乳腺癌MRI影像组学特征与分子标记物的相关性研究
|
摘要
目的:探讨乳腺癌MRI动态增强扫描(DCE-MRI)影像组学特征与病理免疫组化分子标记物的相关性。方法:回顾性分析140例经手术病理证实、行DCE-MRI扫描的乳腺癌患者的病例资料,对所有患者提取乳腺癌病灶的MRI影像组学特征(包括11个形态学特征和13个纹理特征),并采用病理学免疫组化方法检测分子标记物,比较分子标记物表达阳性者与表达阴性者间的形态学特征及纹理特征的差异,对具有组间差异的影像组学特征进行受试者工作特征(ROC)曲线分析。结果:雌激素受体(ER)阳性者的和平均(290.28±28.90)明显高于ER阴性者(266.26±33.76),鉴别病灶是否表达ER时,ROC曲线下面积(AUC)为0.701;孕激素受体(PR)阳性者的圆度(3.99±2.75)明显低于PR阴性者(6.11±4.18),联合圆度、紧致度、实体度这3个影像组学特征鉴别病灶是否表达PR时,AUC为0.678;增殖细胞核抗原(Ki-67)阳性者的和熵(7.76±0.53)明显高于Ki-67阴性者(7.36±0.50),联合和熵、毛刺度、面积、熵这4个影像组学特征鉴别病灶是否表达Ki-67时,AUC为0.767;P53蛋白(P53)阳性者的紧致度(2.56±1.33)明显低于P53阴性者(4.03±2.79),联合紧致度、分形度、实体度、圆度这4个影像组学特征鉴别病灶是否表达P53时,AUC为0.669。结论:乳腺癌MRI影像组学特征与分子标记物的表达具有相关性,可作为临床术前无创性预测乳腺癌分子标记物表达的手段。
Objective:To investigate the correlation between radiomics features of dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI) and pathological immunohistochemical molecular markers in breast cancer.Methods:A total of 140 patients with pathologically confirmed breast cancer were retrospective enrolled in the study.24 radiomics features were extracted from all lesions,including 11 morphologic and 13 texture features.Pathological immunohistochemistry was used to define the molecular markers.The differences in morphological and textural characteristics between the groups with positive and negative expression of molecular markers were compared.The imaging traits with differences between groups were further analyzed by the receiver operating characteristic(ROC) curve.Results:Sum average of ER-positive group(290.28±28.90) were significantly higher than that in ER-negative group(266.26±33.76);the area under the ROC curve(AUC) was 0.701.Circularity of PR-positive group(3.99±2.75) were significantly lower than that in PR-negative group(6.11±4.18).The AUC was 0.678 when combining three imaging features of circularity,compactness and solidity for the classification of PR expression.Sum entropy of Ki-67-positive group(7.76±0.53) were significantly higher than Ki-67-negative group(7.36±0.50).The AUC was 0.767 when combining four imaging features of sum entropy,spiculation,area and entropy for the classification of Ki-67 expression.Compactness of P53-positive group(2.56±1.33) were significantly lower than P53-negative group(4.03±2.79).The AUC was 0.669 when combining four imaging features of compactness,fractal,solidity and circularity for the classification of P53 expression.Conclusion:Radiomics features of breast cancer extracted from DCE-MRI are partly correlated with the expression of molecular markers,which may be useful to non-invasively predict the expression of molecular markers in breast cancer before surgery.
Objective:To investigate the correlation between radiomics features of dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI) and pathological immunohistochemical molecular markers in breast cancer.Methods:A total of 140 patients with pathologically confirmed breast cancer were retrospective enrolled in the study.24 radiomics features were extracted from all lesions,including 11 morphologic and 13 texture features.Pathological immunohistochemistry was used to define the molecular markers.The differences in morphological and textural characteristics between the groups with positive and negative expression of molecular markers were compared.The imaging traits with differences between groups were further analyzed by the receiver operating characteristic(ROC) curve.Results:Sum average of ER-positive group(290.28±28.90) were significantly higher than that in ER-negative group(266.26±33.76);the area under the ROC curve(AUC) was 0.701.Circularity of PR-positive group(3.99±2.75) were significantly lower than that in PR-negative group(6.11±4.18).The AUC was 0.678 when combining three imaging features of circularity,compactness and solidity for the classification of PR expression.Sum entropy of Ki-67-positive group(7.76±0.53) were significantly higher than Ki-67-negative group(7.36±0.50).The AUC was 0.767 when combining four imaging features of sum entropy,spiculation,area and entropy for the classification of Ki-67 expression.Compactness of P53-positive group(2.56±1.33) were significantly lower than P53-negative group(4.03±2.79).The AUC was 0.669 when combining four imaging features of compactness,fractal,solidity and circularity for the classification of P53 expression.Conclusion:Radiomics features of breast cancer extracted from DCE-MRI are partly correlated with the expression of molecular markers,which may be useful to non-invasively predict the expression of molecular markers in breast cancer before surgery.
引文
[1] Siegel RL,Miller KD,Jemal A.Cancer statistics,2018[J].CA Cancer J Clin,2018,68(1):7-30.
[2] 林帆,胡若凡,梁超,等.磁共振纹理动态特征鉴别乳腺良恶性肿块[J].放射学实践,2017,32(10):1037-1040.
[3] Crivelli P,Ledda RE,Parascandolo N,et al.A new challenge for radiologists:radiomics in breast cancer[J].Biomed Res Int,2018:1-10.
[4] Wu KL.Analysis of parameter selections for fuzzy c-means[J].Patt Recog,2012,45(1):407-415.
[5] Xu C,Prince JL.Snakes,shapes,and gradient vector flow[J].IEEE Trans Image Process,1998,7(3):359-369.
[6] Voduc KD,Cheang MC,Tyldesley S,et al.Breast cancer subtypes and the risk of local and regional relapse[J].J Clin Oncol,2010,28(10):1481-1491.
[7] Grimm LJ,Zhang J,Mazurowski MA.Computational approach to radiogenomics of breast cancer:Luminal A and luminal B molecular subtypes are associated with imaging features on routine breast MRI extracted using computer vision algorithms[J].J Magn Reson Imaging,2015,42(4):902-907.
[8] Agner SC,Rosen MA,Englander S,et al.Computerized image analysis for identifying triple-negative breast cancers and differentiating them from other molecular subtypes of breast cancer on dynamic contrast-enhanced MR images:a feasibility study[J].Radiology,2014,272(1):91-99.
[9] Fan M,Li H,Wang S,et al.Radiomic analysis reveals DCE-MRI features for prediction of molecular subtypes of breast cancer[J].PLoS One,2017,12(2):1-15.
[10] Mazurowski MA,Zhang J,Grimm LJ,et al.Radiogenomic analysis of breast cancer:Luminal B molecular subtype is associated with enhancement dynamics at MR imaging[J].Radiology,2014,273(2):365-372.
[11] Wang J,Kato F,Oyama-Manabe N,et al.Identifying triple-Negative breast cancer using background parenchymal enhancement heterogeneity on dynamic contrast-enhanced MRI:a pilot radiomics study[J].PLoS One,2015,10(11):1-10.
[12] Li H,Zhu Y,Burnside ES,et al.Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set[J].NPJ Breast Cancer,2016,2.pii:16012.Epub 2016 May 11.
[13] Ma W,Ji Y,Qi L,et al.Breast cancer Ki-67 expression prediction by DCE-MRI radiomics features[J].Clin Radiol,2018,73(10):1-5.
[14] 李建灵,殷洁,廖珍,等.乳腺癌MRI表现与生物因子ER、PR、c-erbB-2、p53的相关性研究[J].实用放射学杂志,2015,31(7):1095-1099.
[2] 林帆,胡若凡,梁超,等.磁共振纹理动态特征鉴别乳腺良恶性肿块[J].放射学实践,2017,32(10):1037-1040.
[3] Crivelli P,Ledda RE,Parascandolo N,et al.A new challenge for radiologists:radiomics in breast cancer[J].Biomed Res Int,2018:1-10.
[4] Wu KL.Analysis of parameter selections for fuzzy c-means[J].Patt Recog,2012,45(1):407-415.
[5] Xu C,Prince JL.Snakes,shapes,and gradient vector flow[J].IEEE Trans Image Process,1998,7(3):359-369.
[6] Voduc KD,Cheang MC,Tyldesley S,et al.Breast cancer subtypes and the risk of local and regional relapse[J].J Clin Oncol,2010,28(10):1481-1491.
[7] Grimm LJ,Zhang J,Mazurowski MA.Computational approach to radiogenomics of breast cancer:Luminal A and luminal B molecular subtypes are associated with imaging features on routine breast MRI extracted using computer vision algorithms[J].J Magn Reson Imaging,2015,42(4):902-907.
[8] Agner SC,Rosen MA,Englander S,et al.Computerized image analysis for identifying triple-negative breast cancers and differentiating them from other molecular subtypes of breast cancer on dynamic contrast-enhanced MR images:a feasibility study[J].Radiology,2014,272(1):91-99.
[9] Fan M,Li H,Wang S,et al.Radiomic analysis reveals DCE-MRI features for prediction of molecular subtypes of breast cancer[J].PLoS One,2017,12(2):1-15.
[10] Mazurowski MA,Zhang J,Grimm LJ,et al.Radiogenomic analysis of breast cancer:Luminal B molecular subtype is associated with enhancement dynamics at MR imaging[J].Radiology,2014,273(2):365-372.
[11] Wang J,Kato F,Oyama-Manabe N,et al.Identifying triple-Negative breast cancer using background parenchymal enhancement heterogeneity on dynamic contrast-enhanced MRI:a pilot radiomics study[J].PLoS One,2015,10(11):1-10.
[12] Li H,Zhu Y,Burnside ES,et al.Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set[J].NPJ Breast Cancer,2016,2.pii:16012.Epub 2016 May 11.
[13] Ma W,Ji Y,Qi L,et al.Breast cancer Ki-67 expression prediction by DCE-MRI radiomics features[J].Clin Radiol,2018,73(10):1-5.
[14] 李建灵,殷洁,廖珍,等.乳腺癌MRI表现与生物因子ER、PR、c-erbB-2、p53的相关性研究[J].实用放射学杂志,2015,31(7):1095-1099.