联合决策树及logistic回归建立乳腺癌相对风险预测模型
|
摘要
目的运用决策树和logistic回归建立乳腺癌相对风险预测模型,实现分级管理。方法本研究为病例对照设计,病例来源于2014-2015年就诊于四川大学华西医院、四川省肿瘤医院和四川省人民医院的原发性乳腺癌783例;对照来源于2009-2012年四川省妇幼中心和成都市双流区妇幼保健院的筛查队列3 879例。运用问卷收集相关信息,构建决策树,采用logistic回归模型计算OR值,并评价回归拟合效果。结果决策树选择乳腺良性疾病史、绝经状态、初产年龄(22/24岁)、活产次数(1个)、年龄(50岁)及行经时间(34年)作为危险因素,组合危险因素logistic回归模型中OR取值范围1-65.71(95%CI:32.19,134.13),其中有乳腺良性疾病史的未绝经妇女患病风险最高(OR=65.71,95%CI:32.19,134.13),回归模型灵敏度0.68,特异度0.79,AUC 0.8038。结论联合决策树和组合危险因素logistic回归能区分不同患病风险的危险因素组合,可用于构建相对风险预测模型。
Objective To establish a prediction model of relative risk for breast cancer using decision tree and logistic regression,and to provide different management strategies for breast cancer.Methods A case-control design was conducted with 783 newly diagnosed cases in West China Hospital of Sichuan University,Sichuan Cancer Hospital and Sichuan Provincial People′s Hospital from 2014 to 2015,and 3,879 controls in Sichuan Women and Children Center and Chengdu Shuangliu District Maternal and Child Health Hospital from 2009 to 2012.A standard questionnaire was used to collect information,and a decision tree model was constructed.The logistic regression was used to calculate the OR value and evaluate predictive power.Results The decision tree selected 6 variables including the history of benign breast diseases,menopausal status,the age of first delivery,number of living birth,age,and total menstrual duration as risk factors.The OR value ranges of logistic regression model using a combination of risk factors were from 1 to 65.71(95%CI:32.19-134.13),among which the risk of non-menopausal women with a history of benign breast disease was the highest(OR=65.71,95%CI:32.19-134.13).For the regression model,the sensitivity was 0.68,and the specificity was 0.79 with the AUC of 0.8038.Conclusion Decision tree and logistic regression can be used to construct a prediction model of relative risk.
Objective To establish a prediction model of relative risk for breast cancer using decision tree and logistic regression,and to provide different management strategies for breast cancer.Methods A case-control design was conducted with 783 newly diagnosed cases in West China Hospital of Sichuan University,Sichuan Cancer Hospital and Sichuan Provincial People′s Hospital from 2014 to 2015,and 3,879 controls in Sichuan Women and Children Center and Chengdu Shuangliu District Maternal and Child Health Hospital from 2009 to 2012.A standard questionnaire was used to collect information,and a decision tree model was constructed.The logistic regression was used to calculate the OR value and evaluate predictive power.Results The decision tree selected 6 variables including the history of benign breast diseases,menopausal status,the age of first delivery,number of living birth,age,and total menstrual duration as risk factors.The OR value ranges of logistic regression model using a combination of risk factors were from 1 to 65.71(95%CI:32.19-134.13),among which the risk of non-menopausal women with a history of benign breast disease was the highest(OR=65.71,95%CI:32.19-134.13).For the regression model,the sensitivity was 0.68,and the specificity was 0.79 with the AUC of 0.8038.Conclusion Decision tree and logistic regression can be used to construct a prediction model of relative risk.
引文
[1] 陈万青,孙可欣,郑荣寿,等.2014年中国分地区恶性肿瘤发病和死亡分析[J].中国肿瘤,2018(1):1-14.
[2] Meads C,Ahmed I,Riley RD.A systematic review of breast cancer incidence risk prediction models with meta-analysis of their performance[J].Breast Cancer Research and Treatment,2012,132(2):365-377.
[3] Cintolo-Gonzalez JA,Braun D,Blackford AL,et al.Breast cancer risk models:a comprehensive overview of existing models,validation,and clinical applications[J].Breast Cancer Research and Treatment,2017,164(2):263-284.
[4] 张瑾,陈薇,刘蕾.2017年《NCCN乳腺癌筛查和诊断临床实践指南》更新与解读[J].中国全科医学,2017,20(24):2939-2943.
[5] 吴辰文,李长生,王伟,等.一种改进的SVM算法在乳腺癌诊断方面的应用[J].计算机工程与科学,2017,39(3):562-566.
[6] Zhao Y,Xiong P,Mccullough LE,et al.Comparison of breast cancer risk predictive models and screening strategies for Chinese women[J].Journal of Women′s Health,2017,26(3):294-302.
[7] 帅健,李丽萍,陈业群.决策树模型与Logistic回归模型在伤害发生影响因素分析中的作用[J].中华疾病控制杂志,2015,19(2):185-189.
[8] 李泓波,白劲波,杨高明,等.决策树技术研究综述[J].电脑知识与技术,2015,11(24):1-4.
[9] Huang ML,Hung YH,Lee WM,et al.Usage of Case-Based reasoning,neural network and adaptive Neuro-Fuzzy inference system classification techniques in breast cancer dataset classification diagnosis[J].Journal of Medical Systems,2012,36(2):407-414.
[10] Li S,Yu KD,Fan L,et al.Predicting breast cancer recurrence following breast-conserving therapy:a single-institution analysis consisting of 764 Chinese breast cancer cases[J].Annals of Surgical Oncology,2011,18(9):2492-2499.
[11] 李芹,刁莎,李卉,等.运用决策树建立中国西南地区女性乳腺癌非遗传因素风险等级模型[J].中华肿瘤杂志,2018,40(11):872-877.
[12] Fan L,Strasser-Weippl K,Li JJ,et al.Breast cancer in China[J].The lancet oncology,2014,15(7):e279-e289.
[13] 刘璐,郑新宇.乳腺良性病变的组织学分型及其乳腺癌风险[J].中国实用外科杂志,2016,36(7):720-724.
[14] Dyrstad SW,Yan Y,Fowler AM,et al.Breast cancer risk associated with benign breast disease:systematic review and meta-analysis[J].Breast Cancer Research and Treatment,2015,149(3):569-575.
[15] 葛冰磊,洪学军,王青青.雌激素和孕激素对人乳腺癌细胞MCF-7 RCAS1表达的影响[J].检验医学,2007,22(2):114-118.
[16] Li J,Zhang BN,Fan JH,et al.A nation-wide multicenter 10-year(1999-2008)retrospective clinical epidemiological study of female breast cancer in China[J].BMC Cancer,2011,11:364.
[17] Lambertini M,Santoro L,Del Mastro L,et al.Reproductive behaviors and risk of developing breast cancer according to tumor subtype:a systematic review and meta-analysis of epidemiological studies[J].Cancer Treatment Reviews,2016,49:65-76.
[18] 章进,杨玉欢,梅勇,等.生育次数与中国女性乳腺癌相关性的Meta分析[J].中国循证医学杂志,2015(10):1148-1152.
[19] Kocdor H,Kocdor MA,Russo J,et al.Human chorionic gonadotropin(hCG)prevents the transformed phenotypes induced by 17 β-estradiol in human breast epithelial cells[J].Cell Biology International,2009,33(11):1135-1143.
[20] Kobayashi S,Sugiura H,Ando Y,et al.Reproductive history and breast cancer risk[J].Breast Cancer,2012,19(4):302-308.
[21] Woods KL,Smith SR,Morrison JM.Parity and breast cancer:evidence of a dual effect[J].British Medical Journal,1980,281(6237):419-421.
[2] Meads C,Ahmed I,Riley RD.A systematic review of breast cancer incidence risk prediction models with meta-analysis of their performance[J].Breast Cancer Research and Treatment,2012,132(2):365-377.
[3] Cintolo-Gonzalez JA,Braun D,Blackford AL,et al.Breast cancer risk models:a comprehensive overview of existing models,validation,and clinical applications[J].Breast Cancer Research and Treatment,2017,164(2):263-284.
[4] 张瑾,陈薇,刘蕾.2017年《NCCN乳腺癌筛查和诊断临床实践指南》更新与解读[J].中国全科医学,2017,20(24):2939-2943.
[5] 吴辰文,李长生,王伟,等.一种改进的SVM算法在乳腺癌诊断方面的应用[J].计算机工程与科学,2017,39(3):562-566.
[6] Zhao Y,Xiong P,Mccullough LE,et al.Comparison of breast cancer risk predictive models and screening strategies for Chinese women[J].Journal of Women′s Health,2017,26(3):294-302.
[7] 帅健,李丽萍,陈业群.决策树模型与Logistic回归模型在伤害发生影响因素分析中的作用[J].中华疾病控制杂志,2015,19(2):185-189.
[8] 李泓波,白劲波,杨高明,等.决策树技术研究综述[J].电脑知识与技术,2015,11(24):1-4.
[9] Huang ML,Hung YH,Lee WM,et al.Usage of Case-Based reasoning,neural network and adaptive Neuro-Fuzzy inference system classification techniques in breast cancer dataset classification diagnosis[J].Journal of Medical Systems,2012,36(2):407-414.
[10] Li S,Yu KD,Fan L,et al.Predicting breast cancer recurrence following breast-conserving therapy:a single-institution analysis consisting of 764 Chinese breast cancer cases[J].Annals of Surgical Oncology,2011,18(9):2492-2499.
[11] 李芹,刁莎,李卉,等.运用决策树建立中国西南地区女性乳腺癌非遗传因素风险等级模型[J].中华肿瘤杂志,2018,40(11):872-877.
[12] Fan L,Strasser-Weippl K,Li JJ,et al.Breast cancer in China[J].The lancet oncology,2014,15(7):e279-e289.
[13] 刘璐,郑新宇.乳腺良性病变的组织学分型及其乳腺癌风险[J].中国实用外科杂志,2016,36(7):720-724.
[14] Dyrstad SW,Yan Y,Fowler AM,et al.Breast cancer risk associated with benign breast disease:systematic review and meta-analysis[J].Breast Cancer Research and Treatment,2015,149(3):569-575.
[15] 葛冰磊,洪学军,王青青.雌激素和孕激素对人乳腺癌细胞MCF-7 RCAS1表达的影响[J].检验医学,2007,22(2):114-118.
[16] Li J,Zhang BN,Fan JH,et al.A nation-wide multicenter 10-year(1999-2008)retrospective clinical epidemiological study of female breast cancer in China[J].BMC Cancer,2011,11:364.
[17] Lambertini M,Santoro L,Del Mastro L,et al.Reproductive behaviors and risk of developing breast cancer according to tumor subtype:a systematic review and meta-analysis of epidemiological studies[J].Cancer Treatment Reviews,2016,49:65-76.
[18] 章进,杨玉欢,梅勇,等.生育次数与中国女性乳腺癌相关性的Meta分析[J].中国循证医学杂志,2015(10):1148-1152.
[19] Kocdor H,Kocdor MA,Russo J,et al.Human chorionic gonadotropin(hCG)prevents the transformed phenotypes induced by 17 β-estradiol in human breast epithelial cells[J].Cell Biology International,2009,33(11):1135-1143.
[20] Kobayashi S,Sugiura H,Ando Y,et al.Reproductive history and breast cancer risk[J].Breast Cancer,2012,19(4):302-308.
[21] Woods KL,Smith SR,Morrison JM.Parity and breast cancer:evidence of a dual effect[J].British Medical Journal,1980,281(6237):419-421.