• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Validation and proposal of a framework for the application of artificial intelligence techniques to epidemiological data

Research Project

Project/Area Number 19K19433
Research Category

Grant-in-Aid for Early-Career Scientists

Allocation TypeMulti-year Fund
Review Section Basic Section 58030:Hygiene and public health-related: excluding laboratory approach
Research InstitutionUniversity of Yamanashi

Principal Investigator

Ooka Tadao  山梨大学, 大学院総合研究部, 助教 (40803987)

Project Period (FY) 2019-04-01 – 2022-03-31
Project Status Completed (Fiscal Year 2021)
Budget Amount *help
¥3,510,000 (Direct Cost: ¥2,700,000、Indirect Cost: ¥810,000)
Fiscal Year 2020: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2019: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Keywords疾患予測モデル / 糖尿病 / 健康診断 / 機械学習 / ランダムフォレスト
Outline of Research at the Start

健康診断データに人工知能技術(RandomForest,DeepLearning)を適用する事で糖尿病の発症予測と予測因子の同定を行い、既存の疫学的アプローチ(線形モデルによる)により示される事実との一致性を確認する事で解析の妥当性を検証する。加えて、人工知能技術による解析の優位性や欠点を提示する事で、疫学データに人工知能技術を導入する是非を検討し、疫学データへの人工知能技術適用の枠組みを提案する。解析データには、複数の健診施設における13年分(2005-2017:延べ約15万人)の検査結果を使用し、HbA1cの上昇をモデルのアウトカムに設定する。

Outline of Final Research Achievements

By utilizing data from about 20 years of health checkups at health checkup facilities, we have succeeded in developing an artificial intelligence model that can accurately predict who will have a sharp rise in HbA1c, an important indicator of type 2 diabetes, based on the results of the previous year's health checkups. By validating these models, we identified factors (e.g., cholesterol levels, blood pressure) that are important in predicting the onset of type 2 diabetes.
Furthermore, by developing this model, we have also developed an artificial intelligence model that can accurately predict the results of health checkups one and three years from now, based on the results of past health checkups.
In the future, a randomized controlled trial will be conducted to confirm whether the model can be used in actual health checkups and health guidance to promote the health of examinees.

Academic Significance and Societal Importance of the Research Achievements

様々な機械学習モデルを疫学データに活用する事で、疫学データへの人工知能(機械学習)技術適応の枠組みの検証を行うことが出来た。また、将来の健康診断結果を高精度に予測する機械学習モデルの開発にも成功した。今後は、開発した予測モデルをどのように使うか、研究の枠組みをどのように活用していくかを検討するために、ランダム化比較試験を含めた更なる検討を進めていく。

Report

(4 results)
  • 2021 Annual Research Report   Final Research Report ( PDF )
  • 2020 Research-status Report
  • 2019 Research-status Report
  • Research Products

    (9 results)

All 2021 2020 2019

All Journal Article (2 results) (of which Peer Reviewed: 2 results,  Open Access: 2 results) Presentation (6 results) (of which Int'l Joint Research: 1 results) Book (1 results)

  • [Journal Article] 425Artificial Intelligence Approaches to Type 2 Diabetes Risk Prediction and Exploration of Predictive Factors2021

    • Author(s)
      Ooka Tadao、Yokomichi Hiroshi、Yamagata Zentaro
    • Journal Title

      International Journal of Epidemiology

      Volume: 50 Issue: Supplement_1 Pages: 106989-106989

    • DOI

      10.1093/ije/dyab168.515

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Random forest approach for determining risk prediction and predictive factors of type 2 diabetes: large-scale health check-up data in Japan2021

    • Author(s)
      Tadao Ooka, Hisashi Johno, Kazunori Nakamoto, Yoshioki Yoda, Hiroshi Yokomichi, Zentaro Yamagata
    • Journal Title

      BMJ Nutrition, Prevention & Health

      Volume: -

    • Related Report
      2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Presentation] Artificial Intelligence Approaches to Type 2 Diabetes Risk Prediction and Exploration of Predictive Factors2021

    • Author(s)
      Tadao Ooka
    • Organizer
      International Journal of Epidemiology
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 機械学習を活用した将来の健康診断検査値の予測方法の検討2021

    • Author(s)
      大岡忠生、横道洋司、山縣然太朗
    • Organizer
      第31回日本疫学会学術総会
    • Related Report
      2020 Research-status Report
  • [Presentation] Deep Learning を活用して健康診断結果から糖尿病発症を予測する方法の検討2020

    • Author(s)
      大岡忠生、横道洋司、山縣然太朗
    • Organizer
      第79回日本公衆衛生学会総会
    • Related Report
      2020 Research-status Report
  • [Presentation] 機械学習技術を用いて健康診断結果から糖尿病発症を予測する方法の検討2020

    • Author(s)
      大岡忠生、横道洋司、山縣然太朗
    • Organizer
      第30回日本疫学会学術総会
    • Related Report
      2019 Research-status Report
  • [Presentation] 予防医療分野における疫学データへの機械学習技術活用について ~スパースモデリングを活用した糖尿病発症予測と予測因子探索~2020

    • Author(s)
      大岡忠生、日野英逸、横道洋司、山縣然太朗
    • Organizer
      統計数理研究所共同利用研究集会 ~統計的機械学習の新展開~
    • Related Report
      2019 Research-status Report
  • [Presentation] 人工知能技術を活用した2型糖尿病のリスク予測手法の検証と疾患予測因子の探索2019

    • Author(s)
      大岡忠生、横道洋司、山縣然太朗
    • Organizer
      第78回日本公衆衛生学会総会
    • Related Report
      2019 Research-status Report
  • [Book] 公衆衛生2021

    • Author(s)
      大岡忠生
    • Total Pages
      6
    • Publisher
      医学書院
    • Related Report
      2020 Research-status Report

URL: 

Published: 2019-04-18   Modified: 2023-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi