基于多参数MRI影像组学结合PI-RADS v2.1和临床指标的新型列线图预测临床显著性前列腺癌

A Novel Nomogram for Predicting Clinical Significate Prostate Cancer:Radiomics Based on Multi⁃Parameter MRI Combined with PI⁃RADS v2.1 and Clinical Parameters

目的建立基于多参数MRI影像组学结合PI-RADS v2.1和临床指标的新型列线图,评价其预测临床显著性前列腺癌(csPCa)的价值。方法回顾性分析204例患者的资料,进行PI-RADS v2.1评分和影像组学分析。应用受试者工作特征曲线和临床决策曲线评估临床模型、PI-RADS模型、影像组学模型及各联合模型诊断csPCa的效能和临床获益,基于效能最优模型建立列线图并验证。结果影像组学模型诊断效能显著优于临床模型和PI-RADS评分模型,差异具有统计学意义(P<0.05)。在临床模型或PI-RADS模型中增加影像组学的特征,其联合诊断效能会显著提高(P<0.05)。结论基于多参数MRI影像组学结合PI-RADS v2.1和临床指标的联合模型所建立的列线图为术前预测csPCa提供了一种无创性的新方法。

Objective To develop a novel nomogram based on multi⁃parameter magnetic resonance imaging radiomics combined with PI⁃RADS v2.1 and clinical indicators,and evaluate its application value in predicting clinically significant prostate cancer(csPCa).Methods The data of 204 patients in our hospital from January 2018 to December 2019 were retrospectively analyzed.The PI⁃RADS v2.1 score was used to evaluate images and PI⁃RADS v2.1 score and radiomics a⁃nalysis were performed.Applying receiver operating characteristic(ROC)curve and clinical decision curve to evaluate the diagnostic efficacy and clinical benefit of clinical model,PI⁃RADS model,radiomics model,and each combined model to de⁃tect csPCa,and a nomogram was developed and validated based on the optimal model.Results The diagnostic perform⁃ance of radiomics model was significantly better than the clinical model and PI⁃RADS model,and the difference is statisti⁃cally significant(P<0.05).By adding the radiomics features to the clinical model or PI⁃RADS model,the combined model significantly improved the diagnostic efficiency of csPCa(P<0.05).Conclusion The nomogram based on the combined model of multiparametric MRI radiomics combined with PI⁃RADS v2.1 and clinical indicators provides a noninvasive new method for preoperative prediction of csPCa.