中国中西医结合影像学杂志2025，Vol.23Issue(5)：647-651,665,6.DOI:10.3969/j.issn.1672-0512.2025.05.022
新型关联规则模型对PSA灰区前列腺癌的预测价值
Predictive value of a new association rule model for prostate cancer in PSA grey zone
摘要
Abstract
Objective:By creating a new association rule model based on clinical data such as Prostate Imaging Reporting and Data System version 2.1(PI-RADS V2.1)score,to evaluate the efficiency of the model for predicting prostate cancer(PCa)in the prostate-specific antigen(PSA)grey zone.Methods:A total of 1 000 patients with prostate disease were retrospectively collected.The patients were randomly divided into a modeling group(700 cases)and a validation group(300 cases).The association rule analysis was performed on the clinical data of the modeling group,taking the clinical data such as PI-RADS V2.1 score as the antecedent and PCa as the consequent,the effective strong association rules were calculated and the predictive model was created.The calibration performance was assessed using the calibration curve.Internal validation was also performed.ROC curve was used to analyze the AUC of the model for predicting PCa.External validation was also performed.Results:The prediction results of the new association rule model showed that the incidence of PCa was 79%when PI-RADS V2.1 score≥3 points,age≥70 years,prostate volume<50 mL,and total PSA(tPSA)≥7 ng/mL.The internal validation showed that the concordance index(C-index)of the model was 0.762(95%CI 0.639-0.810),and there was a good consistency between the predicted and observed values on the calibration curve.The AUC of the model for predicting PCa was 0.893,which was higher than the PI-RADS V2.1 score(0.831),age(0.686),prostate volume(0.722),and tPSA(0.634).The external validation showed that the AUC of the model for predicting PCa in the validation group was 0.884.The result was in good agreement with the ROC curve of the modeling group,and the difference was not statistically significant(χ2=8.098,P>0.05).The prediction performance of the model for PCa in different prostate regions could meet the clinical requirements.Conclusions:The new association rule model based on clinical data such as PI-RADS V2.1 score shows favorable performance in predicting PCa and has potential clinical application value.关键词
前列腺影像报告和数据系统2.1版/关联规则分析/前列腺特异性抗原/灰区/前列腺癌/预测Key words
Prostate Imaging Reporting and Data System version 2.1/Association rule analysis/Prostate specific antigen/Grey area/Prostate cancer/Forecast引用本文复制引用
王晸,田龙,崔书君,杨松林,刘庆啸,朱晓龙..新型关联规则模型对PSA灰区前列腺癌的预测价值[J].中国中西医结合影像学杂志,2025,23(5):647-651,665,6.基金项目
张家口市重点研发计划项目(2421018D). （2421018D）
