摘要
Abstract
Objective:To investigate the value of ultrasonography in prognosis of prostate cancer treated with iodine-125 particle implantation.Methods:The data of 283 patients with PCa from August 2022 to January 2024 were retrospectively analyzed.The enrolled patients were randomly divided into the training set(n=198)and the test set(n=85)at a ratio of 7∶3.tPSA＞10 ng/mL was defined as poor prognosis,and tPSA＜10 ng/mL was defined as good prognosis.Logistic regression was used to analyze clinical covariates.Intraclass correlation coefficient(ICC)and linear correlation test were used to establish LASSO regression to select the best features,and support vector machine(SVM)was used to construct the radiomics signature.Logistic regression was used to construct a visual nomogram model.The ROC curve was used to evaluate the discrimination,the calibration curve was used to evaluate the accuracy,and the decision curve was used to evaluate the clinical benefit.Results:The tPSA,PSAD,radiomics score,and prostate volume are independent factors associated with a favorable prognosis after implantation of iodine-125 particles in patients with prostate cancer(P＜0.05).For each unit increase in tPSA,PSAD,and radiomics score,the likelihood of a favorable prognosis for patients increases by 1.114,1.582,and 1.873 fold,respectively,with 95％confidence intervals of 1.109-1.283,1.431-1.624,and 1.656-5.361,respectively.However,for each unit increase in prostate volume,the likelihood of a favorable prognosis for patients decreases by 3.3％,with a 95％confidence interval of 0.943-0.985.Finally,five radiomics features(ICC＞0.750)related to the prognosis of prostate cancer were selected.In the training set,the AUG values of the nomogram model,radiomics model,and clinical model were 0.846(0.802-0.884),0.788(0.752-0.823),and 0.767(0.685-0.817),respectively.Among them,the nomogram model had the highest AUC value and the best predictive performance.Both the calibration curve and the Hosmer-Lemeshow test results showed that the nomogram model had high predictive performance.The test results showed that the goodness-of-fit between the predicted and actual values was(x2=1.827,P=0.216),indicating no statistically significant difference and a high degree of fit.The net benefit of the model was much higher than that of imaging and clinical features.Conclusion:The tPSA,PSAD,radiomics score,and prostate volume are independent factors associated with a favorable prognosis after implantation of iodine-125 particles in patients with prostate cancer.The nomogram model constructed by combining ultrasonic radiomics with clinical covariates has high discrimination and accuracy in predicting the effect of iodine-125 particle implantation in patients with prostate cancer,and has high clinical decision-making value in clinical applications.关键词
超声组学/前列腺癌/碘125粒子/预后/列线图模型Key words
ultrasonography-based radiomics/prostatic cancer/iodine-125 particle/prognosis/nomogram model分类
医药卫生
