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

doi:10.6040/j.issn.1671-7554.0.2021.0031

摘要/Abstract

摘要： 目的探讨基于可解释性心脏磁共振(CMR)参数的机器学习模型对儿童心肌炎患者预后的预测价值。方法回顾性收集2012年9月至2017年11月临床诊断为儿童心肌炎患者45例,其中男28例,女17例,4~16岁,平均(9.8±3.4)岁。根据随访过程中是否出现心血管不良事件(ACE),将患者分为预后不良组(n=18例)和预后良好组(n=27例)。所有患者于住院治疗后进行CMR扫描,获取心功能、心肌应变、首过灌注及延迟强化(LGE)相关方面共206个可解释性的CMR参数。利用MATLAB分类学习应用程序对参数进行训练,挑选精度最高的模型作为预测模型。采用受试者工作特征曲线(ROC)对模型的预测效能进行评估。结果提取出14个可解释性的CMR参数,挑选其中无显著相关性的参数构建组合参数。单一参数中, 美国心脏协会(AHA)分段法中的第7节段最大信号强度百分比(SI %)预测性能最佳,曲线下面积(AUC)、预测敏感性和特异性分别为0.790、0.667和0.833;组合参数达到了最高的预测性能,AUC、预测敏感性和特异性分别为0.940、0.750和0.889。结论根据可解释性的CMR参数建立的机器学习模型对儿童心肌炎患者预后的预测具有良好价值,且在预后评估中组合参数比单一参数预测性能更高。

关键词: 心脏磁共振, 心肌应变, 心肌炎, 机器学习, 预后

Abstract: Objective To develop and validate machine learning models based on interpretive cardiac magnetic resonance(CMR)parameters for prognosis evaluation of pediatric myocarditis. Methods A retrospective analysis of 45 pediatric patients with myocarditis was conducted. According to whether adverse cardiac events(ACE)occurred, the patients were divided into poor prognosis group(n=18)and good prognosis group(n=27). CMR scans were performed after hospitalization and 206 interpretive CMR parameters about myocardial function, myocardial strain, first-pass perfusion and late gadolinium enhancement(LGE)were obtained. The parameters were trained by the classification learner App in MATLAB and the training model with the highest accuracy was chosen as the best model. The receiver-operating characteristics(ROC)curve of the machine learning model was drawn to determine the prognostic performance. Results A total of 14 CMR parameters were selected as predictive factors, and those without correlation were used to construct the combined parameters. Among all these parameters, maximal signal intensity percentage(SI %)of the 7th segment of AHA had the best performance(AUC: 0.790, sensitivity: 0.667, specificity: 0.833). Combined parameters achieved the highest performance(AUC: 0.940, sensitivity: 0.750, specificity: 0.889). Conclusion The machine learning models based on interpretive CMR parameters can be used for prognosis evaluation of pediatric myocarditis, and combination of interpretive CMR parameters training with machine learning is more accurate than single ones.

Key words: Cardiac magnetic resonance, Myocardial strain, Myocarditis, Machine learning, Prognosis

中图分类号:

R574

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.

TIAN Yaotian, WANG Bao, LI Yeqin, WANG Teng, TIAN Liwen, HAN Bo, WANG Cuiyan. Machine learning models based on interpretive CMR parameters can predict the prognosis of pediatric myocarditis[J]. Journal of Shandong University (Health Sciences), 2021, 59(7): 43-49.

参考文献

[1] 中华医学会心血管病学分会精准医学学组, 中华心血管病杂志编辑委员会, 成人暴发性心肌炎工作组. 成人暴发性心肌炎诊断与治疗中国专家共识[J]. 中华心血管病杂志, 2017, 45(9): 742-752.
[2] 中华医学会儿科学分会心血管学组, 中华医学会儿科学分会心血管学组心肌炎协作组, 中华儿科杂志编辑委员会. 儿童心肌炎诊断建议(2018年版)[J]. 中华儿科杂志, 2019, 57(2): 87-89.
[3] Sagar S, Liu PP, Cooper LT. Myocarditis [J]. Lancet, 2012, 379(9817): 738-747.
[4] Ferreira VM, Schulz-Menger J, Holmvang G, et al. Cardiovascular magnetic resonance in nonischemic myocardial inflammation: expert recommendations [J]. J Am Coll Cardiol, 2018, 72(24): 3158-3176.
[5] Wang HP, Zhao B, Jia HP, et al. A retrospective study: cardiac MRI of fulminant myocarditis in children-can we evaluate the short-term outcomes? [J]. PeerJ, 2016, 4: e2750. doi:10.7717/peerj.2750.
[6] Di Filippo S. Improving outcomes of acute myocarditis in children[J]. Expert Rev Cardiovasc Ther, 2016, 14(1): 117-125.
[7] 曾国飞, 梁仁容. 急性心肌炎的CMR应用进展[J]. 国际医学放射学杂志, 2020, 43(1): 54-58. ZENG Guofei, LIANG Renrong. The application progress of CMR in acute myocarditis [J]. Int J Med Radiol, 2020, 43(1): 54-58.
[8] Leiner T, Rueckert D, Suinesiaputra A, et al. Machine learning in cardiovascular magnetic resonance: basic concepts and applications [J]. J Cardiovasc Magn Reson, 2019, 21(1): 61.
[9] Zhang N, Yang G, Gao Z, et al. Deep learning for diagnosis of chronic myocardial infarction on nonenhanced cardiac cine MRI [J]. Radiology, 2019, 291(3): 606-617.
[10] Baessler B, Mannil M, Oebel S, et al. Subacute and chronic left ventricular myocardial scar: accuracy of texture analysis on nonenhanced cine MR Images [J]. Radiology, 2018, 286(1): 103-112.
[11] Banka P, Robinson JD, Uppu SC, et al. Cardiovascular magnetic resonance techniques and findings in children with myocarditis: a multicenter retrospective study [J]. J Cardiovasc Magn Reson, 2015, 17: 96. doi: 10.1186/s12968-015-0201-6.
[12] Aquaro GD, Ghebru Habtemicael Y, Camastra G, et al. Prognostic value of repeating cardiac magnetic resonance in patients with acute myocarditis [J]. J Am Coll Cardiol, 2019, 74(20): 2439-2448.
[13] Aquaro GD, Perfetti M, Camastra G, et al. Cardiac MR with late gadolinium enhancement in acute myocarditis with preserved systolic function: ITAMY study [J]. J Am Coll Cardiol, 2017, 70(16): 1977-1987.
[14] Blissett S, Chocron Y, Kovacina B, et al. Diagnostic and prognostic value of cardiac magnetic resonance in acute myocarditis: a systematic review and meta-analysis [J]. Int J Cardiovasc Imaging, 2019, 35(12): 2221-2229.
[15] Yang F, Wang J, Li W, et al. The prognostic value of late gadolinium enhancement in myocarditis and clinically suspected myocarditis: systematic review and meta-analysis [J]. Eur Radiol, 2020, 30(5): 2616-2626.
[16] 李浩杰, 朱慧, 杨朝霞, 等. MR心肌应变在暴发性心肌炎初步应用及与心肌水肿相关性分析[J]. 影像诊断与介入放射学, 2020, 29(1): 48-53. LI Haojie, ZHU Hui, YANG Chaoxia, et al. Value of MR myocardial strain analysis in fulminant myocarditis[J]. Diagnostic Imaging Interventional Radiology, 2020, 29(1): 48-53.
[17] Awadalla M, Mahmood SS, Groarke JD, et al. Global longitudinal strain and cardiac events in patients with immune checkpoint inhibitor-related myocarditis [J]. J Am Coll Cardiol, 2020, 75(5): 467-478.
[18] Luetkens JA, Schlesinger-Irsch U, Kuetting DL, et al. Feature-tracking myocardial strain analysis in acute myocarditis: diagnostic value and association with myocardial oedema [J]. Eur Radiol, 2017, 27(11): 4661-4671.
[19] Buss SJ, Breuninger K, Lehrke S, et al. Assessment of myocardial deformation with cardiac magnetic resonance strain imaging improves risk stratification in patients with dilated cardiomyopathy [J]. Eur Heart J Cardiovasc Imaging, 2015, 16(3): 307-315.
[20] Pedrizzetti G, Claus P, Kilner PJ, et al. Principles of cardiovascular magnetic resonance feature tracking and echocardiographic speckle tracking for informed clinical use [J]. J Cardiovasc Magn Reson, 2016, 18(1): 51.
[21] Amzulescu MS, De Craene M, Langet H, et al. Myocardial strain imaging: review of general principles, validation, and sources of discrepancies [J]. Eur Heart J Cardiovasc Imaging, 2019, 20(6): 605-619.
[22] Claus P, Omar AMS, Pedrizzetti G, et al. Tissue tracking technology for assessing cardiac mechanics: principles, normal values, and clinical applications [J]. JACC Cardiovasc Imaging, 2015, 8(12): 1444-1460.
[23] Messroghli DR, Moon JC, Ferreira VM, et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: a consensus statement by the Society for Cardiovascular Magnetic Resonance(SCMR)endorsed by the European Association for Cardiovascular Imaging(EACVI)[J]. J Cardiovasc Magn Reson, 2017, 19(1): 75.
[24] Florian A, Ludwig A, Rösch S, et al. Myocardial fibrosis imaging based on T1-mapping and extracellular volume fraction(ECV)measurement in muscular dystrophy patients: diagnostic value compared with conventional late gadolinium enhancement(LGE)imaging [J]. Eur Heart J Cardiovasc Imaging, 2014, 15(9): 1004-1012.
[25] Gräni C, Eichhorn C, Bière L, et al. Prognostic value of cardiac magnetic resonance tissue characterization in risk stratifying patients with suspected myocarditis [J]. J Am Coll Cardiol, 2017, 70(16): 1964-1976. doi: 10.1016/j.jacc.2017.08.050.
[26] Kotanidis CP, Bazmpani MA, Haidich AB, et al. Diagnostic accuracy of cardiovascular magnetic resonance in acute myocarditis: a systematic review and meta-analysis [J]. JACC: Cardiovasc Imaging, 2018, 11(11): 1583-1590.

相关文章 15

[1]	廖原,门丹,李一帆,李怀臣,龙飞,刘铱. PM_2.5短期暴露对结核病发病风险的个体精准评估[J]. 山东大学学报 (医学版), 2026, 64(3): 116-123.
[2]	刘禹,霍娅娅,龚丞,梁婷,李斌. 基于机器学习的骨科生物材料设计与优化[J]. 山东大学学报 (医学版), 2026, 64(2): 22-33.
[3]	王建民,李晓峰,由志涛,董圣杰,赵宇驰,李占菊,邹德鑫,张剑锋,孙涛,杜伟. 基于可解释机器学习的后路腰椎椎体间融合术后慢性疼痛风险预测模型构建[J]. 山东大学学报 (医学版), 2026, 64(2): 78-88.
[4]	李梓绮,魏闫若雪,刘晓晗,刘春铖,赵然,刘玉昆. 长链非编码RNA HEATR3反义RNA 1参与结直肠癌发生发展的功能及其临床意义[J]. 山东大学学报 (医学版), 2025, 63(9): 108-115.
[5]	刘保国,宋翔,赵晓文,毛亚丽. 血清STAT5B、NKAIN1 mRNA检测在乳腺癌中的应用价值[J]. 山东大学学报 (医学版), 2025, 63(7): 68-74.
[6]	王磊,常霄,王梓萌,李娇娇,崔书君,杨飞,朱月香. 瘤内及瘤周DCE-MRI影像组学对宫颈癌患者无进展生存期的预测价值[J]. 山东大学学报 (医学版), 2025, 63(6): 45-54.
[7]	王丽云,高天勤,刘雨佳,陈青,陈柳,沙凯辉. 基于机器学习产后压力性尿失禁风险预测模型的构建及验证[J]. 山东大学学报 (医学版), 2025, 63(6): 55-66.
[8]	贾若曦,吕丽,刘涵云,吴寅平,李凤彩,赵泽华,王凯,范玉琛. 血细胞计数相关标志物对慢加急性乙型肝炎肝衰竭患者28天预后的诊断价值[J]. 山东大学学报 (医学版), 2025, 63(6): 89-99.
[9]	杜雪,李春霞,刘云霞,张涛. 基于MFPC-Cox的结直肠癌患者预后动态预测模型[J]. 山东大学学报 (医学版), 2025, 63(5): 101-110.
[10]	杨卫芳,徐宏,刘元涛,赵蕙琛. 促甲状腺激素受体抗体在Graves病复发中的作用机制及其临床意义[J]. 山东大学学报 (医学版), 2025, 63(4): 116-121.
[11]	刘文钊,张远,马湘萍,魏峰涛,卜培莉. 内皮活化和应激指数预测值与心力衰竭患者死亡风险的关联[J]. 山东大学学报 (医学版), 2025, 63(11): 27-35.
[12]	陈文亮,王欢欢,郝金锦,弓蕊,赵强,张飞,高磊,董静逊. lncRNA PVT1表达在胃癌预后评估及恶性进展中的作用:基于列线图模型与细胞功能实验的研究[J]. 山东大学学报 (医学版), 2025, 63(10): 61-71.
[13]	王静,刘晓菲,曾荣,许长娟,张锦涛,董亮. 基于机器学习算法鉴定哮喘的坏死性凋亡相关生物标志物[J]. 山东大学学报 (医学版), 2024, 62(7): 21-32.
[14]	郭振江,王宁,赵光远,杜立强,崔朝勃,刘防震. 基于机器学习建立术前预测近端胃癌食管切缘阳性模型[J]. 山东大学学报 (医学版), 2024, 62(7): 78-83.
[15]	郭姝画,樊扬,田风,王传新,杜鲁涛,李培龙,郭兴,徐硕. 微原纤维相关蛋白3在调控胶质瘤干细胞间充质表型转化中的作用[J]. 山东大学学报 (医学版), 2024, 62(6): 9-16.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed