山东大学学报 (医学版) ›› 2023, Vol. 61 ›› Issue (12): 44-50.doi: 10.6040/j.issn.1671-7554.0.2023.0770
• 医学影像人工智能的创新与挑战—临床研究 • 上一篇
朱正阳1,沈靖菲2,陈思璇1,叶梅萍1,杨惠泉1,周佳南1,梁雪1,张鑫1,张冰1
ZHU Zhengyang1, SHEN Jingfei2, CHEN Sixuan1, YE Meiping1, YANG Huiquan1, ZHOU Jianan1, LIANG Xue1, ZHANG Xin1, ZHANG Bing1
摘要: 目的 探讨磁敏感加权成像(SWI)不同影像组学模型术前预测胶质瘤异柠檬酸脱氢酶(IDH)突变状态的效能。 方法 回顾性分析加州大学旧金山分校弥漫性胶质瘤术前磁共振成像数据集(UCSF-PDGM)中经病理证实的493例成人弥漫性胶质瘤患者的影像资料,其中IDH野生型393例,IDH突变型100例,按照8∶2分为训练集(395例)和测试集(98例),根据数据集中提供的分割标签,在SWI序列上按照肿瘤核心(TC)和全肿瘤区(WT)两个感兴趣区(ROI)提取了共计1 316个影像组学特征,采用z-score法对特征进行标准化,通过主成分分析法进行降维,运用方差分析进行特征筛选,采用支持向量机(SVM)、线性判别分析(LDA)、自编码器(AE)、逻辑回归(LR)算法 、Lasso的逻辑回归算法(LR-Lasso)、贝叶斯分类器(NB)建模。通过受试者工作特征(ROC)曲线构建计算模型的准确率、敏感性、特异性,用测试集的曲线下面积(AUC)评估模型效能。 结果 20项影像组学特征被筛选用于建立影像组学诊断模型。SVM模型的AUC和准确率分别为0.841和0.755;LDA模型的AUC和准确率分别为0.800和0.735;AE模型的AUC和准确率分别为0.743和0.745;LR模型的AUC和准确率分别为0.842和0.725;LR-LASSO模型的AUC和准确率分别为0.880和0.857;NB模型的AUC和准确率分别为0.806和0.725。 结论 SWI影像组学特征在预测胶质瘤IDH基因突变中有一定的价值,基于LR-LASSO模型的预测效果最好。
中图分类号:
[1] Schaff LR, Mellingoff I K. Glioblastoma and other primary brain malignancies in adults: a review [J]. JAMA, 2023, 329(7): 574-587. [2] Louis DN, Perry A, Wesseling P, et al. The 2021 WHO classification of tumors of the central nervous system: a summary [J]. Neuro Oncol, 2021, 23(8): 1231-1251. [3] Pirozzi CJ, Yan H. The implications of IDH mutations for cancer development and therapy [J]. Nat Rev Clin Oncol, 2021, 18(10): 645-661. [4] Van-der-voort SR, Incekara F, Wijnenga MM J, et al. Combined molecular subtyping, grading, and segmentation of glioma using multi-task deep learning [J]. Neuro Oncol, 2023, 25(2): 279-289. [5] Yang X, Hu C, Xing Z, et al. Prediction of Ki-67 labeling index, ATRX mutation, and MGMT promoter methylation status in IDH-mutant astrocytoma by morphological MRI, SWI, DWI, and DSC-PWI [J]. Eur Radiol, 2023, 33(10): 7003-7014. [6] Yang X, Xing Z, She D, et al. Grading of IDH-mutant astrocytoma using diffusion, susceptibility and perfusion-weighted imaging [J]. BMC Medical Imaging, 2022, 22(1): 105. doi: 10.1186/s12880-022-00832-3. [7] Kong LW, Chen J, Zhao H, et al. Intratumoral susceptibility signals reflect biomarker status in gliomas [J]. Sci Rep, 2019, 9(1): 17080. doi: 10.1038/s41598-019-53629-w. [8] 李欣, 谢继承, 王静, 等. 磁共振MRS、DWI及SWI序列在脑胶质瘤分级诊断中的应用价值[J]. 中华全科医学, 2022, 20(9): 1541-1544. LI Xin, XIE Jichen, WANG Jing, et al. The application value of magnetic resonance MRS, DWI and SWI sequences in the grading diagnosis of glioma [J]. Chinese Journal of General Practice, 2022, 20(9): 1541-1544. [9] Rudie JD, Rauschecker AM, Bryan RN, et al. Emerging applications of artificial intelligence in neuro-oncology [J]. Radiology, 2019, 290(3): 607-618. [10] Yan J, Zhang B, Zhang S, et al. Quantitative MRI-based radiomics for noninvasively predicting molecular subtypes and survival in glioma patients [J]. NPJ Precis Oncol, 2021, 5(1): 72. doi: 10.1038/s41698-021-00205-z. [11] Calabrese E, Villanueva-meyer JE, Rudie JD, et al. The university of california san francisco preoperative diffuse glioma MRI dataset [J]. Radiol Artif Intell, 2022, 4(6): e220058. doi: 10.1148/ryai.220058. [12] Clark K, Vendt B, Smith K, et al. The cancer imaging archive(TCIA): maintaining and operating a public information repository [J]. J Digit Imaging, 2013, 26(6): 1045-1057. [13] Song Y, Zhang J, Zhang YD, et al. FeAture explorer(FAE): a tool for developing and comparing radiomics models [J]. PLoS One, 2020, 15(8): e0237587. doi: 10.1371/journal.pone.0237587. [14] 涂佳琪, 朱凤平, 李郁欣. 磁敏感加权成像在脑胶质瘤诊疗中的应用进展[J]. 国际医学放射学杂志, 2023, 46(4): 414-416. TU Jiaqi, ZHU Fengping, LI Yuxin. Advances in the application of susceptibility-weighted imaging in the diagnosis and treatment of brain gliomas [J]. International Journal of Medical Radiology, 2023, 46(4): 414-416. [15] Lin Y, Xing Z, She D, et al. IDH mutant and 1p/19q co-deleted oligodendrogliomas: tumor grade stratification using diffusion-, susceptibility-, and perfusion-weighted MRI [J]. Neuroradiology, 2017, 59(6): 555-562. [16] Saini J, Gupta PK, Sahoo P, et al. Differentiation of grade II/III and grade IV glioma by combining "T1 contrast-enhanced brain perfusion imaging" and susceptibility-weighted quantitative imaging [J]. Neuroradiology, 2018, 60(1): 43-50. [17] 杜常月, 齐旭红, 温智勇, 等. SWI、3D-ASL及IVIM鉴别高低级别脑胶质瘤的研究[J]. 中国CT和MRI杂志, 2023, 21(1): 9-11. DU Changyue, QI Xuhong, WEN Zhiyong, et al. The study of SWI, 3D-ASL and IVIM in distinguishing high and low grade glioma [J]. Chinese Journal of CT and MRI, 2023, 21(1): 9-11. [18] Yang X, Lin Y, Xing Z, et al. Predicting 1p/19q codeletion status using diffusion-, susceptibility-, perfusion-weighted, and conventional MRI in IDH-mutant lower-grade gliomas [J]. Acta Radiol, 2021, 62(12): 1657-1665. [19] Natsumeda M, Matsuzawa H, Watanabe M, et al. SWI by 7T MR Imaging for the microscopic imaging diagnosis of astrocytic and oligodendroglial tumors [J]. AJNR Am J Neuroradiol, 2022, 43(11): 1575-1581. [20] Kihira S, Mei X, Mahmoudi K, et al. U-net based segmentation and characterization of gliomas [J]. Cancers(Basel), 2022, 14(18): 4457. doi: 10.3390/cancers14184457. [21] 宋珍珍, 孙小玲, 李海鸥, 等. 120例胶质瘤及瘤周水肿MRI影像组学在评估肿瘤复发中的价值[J]. 山东大学学报(医学版), 2021, 59(11): 53-60. SONG Zhenzhen, SUN Xiaoling, LI Haiou, et al. Value of MRI radiomics of glioma and peritumoral edema in evaluating tumor recurrence [J]. Journal of Shandong University(Health Sciences), 2021, 59(11): 53-60. [22] 袁宏涛, 纪淙山, 康冰, 等. CT影像组学对肾上腺乏脂腺瘤与结节样增生的诊断价值[J]. 山东大学学报(医学版), 2022, 60(4): 68-75. YUAN Hongtao, JI Congshan, KANG Bing, et al. Diagnostic value of CT radiomics nomogram for adrenal lipid-poor adenoma and nodular hyperplasia [J]. Journal of Shandong University(Health Sciences), 2022, 60(4): 68-75. [23] Li Y, Ammari S, Lawrance L, et al. Radiomics-based method for predicting the glioma subtype as defined by tumor grade, IDH mutation, and 1p/19q codeletion [J]. Cancers(Basel), 2022, 14(7):1778. doi: 10.3390/cancers14071778. [24] Li Y, Wei D, Liu X, et al. Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learning [J]. Eur Radiol, 2022, 32(2): 747-758. [25] Xiong D, Ren X, Huang W, et al. Noninvasive classification of glioma subtypes using multiparametric MRI to improve deep learning [J]. Diagnostics(Basel), 2022, 12(12):3063. doi: 10.3390/diagnostics12123063. |
[1] | 王琳琳 孙美丽 孙玉萍 张楠 刘传勇. 中心体α-微管蛋白、γ-微管蛋白在脑胶质瘤中的表达及其与Survivin表达的相关性研究[J]. 山东大学学报(医学版), 2209, 47(6): 103-. |
[2] | 李夕凤,李红梅. 脑脊液CXCL10:抗NMDAR脑炎潜在的生物学标志物[J]. 山东大学学报 (医学版), 2023, 61(6): 47-52. |
[3] | 靳新娟,左立平,邓展昊,李安宁,于德新. MRI影像组学对135例肝癌耐药蛋白PFKFB3的预测价值[J]. 山东大学学报 (医学版), 2023, 61(6): 79-86. |
[4] | 刘艳,冷珊珊,夏晓娜,董昊,黄陈翠,孟祥水. 基于影像组学参数评估376例幕上自发性脑出血患者的功能状态[J]. 山东大学学报 (医学版), 2023, 61(5): 59-67. |
[5] | 刘亚军,郎昭,郭安忆,刘文勇. 骨科冲击波治疗的智能化发展现状及趋势分析[J]. 山东大学学报 (医学版), 2023, 61(3): 7-13. |
[6] | 吴南,仉建国,朱源棚,陈癸霖,陈泽夫. 人工智能在脊柱畸形诊疗中的应用[J]. 山东大学学报 (医学版), 2023, 61(3): 14-20. |
[7] | 巨艳丽,王丽华,成芳,黄凤艳,陈学禹,贾红英. 基于机器学习构建放射性碘治疗疗效的预测模型[J]. 山东大学学报 (医学版), 2023, 61(1): 94-99. |
[8] | 况利,徐小明,曾琪. 机器学习用于自杀研究的综述[J]. 山东大学学报 (医学版), 2022, 60(4): 10-16. |
[9] | 袁宏涛,纪淙山,康冰,秦松楠,于鑫鑫,高琳,王锡明. CT影像组学对肾上腺乏脂腺瘤与结节样增生的诊断价值[J]. 山东大学学报 (医学版), 2022, 60(4): 68-75. |
[10] | 李华玉,时萧寒,张新蕊,李峰. 203例胶质瘤患者睡眠障碍与炎症细胞因子的关联分析[J]. 山东大学学报 (医学版), 2022, 60(12): 26-30. |
[11] | 姜震,孙静,邹雯,王唱唱,高琦. 基于两种机器学习算法的双相情感障碍患者自杀行为影响因素模型比较研究[J]. 山东大学学报 (医学版), 2022, 60(1): 101-108. |
[12] | 孙庆杰,张怡莎,管尚慧,凤志慧. 丙戊酸对134例放疗神经胶质瘤患者预后生存和肿瘤复发的影响[J]. 山东大学学报 (医学版), 2021, 59(8): 80-85. |
[13] | 田瑶天,王宝,李叶琴,王滕,田力文,韩波,王翠艳. 基于可解释性心脏磁共振参数的机器学习模型预测儿童心肌炎的预后[J]. 山东大学学报 (医学版), 2021, 59(7): 43-49. |
[14] | 顾金海,路宁,顾珈榕,文玉军,强媛媛,和祯泉,杨勇,王峰,孙涛,牛建国. 胶质瘤细胞与血管内皮细胞的信号Crosstalk对肿瘤细胞增殖和侵袭的影响[J]. 山东大学学报 (医学版), 2021, 59(2): 1-6. |
[15] | 宋珍珍,孙小玲,李海鸥,王芳,张冉,于德新. 120例胶质瘤及瘤周水肿MRI影像组学在评估肿瘤复发中的价值[J]. 山东大学学报 (医学版), 2021, 59(11): 53-60. |
|