Ultrasound radiomics-based dynamic nomogram to predict histologic grade in soft tissue sarcoma: a multicenter cohort study.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
OBJECTIVES: To predict histologic grade of soft tissue sarcoma (STS) with preoperative ultrasound images, aiding in the selection of personalized treatment plans and improving long-term prognosis. METHODS: In total, 238 patients with histologically proven STS were retrospectively enrolled from April 2016 to December 2023 and divided into the training and internal validation cohorts. Seventy patients were prospectively enrolled from 3 centers between January 2024 and December 2024 as the external validation cohort. Radiomics features were extracted from preoperative grayscale ultrasound images. The dynamic nomogram (DynNom) was developed by using multivariable logistic regression analysis. Predictive performance was evaluated with the receiving operating characteristic curve, calibration curve, Hosmer-Lemeshow test, decision curve analysis (DCA), and clinical impact curve (CIC). RESULTS: The DynNom based on clinical-US characteristics (metastasis status, echogenicity, fascia layer, and vascularity) and radiomics features yielded an optimal AUC of 0.915 (95% CI, 0.873-0.947), 0.87 (95% CI, 0.79-0.93), and 0.90 (95% CI, 0.80-0.96) for predicting the STS histologic grade in the training, internal, and external validation cohorts, respectively. The DynNom outperformed the conventional model and radiomics model (P < .05). Calibration curves and Hosmer-Lemeshow tests indicated its satisfactory calibration ability. DCA confirmed that the DynNom outperformed other models in overall net benefit, meanwhile CIC suggested that the DynNom had great clinical applicability in predicting histologic grade. CONCLUSIONS: The dynamic nomogram is a practical tool that could predict the histologic grade of STS, which might help clinicians to screen histologic high-grade STSs as neoadjuvant treatment candidates. ADVANCES IN KNOWLEDGE: The dynamic nomogram had the potential to accurately predict histologic grade in STS patients before surgery. High-risk patients defined by the dynamic nomogram were potential candidates for preoperative radiotherapy and neoadjuvant chemotherapy.