
浏览全部资源
扫码关注微信
1.陆军军医大学(第三军医大学)高原军事医学系:,寒区医学教研室,重庆
2.陆军军医大学(第三军医大学)高原军事医学系:,高原作业医学教研室,重庆
3.陆军军医大学(第三军医大学)高原军事医学系:,极端环境医学教育部重点实验室,重庆
张钢, E-mail:zhanggang196868@163.com
纸质出版日期:2025-01-15,
收稿日期:2024-08-26,
修回日期:2024-09-13,
移动端阅览
丁宇,王泽军,谢佳新,等.高原脱适应症发生风险预测模型的构建与验证[J].陆军军医大学学报,2025,47(01):20-29.
DING Yu,WANG Zejun,XIE Jiaxin,et al.Construction and validation of a risk prediction model for high altitude de-acclimatization syndrome[J].J Army Med Univ,2025,47(1):20-29.
丁宇,王泽军,谢佳新,等.高原脱适应症发生风险预测模型的构建与验证[J].陆军军医大学学报,2025,47(01):20-29. DOI:10.16016/j.2097-0927.202408110.
DING Yu,WANG Zejun,XIE Jiaxin,et al.Construction and validation of a risk prediction model for high altitude de-acclimatization syndrome[J].J Army Med Univ,2025,47(1):20-29. DOI:10.16016/j.2097-0927.202408110.
目的
2
应用不同机器学习算法构建从高原返回平原人群高原脱适应症(high altitude de-acclimatization syndrome,HADAS)发病的风险预测模型,并验证其预测效能。
方法
2
于2020年11月至2024年2月对结束高原生活返回内地的人群实地或线上发放问卷调查。收集基本资料、慢性高原病(chronic moun
tain sickness,CMS)情况和脱适应症状等调查资料,经筛选最终纳入1 095例作为建模组。阳性事件定义为脱适应症状评分>5分。将建模组按7∶3随机分为训练集(
n
=766)和内部测试集(
n
=329),采用最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)回归筛选自变量,基于多因素Logistic回归(multiple factor logistic regression,LR)、决策树(decision tree,DT)、随机森林(random forest,RF)、极端梯度提升(eXtreme gradient boosting,XGB)、支持向量机(support vector machine,SVM)、K最邻近(K-nearest neighbor,KNN)、轻度梯度提升(light gradient boosting,LGB)、朴素贝叶斯(naïve bayes,NB)8种机器学习方法构建预测模型。采用受试者工作特性曲线(receiver operating characteristic,ROC)、校准曲线和混淆矩阵对模型进行比较并进行内部测试;使用列线图或Shapley加性解释(shapley additive explanations,SHAP)图对最终模型进行展示。于2024年8月收集筛选结束高原生活返回平原132例作为外部验证组,对模型进行外部验证。
结果
2
1 095例调查人群中有脱适应症者549人(50.14%)。CMS评分、年龄、高原居住时间为LASSO回归筛选出的预测因子。8种机器学习算法建立的HADAS预测模型中,以LR模型最优,ROC的曲线下面积(area under curve,AUC)训练集为0.819(95%
CI
:0.789~0.850),内部测试集为0.841(95%
CI
:0.799~0.884),F1评分内部测试集为0.801,内部测试集的AUC、F1得分在8个模型中均为最大;LR模型校准曲线的Spiegelhalter
Z
检验显示训练集
P
=0.703、内部测试集
P
=0.281;LR模型外部验证集AUC为0.867(95%
CI
:0.765~0.969)。
结论
2
以CMS评分、年龄和高原居住时间为预测因子建立的LR模型在内部测试集的综合表现最好,在外部验证集中区分度好,构建的列线图便于应用。
Objective To construct risk models for predicting the occurrence of high altitude de-acclimatization syndrome (HADAS) in the population returning from the plateau to the plain based on different machine learning algorithms and validate the predicting efficiency of these models. Methods Field or online surveys were conducted on the individuals who had ended their high-altitude living and returned to the plain areas from November 2020 to February 2024. Basic information, chronic mountain sickness (CMS), HADAS symptoms and other data were collected. With the inclusion and exclusion criteria, totally 1 095 individuals were subjected and assigned into the modeling group. Positive events were defined as HADAS score
>
5. Then the m
odelling group was divided into a training set (
n
=766) and an internal test set (
n
=329) in a 7∶3 ratio. Least absolute shrinkage and selection operator (LASSO) regression was used to select independent variables. Risk prediction models for high-altitude adaptation symptoms were built based on 8 machine learning methods, including multiple factor logistic regression (LR), decision tree (DT), random forest (RF), eXtreme gradient boosting (XGB), support vector machine (SVM), K-nearest neighbor (KNN), light gradient boosting (LGB) and naïve bayes (NB). The models were compared and evaluated using receiver operating characteristic (ROC) curves, calibration curves and confusion matrices in the internal test set. The final model was presented using a nomogram or Shapley additive explanations (SHAP) algorithm. In August 2024, another 132 individuals who returned to the plains and met the same criteria were recruited and served as the external validation group. Results There were 549 individuals (50.14%) out of the 1 095 subjects having HADAS symptoms. LASSO regression identified CMS score, age and duration of high-altitude residence as significant predictors. Among the 8 machine learning algorithms, the LR model was identified as the best, with an area under the curve (AUC) value of 0.819 (95%
CI
: 0.789~0.850) and 0.841 (95
%CI
: 0.799~0.884), and an F1 score of 0.801 in the internal test set, respectively, and the AUC value and F1 score of the LR model were the largest among the 8 models in the internal test set. Spiegelhalter
Z
test of the calibration curve of the LR model indicated that its
P
=0.703 in the training set while
P
=0.281 in the internal test set. The AUC value of the LR model was 0.867(95
%CI
: 0.765~0.969) in the external validation set. Conclusion The LR model constructed based on indicators including CMS score, age and duration of high-altitude residence has a good overal
l performance in the internal test set, and good discriminating effect in the external validation set. The constructed nomogram is convenient for application.
预测模型高原脱适应症机器学习列线图
prediction modelhigh altitude de-acclimatization syndromemachine learningnomogram
高钰琪. 高原军事医学[M]. 重庆: 重庆出版社, 2005.
GAO Y Q. High altitude military medicine [M]. Chongqing: Chongqing Press, 2005.
周发权, 陈师, 刘若鸿, 等. 高原脱适应症及其临床症状检出率的荟萃分析[J]. 军事医学, 2020, 44(5): 331-337.
ZHOU F Q, CHEN S, LIU R H, et al. Detection rate of high altitude de-adaptation and clinical symptoms: a meta-analysis[J]. Mil Med Sci, 2020, 44(5): 331-337.
范勇, 周其全. 高原脱适应研究进展[J]. 解放军预防医学杂志, 2012, 30(3): 227-230.
FAN Y, ZHOU Q Q. Research progress of altitude maladjustment[J]. China Ind Econ, 2012, 30(3): 227-230.
刘爽, 龙贤哲, 罗勇军. 高原脱适应症诊断及防治研究进展[J]. 人民军医, 2017, 60(9): 941-943.
LIU S, LONG X Z, LUO Y J. Research progress on diagnosis, prevention and treatment of plateau de-indications[J]. People Mil Surg, 2017, 60(9): 941-943.
张卫花,姜钊. 高原脱适应症的危害及防治措施研究进展[J]. 中国疗养医学,2021,30(3):242-245. DOI:10.13517/j.cnki.ccm.2021.03.006http://dx.doi.org/10.13517/j.cnki.ccm.2021.03.006
ZHANG W H,JIANG Z. Advances in research on the hazards and prevention and treatment of plateau de-adaptation [J]. Chin J Convalesc Med,2021,30(3):242-245. DOI:10.13517/j.cnki.ccm.2021.03.006http://dx.doi.org/10.13517/j.cnki.ccm.2021.03.006
伍亚舟,陈锡程,易东. 人工智能在临床领域的研究进展及前景展望[J]. 陆军军医大学学报,2022,44(1):89-102.DOI:10.16016/j.1000-5404.202112092http://dx.doi.org/10.16016/j.1000-5404.202112092.
WU Y Z,CHEN X C,YI D. Advances and perspective of artificial intelligence in clinical area[J].J Army Med Univ,2022,44(1):89-102.DOI:10.16016/j.1000-5404.202112092http://dx.doi.org/10.16016/j.1000-5404.202112092.
周其全, 杨生岳, 袁振才, 等. 高原移居人群返回平原后高原脱适应症的诊断标准: 多中心、随机对照研究[J]. 解放军医学杂志, 2012, 37(2): 146-155.
ZHOU Q Q, YANG S Y, YUAN Z C, et al. A research in diagnostic criteria of high altitude de-adaptation for plateau migrants returning to the Plains: a multicenter, randomized controlled trial[J]. Med J Chin PLA, 2012, 37(2): 146-155.
国际高原医学会慢性高原病专家小组. 第六届国际高原医学和低氧生理学术大会颁布 慢性高原病青海诊断标准[J]. 青海医学院学报, 2005, 26(1): 3-5.
Expert Group of Chronic High Altitude Diseases of International Society for High Altitude Medicine.The 6th International Conference on Plateau Medicine and Hypoxia Physiology promulgated the diagnostic criteria of chronic altitude sickness in Qinghai[J]. J Qinghai Med Coll, 2005, 26(1): 3-5.
阳盛洪, 王引虎, 杨海军, 等. 高压氧治疗海拔5000m以上地区移居人群脱习服的临床疗效观察[J]. 西北国防医学杂志, 2013, 34(3): 205-207. DOI: 10.16021/j.cnki.1007-8622.2013.03.007http://dx.doi.org/10.16021/j.cnki.1007-8622.2013.03.007.
YANG S H, WANG Y H, YANG H J, et al. Clinical observation of hyperbaric oxygen treatment for de-acclimatization to high altitude for people above 5000 m altitude[J]. Med J Natl Defending Forces Northwest China, 2013, 34(3): 205-207. DOI: 10.16021/j.cnki.1007-8622.2013.03.007http://dx.doi.org/10.16021/j.cnki.1007-8622.2013.03.007.
COLLINS G S, REITSMA J B, ALTMAN D G, et al. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement[J]. BJOG, 2015, 122(3): 434-443. DOI: 10.1111/1471-0528.13244http://dx.doi.org/10.1111/1471-0528.13244.
LUO W, PHUNG D, TRAN T, et al. Guidelines for developing and reporting machine learning predictive models in biomedical research: a multidisciplinary view[J]. J Med Internet Res, 2016, 18(12): e323. DOI: 10.2196/jmir.5870http://dx.doi.org/10.2196/jmir.5870.
武松. 预测模型实战:基于R、SPSS和Stata(新时代·技术新未来)[M]. 北京: 清华大学出版社, 2023.
WU S. Predictive model practice with R, SPSS and Stata (New era·technology and new future)[M]. Beijing: Tsinghua University Press, 2023.
HEATH D, DONALD D E, EDWARDS J E. Pulmonary vascular changes in a dog after aortopulmonary anastomosis for four years[J]. Br Heart J, 1959, 21(2): 187-196. DOI: 10.1136/hrt.21.2.187http://dx.doi.org/10.1136/hrt.21.2.187.
丁魁, 李权超, 王燕, 等. 高原驻训士兵返回平原后高原脱适应症与训练倦怠的关系研究[J]. 军事医学, 2016, 40(8): 661-664, 694. DOI: 10.7644/j.issn.1674-9960.2016.08.010http://dx.doi.org/10.7644/j.issn.1674-9960.2016.08.010.
DING K, LI Q C, WANG Y, et al. Relationships between high altitude de-adaptation syndrome and training burnout of plateau soldiers after returning to the plain[J]. Mil Med Sci, 2016, 40(8): 661-664, 694. DOI: 10.7644/j.issn.1674-9960.2016.08.010http://dx.doi.org/10.7644/j.issn.1674-9960.2016.08.010.
WEI C Y, CHEN P N, LIN S S, et al. Using machine learning to determine the correlation between physiological and environmental parameters and the induction of acute mountain sickness[J]. BMC Bioinformatics, 2022, 22(Suppl 5): 628. DOI: 10.1186/s12859-022-04749-0http://dx.doi.org/10.1186/s12859-022-04749-0.
YANG M, WU Y, YANG X B, et al. Establishing a prediction model of severe acute mountain sickness using machine learning of support vector machine recursive feature elimination[J]. Sci Rep, 2023, 13(1): 4633. DOI: 10.1038/s41598-023-31797-0http://dx.doi.org/10.1038/s41598-023-31797-0.
王胜锋, 王俊峰, 谷鸿秋, 等. 临床预测模型方法与应用[M]. 北京: 北京大学医学出版社, 2024.
WANG S F, WANG J F, GU H Q, et al. Clinical predicting models methods and application [M]. Beijing, Peking University Press, 2024.
HONG C, PENCINA M J, WOJDYLA D M, et al. Predictive accuracy of stroke risk prediction models across black and white race, sex, and age groups[J]. JAMA, 2023, 329(4): 306-317. DOI: 10.1001/jama.2022.24683http://dx.doi.org/10.1001/jama.2022.24683.
EFTHIMIOU O, SEO M, CHALKOU K, et al. Developing clinical prediction models: a step-by-step guide[J]. BMJ, 2024, 386: e078276. DOI: 10.1136/bmj-2023-078276http://dx.doi.org/10.1136/bmj-2023-078276.
RILEY R D, ARCHER L, SNELL K I E, et al. Evaluation of clinical prediction models (part 2): how to undertake an external test study[J]. BMJ, 2024, 384: e074820. DOI: 10.1136/bmj-2023-074820http://dx.doi.org/10.1136/bmj-2023-074820.[PubMed]
谢佳新, 吴玉, 董华平, 等. 高原驻训人员高原脱适应症发病特征与影响因素调查分析[J]. 西南国防医药, 2021, 31(6): 556-560.
XIE J X, WU Y, DONG H P, et al. Investigation and analysis on epidemiological features and influencing factors of HADAS among encamping training personnel at high altitude[J]. Med J Natl Defending Forces Southwest China, 2021, 31(6): 556-560.
李政波, 张进, 王雪. 高原低氧环境短期暴露对官兵心脏的影响[J]. 西北国防医学杂志, 2019, 40(3): 174-178. DOI: 10.16021/j.cnki.1007-8622.2019.03.009http://dx.doi.org/10.16021/j.cnki.1007-8622.2019.03.009.
LI Z B, ZHANG J, WANG X. Effect of short-term exposure to plateau hypoxic environment on the heart of officers and soldiers[J]. Med J Natl Defending Forces Northwest China, 2019, 40(3): 174-178. DOI: 10.16021/j.cnki.1007-8622.2019.03.009http://dx.doi.org/10.16021/j.cnki.1007-8622.2019.03.009.
丰丽英, 朱前勇, 王保健, 等. 医疗分队高原驻训后脱适应症发生情况调查[J]. 人民军医, 2017, 60(2): 114-117.
FENG L Y, ZHU Q Y, WANG B J, et al. Investigation on the occurrence of out-of-indications of medical teams after training at high altitude[J]. People Mil Surg, 2017, 60(2): 114-117.
魏征华, 资云海, 董志伟, 等. 高原移居者返回平原的高原脱适应症发生率比较[J]. 解放军医学杂志, 2012, 37(12): 1151-1154.
WEI Z H, ZI Y H, DONG Z W, et al. Analysis on the deadaptation symptoms of plain region migrants back from different high altitude regions[J]. Med J Chin PLA, 2012, 37(12): 1151-1154.
曾以萍, 周其全, 邬云红, 等. 久居高原者返回平原后脱适应症状分析[J]. 西藏医药, 2014, 35(1): 46-48.
ZENG Y P, ZHOU Q Q, WU Y H, et al. Analysis of maladjustment symptoms of people who have lived in plateau for a long time after returning to plain[J]. Tibet Med, 2014, 35(1): 46-48.
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621