切换至 "中华医学电子期刊资源库"

中华诊断学电子杂志 ›› 2019, Vol. 07 ›› Issue (04) : 233 -238. doi: 10.3877/cma.j.issn.2095-655X.2019.04.004

所属专题: 文献

心脑血管疾病诊治

基于随机森林的男性急性心肌梗死诊断模型建立及验证
吕永楠1, 李迪2, 李艳2,()   
  1. 1. 430060 武汉大学人民医院心血管内科
    2. 430060 武汉大学人民医院检验医学中心
  • 收稿日期:2019-05-24 出版日期:2019-11-26
  • 通信作者: 李艳
  • 基金资助:
    国家自然科学基金(81772265); 湖北省自然科学基金(2017CFB172); 武汉大学人民医院引导基金(RMYD2018M15); 武汉大学医学部创新种子基金培育项目(TFZZ2018034)

Establishment and validation of diagnosis model for acute myocardial infarction based on random forest classification in men

Yongnan Lyu1, Di Li2, Yan Li2,()   

  1. 1. Department of Cardiology, People′s Hospital of Wuhan University, Wuhan 430060, China
    2. Department of Laboratory Medicine Centre, People′s Hospital of Wuhan University, Wuhan 430060, China
  • Received:2019-05-24 Published:2019-11-26
  • Corresponding author: Yan Li
  • About author:
    Corresponding author: Li Yan, Email:
引用本文:

吕永楠, 李迪, 李艳. 基于随机森林的男性急性心肌梗死诊断模型建立及验证[J/OL]. 中华诊断学电子杂志, 2019, 07(04): 233-238.

Yongnan Lyu, Di Li, Yan Li. Establishment and validation of diagnosis model for acute myocardial infarction based on random forest classification in men[J/OL]. Chinese Journal of Diagnostics(Electronic Edition), 2019, 07(04): 233-238.

目的

利用随机森林建立及验证男性急性心肌梗死诊断模型。

方法

检测2016年1至6月于武汉大学人民医院心内科住院的205例心绞痛或急性心肌梗死男性患者的血清生化及生物标志物指标,其中151例患者作为训练集,54例患者作为验证集。用随机森林对指标预测急性心肌梗死的重要性进行排序。根据袋外数据误差,赤池信息量准则和贝叶斯信息量准则对排序指标进行筛选并构建诊断模型;多维标度法(MDS)观察模型对急性心肌梗死和心绞痛的区分能力;用验证集数据验证模型对心绞痛和急性心肌梗死的鉴别能力。

结果

19个指标根据平均准确度下降程度和平均基尼(Gini)指数下降程度进行重要性排序。用袋外数据误差,赤池信息量准则和贝叶斯信息量准则筛选出C-反应蛋白、中性粒细胞绝对值和血糖3个变量,并纳入模型。通过MDS法观察到多半样本得到良好的区分,但部分样本仍难以区分开。在外部验证中,31例急性心肌梗死患者有26例(83.87%)被正确识别;在23例心绞痛患者中有19例(82.61%)被正确识别。

结论

基于随机森林的诊断模型建立能较好区分急性心肌梗死与心绞痛。

Objective

To establish and validate the model of forecasting acute myocardial infarction in men.

Methods

From January to June 2016, 205 male patients admitted to the department of cardiology of the People′s Hospital of Wuhan University with angina pectoris or acute myocardial infarction were included in our study. Among them, 151 patients served as training set and 54 patients served as validation set. Random forest was used to rank the importance of predicting acute myocardial infarction. According to the OOB error, AIC and BIC criterion, the sorting indexs were screened and the prediction model was constructed. Multidimensional scaling (MDS) was used to observe the ability of the model to differentiate acute myocardial infarction from angina pectoris, and validation set data was used to investigate whether the random forest could distinguish between acute myocardial infarction and angina pectoris.

Results

The 19 indicators were ranked according to mean Decrease Accuracy and mean Decrease Gini index. C-reactive protein, neutrophil absolute value and blood sugar inclusion model were screened by OOB error, AIC criterion and BIC criterion. In external validation, 26 of 31(83.87%) patients with acute myocardial infarction were correctly identified, and 19 of 23(82.61%) patients with angina pectoris were correctly identified.

Conclusion

Random forest-based predictive model can well distinguish between acute myocardial infarction and angina pectoris.

表1 训练集与验证集患者的临床特征比较
变量 训练集(n=151) 验证集(n=54) 统计量 P 调整的P
年龄(岁,±s) 59.61±9.35 58.96±9.20 t=-0.285 0.662 0.884
急性心肌梗死(例,%) 71 (47.02) 31 (57.41) χ2=1.717 0.125 0.884
高血压(例,%) 86 (56.95) 29 (53.70) χ2=0.171 0.399 0.884
糖尿病(例,%) 22 (14.57) 22 (40.74) χ2=0.377 0.342 0.884
Gensini评分[分,中位数(四分位间距)] 32.00[12.50~64.00] 29.75[15.13~76.25] Z=-0.340 0.734 0.889
C-反应蛋白(mg/L,[中位数(四分位间距)]) 2.20[0.37~11.45] 2.74[0.61~13.41] Z=-0.551 0.582 0.884
白细胞计数(109/L,[中位数(四分位间距)]) 7.60[6.00~9.60] 8.40[6.20~10.05] Z=-1.242 0.214 0.884
中性粒细胞绝对值(109/L,[中位数(四分位间距)]) 4.92[3.48~7.15] 5.83[3.79~8.12] Z=-1.280 0.200 0.884
淋巴细胞绝对值(109/L,[中位数(四分位间距)]) 1.64[1.32~2.07] 1.73[1.34~2.10] Z=-0.396 0.692 0.884
单核细胞绝对值(109/L,[中位数(四分位间距)]) 0.60[0.40~0.70] 0.60[0.48~0.83] Z=-0.695 0.487 0.884
血红蛋白(g/L,[中位数(四分位间距)]) 141.00[131.00~153.00] 143.50[132.75~153.25] Z=-0.491 0.624 0.884
血糖(mmol/L,[中位数(四分位间距)]) 5.80[5.10~6.84] 5.70[5.17~7.35] Z=-0.102 0.919 0.961
尿酸(μmol/L,[中位数(四分位间距)]) 382.00[332.00~440.00] 390.50[340.00~458.50] Z=-0.691 0.490 0.884
尿素氮(mmol/L,[中位数(四分位间距)]) 5.90[4.86~7.30] 6.25[5.07~7.25] Z=-0.646 0.519 0.884
肌酐(μmol/L,[中位数(四分位间距)]) 77.00[67.00~91.00] 74.00[63.75~92.25] Z=-0.127 0.899 0.961
总胆固醇(mmol/L,[中位数(四分位间距)]) 4.00[3.40~4.63] 3.70[3.35~4.59] Z=-1.139 0.255 0.884
总甘油三酯(mmol/L,[中位数(四分位间距)]) 1.46[1.05~2.11] 1.68[0.97~2.34] Z=-0.941 0.347 0.884
高密度脂蛋白胆固醇(mmol/L,[中位数(四分位间距)]) 0.89[0.75~1.06] 0.88[0.70~1.08] Z=-0.711 0.477 0.884
低密度脂蛋白胆固醇(mmol/L,[中位数(四分位间距)]) 2.21[1.71~2.88] 1.99[1.59~2.52] Z=-1.577 0.115 0.884
睾酮(ng/dL,[中位数(四分位间距)]) 300.00[211.00~392.00] 317.50[223.00~379.50] Z=-0.568 0.570 0.884
雌二醇(pmol/L,[中位数(四分位间距)]) 32.00[25.00~41.00] 31.00[25.75~42.00] Z=-0.033 0.973 0.973
睾酮/雌二醇[中位数(四分位间距)] 9.45[6.60~15.77] 9.66[6.98~13.28] Z=-0.513 0.608 0.884
图1 各变量在鉴别心绞痛和急性心肌梗死患者中重要性的排序
表2 基于袋外数据误差、赤池信息量准则和贝叶斯信息量准则的变量筛选
图2 多维标度法展示模型对心绞痛和急性心肌梗死患者的区分能力
[1]
陈凤英. 急性ST段抬高型心肌梗死的救治及中国经皮冠状动脉介入治疗指南(2016)要点解读[J]. 中华急诊医学杂志,2016,25(7):857-861.
[2]
王威廉.机器学习:现在与未来[J].科技导报,2016,34(7):78-79.
[3]
董师师,黄哲学.随机森林理论浅析[J]. 集成技术,2013,2(1):1-7.
[4]
Task force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC), Steg PG, James SK,et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation[J].Eur Heart J,2012,33(20):2569-2619.
[5]
Gensini GG. A more meaningful scoring system for determining the severity of coronary heart disease[J]. Am J Cardiol,1983,51(3):606.
[6]
Sandri M, Zuccolotto P.Variable selection using random forests[M].Berlin:springer-Verlag,2012:263-270.
[7]
夏欣. 冠状动脉血管成像检查与冠状动脉造影诊断冠心病的效果分析及其与C-反应蛋白的相关性[J]. 实用医院临床杂志,2018,15(1):145-148.
[8]
戚厚兴,崔芬,张金国,等. 首发与再发急性心肌梗死患者先兆诱因及危险因素的临床研究[J/CD]. 中华诊断学电子杂志,2017,5(2):98-102.
[9]
Westman PC, Lipinski MJ, Luger D,et al. Inflammation as a driver of adverse left ventricular remodeling after acute myocardial infarction[J].J Am Coll Cardiol,2016,67(17):2050-2060.
[10]
赵莉莉,宋衍秋,张莹,等. 人金属蛋白酶1组织抑制剂基因表达对大鼠心肌梗死后炎症反应的影响及机制[J]. 中华心血管病杂志,2017,45(6):513-518.
[11]
叶飘,雷晓明. 入院首次中性粒细胞/淋巴细胞比值与急性心肌梗死患者院内主要不良心血管事件发生的相关性研究[J]. 广州医药,2018,49(5):13-17.
[12]
Wahab NN, Cowden EA, Pearce NJ, et al. Is blood glucose an independent predictor of mortality in acute myocardial infarction in the thrombolytic era?[J]. J Am Coll Cardiol,2002,40(10):1748-1754.
[13]
Caprnda M, Mesarosova D, Ortega PF,et al.Glycemic variability and vascular complications in patients with type 2 diabetes mellitus[J].Folia Med (Plovdiv),2017,59(3):270-278.
[14]
Gurung RL, Yiamunaa M, Liu S,et al. Association of haptoglobin phenotype with incident acute myocardial infarction in Chinese patients with type 2 diabetes[J]. Cardiovasc Diabetol,2019,18(1):65.
[15]
Breiman L.Random Forests[J].Machine Learning,2001,45(1):5-32.
[16]
Pavey TG, Gilson ND, Gomersall SR,et al.Field evaluation of a random forest activity classifier for wrist-worn accelerometer data[J].J Sci Med Sport,2017,20(1):75-80.
[17]
Hsieh CH, Lu RH, Lee NH, et al. Novel solutions for an old disease:diagnosis of acute appendicitis with random forest,support vector machines,and artificial neural networks[J]. Surgery,2011,149(1):87-93.
[18]
Pang H, Lin A, Holford M,et al. Pathway analysis using random forests classification and regression[J]. Bioinformatics,2006,22(16):2028-2036.
[1] 罗烨, 胡梦铃, 黄小凡, 林金鹏, 李竺蔓, 王少白. 支持向量机用于膝骨关节炎和韧带损伤的分类研究[J/OL]. 中华关节外科杂志(电子版), 2024, 18(02): 201-208.
[2] 王友芳, 李兴超, 刘清敏, 刘德彬, 刘松伍, 郭冬冬, 车峰远. 应激性高血糖指数对经皮冠状动脉介入术后急性心肌梗死患者发生主要不良心脑血管事件的预测价值[J/OL]. 中华危重症医学杂志(电子版), 2024, 17(02): 124-129.
[3] 黄艺承, 梁海祺, 何其焕, 韦发烨, 杨舒博, 谭舒婷, 翟高强, 程继文. 机器学习模型评估RAS亚家族基因对膀胱癌免疫治疗的作用[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(02): 131-140.
[4] 杨龙雨禾, 王跃强, 招云亮, 金溪, 卫娜, 杨智明, 张贵福. 人工智能辅助临床决策在泌尿系肿瘤的应用进展[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(02): 178-182.
[5] 张伟伟, 陈启, 翁和语, 黄亮. 随机森林模型预测T1 期结直肠癌淋巴结转移的初步研究[J/OL]. 中华结直肠疾病电子杂志, 2024, 13(05): 389-393.
[6] 周璐, 温镕博, 刘子璇, 李奕飏, 周乐其, 朱晓明, 龚海峰, 高显华, 楼征, 刘连杰, 郝立强, 于冠宇, 张卫. Changhai-AL-Prediction预测模型指导预防性造口合理实施的卫生经济学研究[J/OL]. 中华结直肠疾病电子杂志, 2024, 13(04): 280-287.
[7] 潘清, 葛慧青. 基于机械通气波形大数据的人机不同步自动监测方法[J/OL]. 中华重症医学电子杂志, 2024, 10(04): 399-403.
[8] 孙铭远, 褚恒, 徐海滨, 张哲. 人工智能应用于多发性肺结节诊断的研究进展[J/OL]. 中华临床医师杂志(电子版), 2024, 18(08): 785-790.
[9] 刘亮, 肖浩, 崔晓磊, 吕宝谱, 张睿, 郑拓康, 孟庆冰, 姚冬奇, 田英平, 高恒波. 急性心肌梗死合并心源性休克患者预后因素分析97例[J/OL]. 中华临床医师杂志(电子版), 2024, 18(02): 183-189.
[10] 林永俭, 谢雪花, 郝莉茹, 刘丽, 马英东. 优化急诊绿色通道对急性心肌梗死介入治疗患者救治时间的影响[J/OL]. 中华介入放射学电子杂志, 2024, 12(02): 185-189.
[11] 刘聪辉, 何浩然, 黄一诺, 张凤, 王凡月, 郝翰. 膳食铜补充对大鼠心肌梗死后心肌基质金属蛋白酶2表达水平及血流动力学的影响[J/OL]. 中华诊断学电子杂志, 2024, 12(03): 166-172.
[12] 郭方明, 赵明俐, 颜凡辉, 刘萌萌, 王阳, 赵英杰, 刘远航, 张艳芬, 詹景冬. 光学相干断层成像在急性心肌梗死冠状动脉分层斑块病变中的应用[J/OL]. 中华诊断学电子杂志, 2024, 12(02): 73-79.
[13] 徐来英, 程效, 戴亨纷, 侯俊凉, 苏怡林, 张彦. 药物联合个体化精准恒定功率运动疗法治疗心肌梗死术后频发室性早搏一例[J/OL]. 中华心脏与心律电子杂志, 2024, 12(03): 176-179.
[14] 单兴华, 唐文栋, 赵仙先. 延伸导管在室间隔穿孔介入治疗的新应用一例[J/OL]. 中华心脏与心律电子杂志, 2024, 12(02): 126-128.
[15] 郑屹, 刘莹, 张煜坤, 李广平, 陈康寅, 刘彤. 既往及新发心房颤动对急性心肌梗死患者远期卒中风险的影响[J/OL]. 中华脑血管病杂志(电子版), 2024, 18(05): 406-417.
阅读次数
全文


摘要