| 陈峻帆,杨静,黄豆豆,罗银灯.体素内不相干运动模型鉴别诊断移行带前列腺癌与基质型前列腺增生[J].中国医学影像技术,2022,38(10):1534~1539 |
| 体素内不相干运动模型鉴别诊断移行带前列腺癌与基质型前列腺增生 |
| Models of intravoxel incoherent motion imaging for differential diagnosis of transitional zone prostate cancer and stromal prostatic hyperplasia |
| 投稿时间：2022-04-08 修订日期：2022-06-09 |
| DOI：10.13929/j.issn.1003-3289.2022.10.021 |
| 中文关键词: 前列腺肿瘤 前列腺增生 磁共振成像 |
| 英文关键词:prostatic neoplasms prostatic hyperplasia magnetic resonance imaging |
| 基金项目:重庆市科技局技术创新与应用发展专项面上项目（cstc2019jscx-msxmX0253）、重庆医科大学附属第二医院江苏恒瑞医药股份有限公司"横向课题"项目（A201905）。 |
|
| 摘要点击次数: 1994 |
| 全文下载次数: 479 |
| 中文摘要: |
| 目的 评价体素内不相干运动（IVIM）单指数模型、双指数模型、灌注扩散比（PDR）及拉伸指数模型各参数鉴别诊断移行带前列腺癌（PCa）与基质型良性前列腺增生（S-BPH）的价值。方法 纳入移行带PCa患者（PCa组）及S-BPH患者（S-BPH组）各24例，以不同图像后处理技术分析IVIM图像，获得标准表观弥散系数（ADCStand）、快速ADC （ADCFast）、慢速ADC （ADCSlow）、灌注分数（f）、弥散分布指数（DDC）及弥散异质性指数（α），计算PDR。比较组间基本资料及MRI参数差异，以单因素logistic回归分析筛选变量，纳入多因素回归分析，构建鉴别诊断移行带PCa与S-BPH的回归模型；绘制受试者工作特征曲线，计算曲线下面积（AUC），评价MRI参数及回归模型鉴别诊断移行带PCa与S-BPH的价值，并以DeLong检验比较各参数AUC的差异。结果 组间患者年龄、血清前列腺特异性抗原（PSA）水平、前列腺影像报告和数据系统（PI-RADS）评分及病灶面积差异均有统计学意义（P均<0.05）。PCa组ADCStand、ADCSlow、DDC及α均低于S-BPH组（P均<0.05），而ADCFast、PDR高于S-BPH组（P均<0.05）。ADCStand、ADCFast、ADCSlow、DDC及PDR用于鉴别诊断移行带PCa与S-BPH的效能均较高，AUC分别为0.948、0.966、0.900、0.931及0.965，均优于α效能（P均<0.05）。年龄、PSA、PI-RADS、ADCStand、ADCFast、ADCSlow、DDC及α纳入多因素logistic回归分析，所构建回归模型的诊断AUC、敏感度及特异度分别为0.988、95.80%及100%。结论 多数IVIM双指数模型、拉伸指数模型参数鉴别诊断移行带PCa与S-BPH的效能均较高；PDR可作为IVIM模型的重要补充；联合临床与影像学指标构建的预测模型鉴别诊断效能优异。 |
| 英文摘要: |
| Objective To observe the value of intravoxel incoherent motion (IVIM) monoexponential model, biexponential model, perfusion diffusion ratio (PDR) and parameters of stretch index model for differential diagnosis of transitional zone prostate cancer (PCa) and stromal benign prostatic hyperplasia (S-BPH). Methods Totally 24 patients with transitional zone PCa (PCa group) and other 24 patients with S-BPH (S-BPH group) were enrolled. IVIM images were analyzed using different image post-processing techniques to obtain standard apparent diffusion coefficients (ADCStand), fast ADC (ADCFast), slow ADC (ADCSlow), perfusion fraction (f), distributed diffusion coefficient (DDC) and diffusion heterogeneity index (α), and to calculate PDR. The basic information and MRI parameters were compared between groups, and variables were screened with univariate logistic regression analysis for multifactor regression analysis to construct a regression model for differential diagnosis of transitional zone PCa and S-BPH. Receiver operating characteristic curves were plotted, and the area under the curves (AUC) were calculated to evaluate the value of MRI parameters and regression models for differential diagnosis of transitional zone PCa versus S-BPH, and the AUC of MRI parameters were compared using DeLong test. Results Significant differences of patient age, serum prostate specific antigen (PSA) level, prostate imaging reporting and data system (PI-RADS) score and area of lesion were found between groups (all P<0.05). ADCStand, ADCSlow, DDC and α in PCa group were lower, while ADCFast and PDR were higher than those in S-BPH group (all P<0.05). ADCStand, ADCFast, ADCSlow, DDC and PDR had high efficacy for differential diagnosis of transitional zone PCa and S-BPH, with AUC of 0.948, 0.966, 0.900, 0.931 and 0.965, respectively, better than that of α (all P<0.05). Patients' age, PSA, PI-RADS, ADCStand, ADCFast, ADCSlow, DDC and α were included in the multifactorial logistic regression analysis, and the diagnostic AUC, sensitivity and specificity of the obtained regression model was 0.988, 95.80% and 100%, respectively. Conclusion Most IVIM biexponential models and stretching exponential model parameters were effective for differential diagnosis of transitional zone PCa and S-BPH. PDR could be an important complement to IVIM model.The predictive model constructed through combination of clinical and imaging indicators had excellent differential diagnostic efficacy. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
