【摘要】 目的 探析对前列腺外周带癌患者应用MRI多模态分级影像检测报告的临床诊断效果。方法 选取2018年5月至2019年10月本院前列腺疾病患者86例,均经病理检查证实,分析MRI多模态分级影像对前列腺外周带癌诊断准确率,比较阳性组与阴性组MRI分级诊断评分,并利用受试者工作特征(ROC)曲线分析MRI分级诊断评分对前列腺外周带癌诊断效能。结果 多模态MRI分级对前列腺外周带癌诊断敏感度90.63%(29/32),特异度88.89%(48/54),准确率89.53%(77/86),阳性预测值82.86%(29/35),阴性预测值94.12%(48/51),Kappa值0.78;阳性组MRI分级诊断评分显著高于阴性组(P<0.05);多模态MRI分级诊断曲线下面积(AUC)0.902,MRI评分临界值为4分,对前列腺外周带癌诊断敏感度91.6%,特异度80.8%。结论 MRI多模态分级影像可为前列腺外周带癌的诊断提供重要资料,其诊断与病理结果 具有高度一致性,诊断效能高,具有重要应用价值。更多还原

【Abstract】 Objective To explore the clinical diagnostic effect of MRI multimodal grading imaging detection report in patients with peripheral prostate cancer.Methods A total of 86 patients with prostate disease in our hospital from May 2018 to October 2019 were selected,all of which were confirmed by pathological examination.The diagnostic accura cy of MRI multimodal gra ding images for peripheral prostate ca ncer was analyzed,and the MRI grading of the positive group and the negative group was compared.Diagnostic score,and the receiver operating characteristic(ROC) curve was used to analyze the diagnostic performance of MRI grading diagnostic score for peripheral prostate cancer.Results The sensitivity of multimodal MRI grading for the diagnosis of peripheral prostate ca ncer was 90.63%(29/32),the specificity was 88.89%(48/54),the accuracy was 89.53%(77/86),and the positive predictive value was 82.86%(29/35),the negative predictive value was 94.12%(48/51),and the Kappa value was0.78;the MRI grading diagnostic score in the positive group was significantly higher than that in the negative group(P<0.05);the area under the multimodal MRI grading diagnostic cu rve(AUC) was 0.902,The critical value of MRI score was 4 points,with a sensitivity of 91.6% and a specificity of 80.8% for the diagnosis of peripheral prostate cancer.Conclusion MRI multimodal grading images can provide importa nt information for the diagnosis of peripheral prostate cancer,and its diagnosis and pathological results are highly consistent,with high diagnostic efficiency and importa nt application value.更多还原