Project/Area Number |
19K12840
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Review Section |
Basic Section 90130:Medical systems-related
|
Research Institution | Niigata University |
Principal Investigator |
Kodama Satoru 新潟大学, 医歯学総合研究科, 特任准教授 (50638781)
|
Co-Investigator(Kenkyū-buntansha) |
加藤 公則 新潟大学, 医歯学総合研究科, 特任教授 (00303165)
藤原 和哉 新潟大学, 医歯学総合研究科, 特任准教授 (10779341)
渡邊 賢一 新潟大学, 医歯学総合研究科, 客員研究員 (70175090)
|
Project Period (FY) |
2019-04-01 – 2022-03-31
|
Project Status |
Completed (Fiscal Year 2021)
|
Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2021: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2020: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2019: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
Keywords | 人工知能 / 糖尿病 / メタ解析 / 機械学習 / 生活習慣病 / エビデンスの基づく医療 / 自然言語処理 |
Outline of Research at the Start |
1) MA対象研究デザイン論文検出、2) 論文テーマ認識、3) テーマ設定および該当論文抽出の3つのAIを開発するため、文献データベース(EMBASE)でヒットした最新のabstractと索引語(EMTREE)がついた論文について、申請者がこれまでの豊富なSR/MA経験で培った思考プロセスに即し、論文タイトルの各用語と主要テーマを関連付けて機械学習させ、目的に沿った論文を検出する能力を評価する。
|
Outline of Final Research Achievements |
Evidence for usefulness of artificial intelligence (AI) in primary prevention of non-communicable diseases has not been established. This project aimed to assess the ability of AI to predict the onset of non-communicable diseases, focusing on type 2 diabetes mellitus (T2D) and hypoglycemia, a major barrier of treating diabetes, using a meta-analytic technique. The results of meta-analysis were interpret as meaning that the ability of current machine learning was acceptable for clinicians to discriminate individuals at high risk of T2D but insufficient for individuals to recognize their risk of T2D and that it is sufficient as a tool for patients with diabetes to prepare for their impeding hypoglycemia. The study is the first step to apply the AIs to clinical practice of non-communicable diseases, especially identifying individuals which were at high risk and thus require strict managements for primary prevention.
|
Academic Significance and Societal Importance of the Research Achievements |
予後予測に必須であるが、原理・解釈が難しく敬遠されがちなhierarchical summary receiver operating characteristicモデルを用いたメタ解析を大々的に行った研究プロジェクトである。人工知能の糖尿病、低血糖予測能力を評価した本研究は、社会的要請の高いAIの糖尿病診療にとって極めて重要な布石であり、今後、他の生活習慣病への拡張も期待大である。
|