Tissue Tumor Reporting and Data System (ST-RADS): Prospective implementation and diagnostic accuracy

To assess clinical performance with prospective implementation of the ST-RADS scoring and compare it to traditional descriptions.

Following prospective implementation of ST-RADS in the electronic dictation system, fellowship-trained musculoskeletal radiologists classified soft tissue tumors and tumor-like lesions routinely. This IRB-approved audit study included 236 consecutive patients with MRIs from 01/2023 to 05/2024 at a tertiary care center. Data on patient demographics, imaging findings, lesion characteristics, biopsy results, treatments, and clinical outcomes were collected. Diagnostic accuracy of ST-RADS versus descriptive reporting was compared using McNemar test.

The cohort (mean age: 54.3 years; 61.02 % female) had both lower extremity (47.03 %) and upper extremity (37.29 %) masses. The distribution included ST-RADS I (2.12 %), ST-RADS II (36.86 %), ST-RADS III (22.03 %), ST-RADS IV (5.93 %), ST-RADS V (4.24 %), and ST-RADS VI (28.81 %). ST-RADS demonstrated superior diagnostic accuracy compared to descriptive reporting overall (99.15 % vs. 92.80 %, p=0.000) and also within biopsy-confirmed cases (97.73 % vs. 82.95 %, p=0.00098). For benign tumors (ST-RADS I-III), ST-RADS achieved 100 % accuracy compared to 93.75 % for descriptive reporting (p=0.003). For malignant tumors (ST-RADS IV-VI), ST-RADS also showed higher accuracy compared to descriptive reporting (96.74 % vs. 89.13 %), although difference did not reach statistical significance (p=0.109).

Prospective application of the ST-RADS scoring system demonstrated significantly superior accuracy compared to traditional descriptive methods. These results advocate for broader implementation of ST-RADS in clinical practice as a standardized method for classifying musculoskeletal soft tissue tumors, providing standardized management recommendations, and longitudinal data collection for tracking patient outcomes.