Gout Staging Diagnosis Method Based on Deep Reinforcement Learning

被引:3
作者
Ma, Chao [1 ]
Pan, Changgang [1 ]
Ye, Zi [1 ]
Ren, Hanbin [1 ]
Huang, Hai [1 ]
Qu, Jiaxing [2 ]
机构
[1] Harbin Univ Sci & Technol, Sch Comp Sci & Technol, Harbin 150080, Peoples R China
[2] Heilongjiang Prov Cyberspace Res Ctr, Harbin 150090, Peoples R China
关键词
electronic medical records; gout; disease staging diagnosis; deep reinforcement learning;
D O I
10.3390/pr11082450
中图分类号
TQ [化学工业];
学科分类号
0817 ;
摘要
In clinical practice, diseases with a prolonged course and disease characteristics at the time of diagnosis are often classified into specific stages. The precision of disease staging significantly impacts the therapeutic and curative outcomes for patients, and the diagnosis of multi-clinical-stage diseases based on electronic medical records is a problem that needs further research. Gout is a multi-stage disease. This paper focuses on the research of gout and proposes a staging diagnosis method for gout based on deep reinforcement learning. This method firstly uses the candidate binary classification model library for accurate diagnosis of gout, and then corrects the results of the binary classification through the set medical rules for diagnosis of gout, and then uses the machine learning model to diagnose different stages of corrected accurate data. In the course of the experiment, deep reinforcement learning was introduced to solve the hyperparameter tuning problem of the staging model. Through experiments conducted on 24,872 electronic medical records, the accuracy rate of gout diagnosis was found to be 90.03%, while the accuracy rate for diagnosing different stages of gout disease reached 86.85%. These findings serve as a valuable tool in assisting clinicians with accurate staging and diagnosis of gout. The application of deep reinforcement learning in gout staging diagnosis demonstrates a significant enhancement in diagnostic accuracy, thereby validating the effectiveness and feasibility of this method.
引用
收藏
页数:18
相关论文
共 14 条
  • [1] Regularized Least Squares Twin SVM for Multiclass Classification
    Ali, Javed
    Aldhaifallah, M.
    Nisar, Kottakkaran Sooppy
    Aljabr, A. A.
    Tanveer, M.
    [J]. BIG DATA RESEARCH, 2022, 27
  • [2] Reinforcement Learning, Fast and Slow
    Botvinick, Matthew
    Ritter, Sam
    Wang, Jane X.
    Kurth-Nelson, Zeb
    Blundell, Charles
    Hassabis, Demis
    [J]. TRENDS IN COGNITIVE SCIENCES, 2019, 23 (05) : 408 - 422
  • [3] Q-Learning: Theory and Applications
    Clifton, Jesse
    Laber, Eric
    [J]. ANNUAL REVIEW OF STATISTICS AND ITS APPLICATION, VOL 7, 2020, 2020, 7 : 279 - 301
  • [4] Towards intelligent robust detection of anatomical structures in incomplete volumetric data
    Ghesu, Florin C.
    Georgescu, Bogdan
    Grbic, Sasa
    Maier, Andreas
    Hornegger, Joachim
    Comaniciu, Dorin
    [J]. MEDICAL IMAGE ANALYSIS, 2018, 48 : 203 - 213
  • [5] Effective multiple cancer disease diagnosis frameworks for improved healthcare using machine learning
    Hsu, Ching-Hsien
    Chen, Xing
    Lin, Weiwei
    Jiang, Chuntao
    Zhang, Youhong
    Hao, Zhifeng
    Chung, Yeh-Ching
    [J]. MEASUREMENT, 2021, 175
  • [6] The Artificial Intelligence Clinician learns optimal treatment strategies for sepsis in intensive care
    Komorowski, Matthieu
    Celi, Leoa
    Badawi, Omar
    Gordon, Anthony C.
    Faisal, A. Aldo
    [J]. NATURE MEDICINE, 2018, 24 (11) : 1716 - +
  • [7] Xrare: a machine learning method jointly modeling phenotypes and genetic evidence for rare disease diagnosis
    Li, Qigang
    Zhao, Keyan
    Bustamante, Carlos D.
    Ma, Xin
    Wong, Wing H.
    [J]. GENETICS IN MEDICINE, 2019, 21 (09) : 2126 - 2134
  • [8] [李占山 Li Zhanshan], 2019, [通信学报, Journal on Communications], V40
  • [9] Liu SQ, 2019, Arxiv, DOI arXiv:1907.09475
  • [10] Maxwell J. Clerk, 2019, DIS PREDICTION MODEL, V42, P2339