The value of the nomogram model based on high-frequency ultrasound and DWI in the diagnosis of benign and malig-nant superficial soft tissue tumors

WANG Yin; XU Yali; YANG Fan

doi:10.20258/j.cnki.1006-9011.2025.12.026

Abstract：Objective To explore the value of nomogram models constructed based on high-frequency ultrasound and diffusion-weighted magnetic resonance imaging(DWI)in the diagnosis of benign and malignant superficial soft tissue tumors(STTs),so as to provide a reference for clinical diagnosis and treatment.Methods A total of 309 patients with superficial STTs(309 tumors in total)were selected.The tumors were confirmed to be benign(237)or malignant(72)by surgical pathol-ogy or biopsy,and were included in the benign group and malignant group,respectively.The clinical manifestations and high-frequency ultrasound and DWI imaging results of the two groups were compared.LASSO regression and multivariate Logistic re-gression were used to analyze the independent factors affecting the malignancy of superficial STTs,and a nomogram model was established based on these independent factors to predict the malignancy of superficial STTs.The receiver operating characteris-tic(ROC)curve was used to evaluate the discriminative power of the nomogram,and the calibration curve was used to analyze the calibration of the nomogram.Results The malignant group had a higher proportion of patients with tumor diameter≥5 cm,ir-regular edge morphology,unclear boundary,blood flow type Ⅳ,uneven T2WI signal intensity,uneven enhancement pattern,and DWI visual signal>50%than the benign group.The apparent diffusion coefficient(ADC)and relative apparent diffusion coeffi-cient(rADC)were lower in the malignant group compared to the benign group(all P<0.05).LASSO regression analysis showed that the variables for diagnosing malignant superficial STTs with excellent model performance and minimal influencing factors un-der the optimal penalty coefficient λ were as follows:edge morphology,blood flow classification,T2WI signal,enhancement pat-tern,ADC,rADC,DWI visual signal.Logistic regression analysis showed that edge morphology,blood flow pattern,T2WI signal intensity,enhancement mode,ADC value,rADC value and DWI visual signal were all independent factors for malignant superfi-cial STTs(P<0.05).The nomogram model and ROC curve were constructed to show that the AUC for diagnosing benign or malig-nant superficial STTs was 0.963(95%CI:0.943-0.983).The calibration curve showed good consistency between the model's di-agnosis and actual observation results.Conclusion There are significant differences in the edge morphology,blood flow classifi-cation,T2WI signal,enhancement pattern,ADC,rADC,and DWI visual signals between benign and malignant superficial STTs.The nomogram model constructed based on high-frequency ultrasound and DWI has high diagnostic performance,discrimination,and calibration,thereby providing a reliable basis for the diagnosis of benign and malignant superficial STTs.