DCE-MRI影像组学联合血细胞参数鉴别管腔型和非管腔型乳腺癌的价值  

Value of radiomics model based on DCE-MRI combined with blood cell parameters in differentiating Luminal and Non-Luminal breast cancer

作　　者：胥豪 杨澳 胡云涛 周鹏[1] 邓和平[1] XU Hao;YANG Ao;HU Yuntao;ZHOU Peng;DENG Heping(Department of Radiology,Sichuan Clinical Research Center for Cancer,Sichuan Cancer Hospital&Institute,Sichuan Cancer Center,University of Electronic Science and Technology of China,Chengdu 610041,China;School of Medicine,University of Electronic Science and Technology of China,Sichuan Cancer Hospital&Institute,Chengdu 610041,China)

机构地区：[1]四川省肿瘤医院·研究所,四川省肿瘤临床医学研究中心,四川省癌症防治中心,电子科技大学附属肿瘤医院影像科,成都610041 [2]电子科技大学医学院,四川省肿瘤医院·研究所,成都610041

出　　处：《磁共振成像》2025年第4期33-40,150,共9页Chinese Journal of Magnetic Resonance Imaging

基　　金：国家自然科学基金项目(编号:U21A20521)。

摘　　要：目的探讨动态对比增强磁共振成像(dynamic contrast-enhanced magnetic resonance imaging,DCE-MRI)影像组学联合血细胞参数在鉴别管腔上皮型(Luminal)和非管腔上皮型(Non-Luminal)乳腺癌中的价值。材料与方法回顾性分析227例经病理证实的乳腺癌患者的DCE-MRI图像,按照7∶3的比例随机分为训练集(n=162)和验证集(n=65)。根据免疫组化结果将病灶分为Luminal型和Non-Luminal型。勾画DCE-MRI图像中肿瘤的瘤体作为感兴趣区进行影像组学特征提取。分析影像组学特征、临床病理特征及血细胞参数,构建了3个模型,用于鉴别Luminal和Non-Luminal,分别为模型1(影像组学)、模型2(血细胞参数)、模型3(影像组学+血细胞参数)。利用受试者工作特征曲线评价模型在二者之间鉴别诊断中的效能。采用决策分析曲线评价不同风险阈值下模型的净获益情况。结果模型3训练集的曲线下面积(area under the curve,AUC)值、敏感度、特异度分别为0.840(0.774~0.893)、87.9%、71.4%,验证集对应的值分别为0.818(0.703~0.903)、87.5%、68.0%。在训练集及验证集中,模型1的AUC均高于模型2(0.817 vs.0.636,0.838 vs.0.515,P=0.001和P<0.001),同时模型3的AUC均高于模型2(0.840 vs.0.636,0.818 vs.0.515,P值均<0.001)。决策曲线示三个模型在鉴别Luminal和Non-Luminal亚型均有明确临床获益,并且模型3、模型1优于模型2。结论联合DCE-MRI影像组学和血细胞参数的模型能帮助鉴别Luminal和Non-Luminal亚型,有助于准确制订乳腺癌的治疗方案。Objective:To explore the value of dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI) radiomics model combined with blood cell parameters in differentiating Luminal and Non-Luminal breast cancer(BC).Materials and Methods:DCE-MRI of two hundred and twenty-seven patients with pathologically confirmed BC were retrospectively analyzed.The patients were randomly split into a training set(n = 162) and a validation set(n = 65) at a ratio of 7∶3.Patients were divided into the Luminal group(139 cases) and the Non-Luminal group(88 cases) according to the immunohistochemical results.The BC lesions on the pretreatment DCE-MRI served as the basis for the volume of interest for feature extraction.Three models were constructed to differentiate Luminal from Non-Luminal by analyzing radiomic feature,clinical pathological feature,and hematological parameters.These models were Model 1(radiomics),Model 2(hematological parameters),and Model 3(radiomics + hematological parameters),respectively.The discrimination performance of the models was evaluated using the receiver operating characteristic curve.Decision curve analysis was conducted to determine the clinical usefulness of the models by quantifying the net benefits at different threshold probabilities.Results:The area under the curve(AUC),sensitivity,and specificity,of Model 3 were 0.840(0.774 to 0.893),87.9%,and 71.4% in the training set,and 0.818(0.703 to 0.903),87.5%,and 68.0% in the validation set,respectively.The AUC of the Model 1 was better than that of the Model 2 in the both cohorts(0.817 vs.0.636,0.838 vs.0.515,P = 0.001 and P < 0.001),and the AUC of the Model 3 was also better than that of the Model 2 in the both cohorts(0.840 vs.0.636,0.818 vs.0.515,both P < 0.001).The Model 3 and Model 1 were both more beneficial than Model 2 in clinical practice,as illustrated by decision curve analysis.Conclusions:The model that integrated the hematological parameters with DCE-MRI radiomics can help differentiate Luminal from Non-Luminal BC,which facilitates the accurate

关 键 词：管腔型乳腺癌 非管腔型乳腺癌 磁共振成像 影像组学 血细胞参数 诊断价值 

分 类 号：R445.2[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]