Predictive value of nomogram model based on quantitative ultrasound parameters for breast cancer

TANG Jin; ZHANG Bo; MIAO Runqin

doi:10.19300/j.2025.L22070

Abstract：Objective To explore the predictive value of a nomogram model based on quantitative ultrasound parameters for breast cancer.Methods Clinical and imaging data of 135 female patients with breast tumors confirmed by biopsy or surgical pathology who underwent ultrasound examination were retrospectively collected.Patients were randomly divided into a modeling set(95 cases)and an internal validation set(40 cases)at a ratio of 7∶3.Additionally,30 patients with breast tumors confirmed by biopsy or surgical pathology who underwent ultrasound examination were collected as an external validation set.The pathological diagnosis results were used as the"gold standard"to divide the modeling set into a benign group(60 cases)and a malignant group(35 cases).Color doppler ultrasound data were collected,including blood flow grading and hemodynamic parameters,along with elasticity imaging parameters.Multivariate Logistic regression was used to identify independent risk factors for breast cancer.A nomogram prediction model of breast cancer was constructed using R software.The predictive performance of the model was assessed using the area under the receiver operating characteristic(ROC)curve(AUC).The goodness-of-fit of the model was evaluated using the Hosmer-Lemeshow test.Results In the modeling set,the sensitivity,specificity,and accuracy of ultrasound diagnosis for breast cancer were 94.29%,93.33%,and 93.68%respectively.Multivariate Logistic regression analysis revealed that a peak systolic velocity(PSV)≥15.00 cm/s,maximum elastic modulus(Emax)≥86.4 kPa,elasticity ratio(Eratio)≥4.35,and standard deviation of elastic modulus(Esd)≥17.0 kPa were independent risk factors for breast cancer(all P＜0.05).The ROC curve showed good predictive performance of the model in the modeling,internal validation,and external validation cohorts,with AUCs of 0.900,0.926,and 0.820,respectively.Calibration curve demonstrated good agreement between predicted and actual outcomes in all three cohorts,and the Hosmer-Lemeshow test showed no significant deviation between predicted and observed probabilities(all P＞0.05).Conclusion The breast cancer nomogram model based on PSV,Emax,Eratio and Esd quantitative ultrasound parameters has high prediction performance and can provide an objective basis for the diagnosis of breast cancer.