北京生物医学工程

基于贝叶斯网络的老年综合评估过程简化方法

A simplified method for elderly comprehensive geriatric assessment process based on Bayesian network

作者：李关东邹函怡王艺桦许杨孙怡宁马祖长高理升

单位：中国科学院合肥物质科学研究院（合肥 230031）<br />中国科学技术大学（合肥  230026）<br />蚌埠医学院护理学院（安徽蚌埠233030）<br />通信作者：高理升, 副研究员。E-mail：lsgao@iim.ac.cn 

关键词：老年综合评估；问卷调查；贝叶斯网络；概率图模型；人工智能

分类号：R318

出版年·卷·期（页码）：2022·41·6（589-596）

摘要：

目的探究基于贝叶斯网络的老年综合评估过程简化方法在社区老年综合评估（comprehensive geriatric assessment, CGA）中的应用价值，为普及社区CGA工作提供一种可行的技术手段。方法以2018年中国老年健康影响因素跟踪调查（Chinese Longitudinal Healthy Longevity Survey，CLHLS）的横断面调查中的958位老年人为研究对象，根据CGA内容筛选有效样本。选择70%数据作为训练集构建贝叶斯网络模型并设计问卷调查算法。构建FPQM模型与基于贝叶斯网络的老年综合评估过程简化方法进行比较分析。以平均准确率、平均简化率、AFβ、占用空间作为模型的评价指标。结果基于贝叶斯网络的老年综合评估过程简化方法在测试集上的平均准确率、AFβ均最高，分别为0.983 6，0.849 9；占用空间最小，为23 kB。FPQM的平均简化率较高，为0.973 7；但平均准确率、AFβ较低，分别为0.842 0，0.787 6；占用空间巨大，达4.13 GB。结论基于贝叶斯网络的老年综合评估过程简化方法在准确率、综合性能和空间占用方面均明显优于已报道的简化算法。在当前老龄化日益严峻背景下，该方法对于基层社区和养老机构提升CGA工作效率具有重要应用价值。

Objective To explore the application value of the simplified method based on Bayesian network in community comprehensive geriatric assessment(CGA), and provides a feasible technical means for the popularization of community CGA. Methods A total of 958 elderly people from The Chinese Longitudinal Healthy Longevity Survey(CLHLS) 2018 cross-sectional survey were selected as the research subjects, and valid samples were selected according to the CGA content. 70% data were selected as training set to construct bayesian network model and design questionnaire algorithm. The FPQM model is compared with the simplified method based on Bayesian network. Average accuracy, average simplification rate, AFβ and occupied space were used as evaluation indexes of the model. Results The average accuracy and AFβ of the simplified method based on Bayesian network on the test set were 0.983 6 and 0.849 9, respectively. The minimum space is 23 kB. The average simplification rate of FPQM was 0.973 7. But the average accuracy and AFβ were lower, 0.842 0 and 0.787 6, respectively. It takes up a huge 4.13 GB of space. Conclusions The simplified method based on Bayesian network is superior to the reported simplified algorithm in terms of accuracy, comprehensive performance and space occupancy. In the context of increasingly severe aging, this method has important application value for grassroots communities and pension institutions to improve the work efficiency of CGA.

参考文献：

[1] 中华人民共和国国家统计局.中国统计年鉴 2021[M]. 北京: 中国统计出版社, 2021.
[2] 国务院办公厅. 国务院办公厅关于推进养老服务发展的意见[EB/OL]. (2022-07-22)[2022-08-02]. http://www.gov.cn/ zhengce/content/2019-04/16/content_5383270.htm.
[3] 陈旭娇，严静，王建业，等. 老年综合评估技术应用中国专家共识[J]. 中华老年医学杂志, 2017, 36(5): 471-477.
Chen XJ, Yan J, Wang JY, et al. Chinese experts consensus on application of comprehensive geriatric assessment[J]. Chinese Journal of Geriatrics, 2017, 36(5): 471-477.
[4] 都永丽，郭青，王英，等. 老年综合评估工具在社区护理中的应用现状及研究进展[J]. 中华现代护理杂志, 2020, 26(19): 2661-2664.
Du YL, Guo Q, Wang Y, et al. Application status and research progress of comprehensive geriatric assessment tool in the community nursing[J]. Chinese Journal of Modern Nursing, 2020, 26(19): 2661-2664.
[5] 李苗苗，王青，马玉霞，等. 基于快速筛查的老年综合评估指标体系构建[J]. 护理学杂志, 2019, 34(24): 80-83.
Li MM, Wang Q, Ma YX, et al. Construction of comprehensive geriatric assessment indicator system based on rapid screening[J]. Journal of Nursing Science, 2019, 34(24): 80-83.
[6] Tai CJ, Yang YH, Huang CY, et al. Development of the brief geriatric assessment for the general practitioner[J]. The Journal of Nutrition, Health & Aging, 2021, 25: 134-140.
[7] Mo T, Zhang R, Li W, et al. An influence-based fast preceding questionnaire model for elderly assessments[J]. Intelligent Data Analysis, 2018, 22(2): 407-437.
[8] Daniel RJ, Longmate J, Vani K, et al. Mining branching rules from past survey data with an illustration using a geriatric assessment survey for older adults with cancer[J]. Algorithms, 2016, 9(2): 33.
[9] 王皓珺，查光玉，曹明国，等. COVID-19疫情期间口腔本科生线上学习决策TAN贝叶斯网络模型分析[J]. 上海口腔医学, 2021, 30(5): 551-555.
Wang HJ, Zha GY, Cao MG, et al. TAN Bayesian network model for online learning decision-making of stomatological undergraduates during COVID-19[J]. Shanghai Journal of Stomatology, 2021, 30(5): 551-555.
[10] 陈雪龙，姜坤. 基于贝叶斯网络的并发型突发事件链建模方法[J]. 中国管理科学, 2021, 29(10): 165-177.
Chen XL, Jiang K. Modeling method of concurrent emergency chain based on Bayesian network[J]. Chinese Journal of Management Science, 2021, 29(10): 165-177.
[11] Aburai T, Okubo A, Suzuki D, et al. Bayesian network analysis for the questionnaire investigation on the needs at Fuji Shopping Street Town [J]. Journal of Management and Strategy, 2018, 9(2): 46-68.
[12] 北京大学健康老龄与发展研究中心. 中国老年健康影响因素跟踪调查(CLHLS)-追踪数据（1998-2018）[DB/OL].(2020-04-03)[2021-09-20]. https://doi.org/10.18170/DVN/WBO7LK.
[13] 中华人民共和国民政部. 中华人民共和国民政行业标准：老年人能力评估 (MZ/T 039—2013)[S].北京：中国标准出版社，2013.
[14] 吕志刚，李叶，王洪喜，等. 贝叶斯网络的结构学习综述[J]. 西安工业大学学报, 2021(1): 1-17.
Lyu ZG, Li Y, Wang HX, et al. Overview of Bayesian network structure learning[J]. Journal of Xi'an Technological University, 2021(1): 1-17.
[15] 宋财文，段霄雨. 贝叶斯网络推理与学习方法研究[J]. 信息记录材料, 2019, 20(6): 18-20.
Song CW, Duan XY. Research on reasoning and learning method of Bayesian network[J]. Information Recording Materials, 2019, 20(6): 18-20.
[16] 王君，陈天勇. 职业压力源量表的编制及信效度检验[J]. 中国健康心理学杂志, 2022, 30(5): 693-698.
Wang J, Chen TY. Development of occupational stressors scale and assessment on reliability and validity[J]. Chinese Journal of Health Psychology, 2022, 30(5): 693-698.
[17] 师亚，王秀华，刘永兵. 老年长期照护分级综合评价模型初步构建研究[J]. 中国全科医学, 2018, 21(34): 4254-4261.
Shi Y, Wang XH, Liu YB. Development of an integrative long-term care services classification model for the elderly[J]. Chinese General Practice, 2018, 21(34): 4254-4261.
[18] Gil-Begue S, Bielza C, Larraaga P. Multi-dimensional Bayesian network classifiers: a survey[J]. Artificial Intelligence Review, 2021, 54: 519-559.
[19] 王利民，刘洋，孙铭会，等. 基于Markov blanket的无约束型K阶贝叶斯集成分类模型[J]. 吉林大学学报(工学版), 2018, 48(6): 1851-1858.
Wang LM, Liu Y, Sun MH, et al. Ensemble of unrestricted K-dependence Bayesian classifiers based on Markov blanket [J]. Journal of Jilin University (Engineering and Technology Edition), 2018, 48(6): 1851-1858.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录