Development of statistical models for data containing heterogeneous subgroups using statistical learning theory and its application to clinical medicine

• Project/Area Number: 18K11197
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 60030:Statistical science-related
• Research Institution: Keio University
• Principal Investigator: Hayashi Kenichi 慶應義塾大学, 理工学部(矢上), 准教授 (70617274)
• Project Period (FY): 2018-04-01 – 2024-03-31
• Project Status: Completed (Fiscal Year 2023)
• Budget Amount: ¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000); Fiscal Year 2021: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000); Fiscal Year 2020: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000); Fiscal Year 2019: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000); Fiscal Year 2018: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
• Keywords: 医学統計学 / IDI / ROC / 欠測値 / 因果推論 / 二値回帰モデル / 生存時間解析 / 経時データ / 生存時間モデル / 再帰分割法 / 回帰モデル / 平均治療効果 / 平均因果効果 / 共変量調整 / odds-IDI / power-IDI / 集団の異質性 / 解釈可能性 / 混合治癒モデル / 多値応答 / ノンコンプライアンス / pAUC / 反応例解析 / 欠測データ / 生物統計 / 機械学習 / 統計科学

Outline of Final Research Achievements

Our research aims to contribute to the development of statistical methods with both predictive power and interpretability for data consisting of heterogeneous subpopulations. The main results are (1) the development of an improved version of the IDI, the odds-IDI, (2) the proposal of statistical analysis methods that take into account various situations in clinical trials, and (3) the proposal of a survival time regression model that takes heterogeneity into account. In (1), the proposed index outperforms existing indices in terms of performance, and the theoretical aspects and interpretability of the proposed index were studied. In (2), statistical analysis methods were studied for the cases where heterogeneity is characterized by missing data and so forth. In (3), a regression model for survival outcomes in the case of a mixture of potentially cured and uncured groups was studied.

Academic Significance and Societal Importance of the Research Achievements

本研究の成果の学術的意義は，従来の統計手法では十分に対応できなかった異質性を含むデータに対する新たな視点からの解析法を提示したことである．予測力と解釈可能性を兼ね備えた新たな統計的手法の開発は，大量・複雑になるデータの特徴を人間が理解する上で重要な意義をもつ．この課題に対し，疫学や臨床試験を想定する諸種のデータについて予測精度のより高い統計モデルを開発し，またそれらのモデルの評価指標を提案した．これらの成果は，様々な分野におけるデータ分析の精度向上と新たな知見の創出に貢献し，社会全体の利益に資する可能性をもつと考えられる．

Research Products

• [Journal Article] 疫学分野での計量生物学の発展 (2024)
  • Author(s): 田栗正隆、高橋邦彦、小向翔、伊藤ゆり、服部聡、船渡川伊久子、篠崎智大、山本倫生、林賢一
  • Journal Title: 計量生物学 Volume: 44 Pages: 129-200
  • Peer Reviewed
• [Journal Article] Micro-Internal Short Circuit Detection in Lithium-Ion Batteries Based on <i>k</i>-Nearest Neighbor Method (2023)
  • Author(s): 志村 重輔、林 沙織、岡安 悟志、板垣 昌幸、林 賢一
  • Journal Title: Bulletin of Data Analysis of Japanese Classification Society Volume: 12 Issue: 1 Pages: 1-15
  • DOI: 10.32146/bdajcs.12.1
  • ISSN: 2186-4195, 2434-3382
  • Year and Date: 2023-09-01
  • Peer Reviewed / Open Access
• [Journal Article] Model Selection with Missing Data Embedded in Missing-at-Random Data (2023)
  • Author(s): Takai Keiji、Hayashi Kenichi
  • Journal Title: Stats Volume: 6 Issue: 2 Pages: 495-505
  • DOI: 10.3390/stats6020031
  • Peer Reviewed / Open Access
• [Journal Article] An Accelerated Failure Time Cure Model with Shifted Gamma Frailty and Its Application to Epidemiological Research (2022)
  • Author(s): Aida Haro、Hayashi Kenichi、Takeuchi Ayano、Sugiyama Daisuke、Okamura Tomonori
  • Journal Title: Healthcare Volume: 10 Issue: 8 Pages: 1383-1383
  • DOI: 10.3390/healthcare10081383
  • Peer Reviewed / Open Access
• [Journal Article] A nationwide randomised, double-blind, placebo-controlled physicians’ trial of loxoprofen for the treatment of fatigue, headache, and nausea after hangovers (2020)
  • Author(s): Hara, M., Hayashi,K., Kitamura, T., Honda, M., Tamaki, M.
  • Journal Title: Alcohol Volume: 84 Pages: 21-25
  • DOI: 10.1016/j.alcohol.2019.10.006
  • Peer Reviewed / Open Access
• [Journal Article] Pulmonary vein isolation alone vs. more extensive ablation with defragmentation and linear ablation of persistent atrial fibrillation: the EARNEST-PVI trial (2020)
  • Author(s): Inoue, K., Hikoso,S., Masuda, M., Furukawa, Y., Hirata, A., Egami, Y., Watanabe, T., Minamiguchi, H., Miyoshi, M., Tanaka, N., Oka, T., Okada, M., Kanda, T., Matsuda, Y., Kawasaki, M., Hayashi, K., Kitamura, T., Dohi, T., Sunaga, A., Mizuno, H., Nakatani, D., Sakata, Y.
  • Journal Title: EP Europace Volume: 23 Issue: 4 Pages: 565-574
  • DOI: 10.1093/europace/euaa293
  • Peer Reviewed
• [Journal Article] The power‐integrated discriminant improvement: an accurate measure of the incremental predictive value of additional biomarkers (2019)
  • Author(s): Hayashi, K., Eguchi, S.
  • Journal Title: Statistics in Medicine Volume: 印刷中 Issue: 14 Pages: 2589-2604
  • DOI: 10.1002/sim.8135
  • Peer Reviewed / Int'l Joint Research
• [Journal Article] Effect of extensive ablation on recurrence in patients with persistent ptrial fibrillation treated with pulmonary cein isolation (EARNEST-PVI) trial: design and rationale (2019)
  • Author(s): Dohi, T., Nakatani, D., Inoue, K., Hikoso, S. Oka, T. Hayashi, K., Masuda, M., Furukawa, Y., Kawasaki, M., Egami,Y., Kashiwase, K., Hirata, A., Watanabe, T., … Sakata,Y.
  • Journal Title: Journal of Cardiology Volume: 74 Issue: 2 Pages: 64-68
  • DOI: 10.1016/j.jjcc.2019.01.010
  • Peer Reviewed
• [Journal Article] Asymptotic comparison of semi-supervised and supervised linear discriminant functions for heteroscedastic normal populations (2018)
  • Author(s): Kenichi Hayashi
  • Journal Title: Advances in Data Analysis and Classification Volume: - Issue: 2 Pages: 315-339
  • DOI: 10.1007/s11634-016-0266-6
  • Peer Reviewed
• [Journal Article] 統計学は錬金術ではない (2018)
  • Author(s): 林賢一
  • Journal Title: 心理学評論 Volume: 61 Pages: 147-155
  • NAID: 130007678122
• [Journal Article] Estimation of a concordance probability for doubly censored time-to-event data (2018)
  • Author(s): Kenichi Hayashi, Yasutaka Shimizu
  • Journal Title: Statistics in Biosciences Volume: - Issue: 3 Pages: 546-567
  • DOI: 10.1007/s12561-018-9216-5
  • Peer Reviewed
• [Presentation] Asymptotic properties of Matthews correlation coefficient (2023)
  • Author(s): Yuki Itaya, Jun Tamura, Kenichi Hayashi, Kouji Yamamoto
  • Organizer: The 8th Japanese-German Symposium on Classification
  • Int'l Joint Research
• [Presentation] 生存ー経時同時モデルに対する異質な部分集団への再帰分割法 (2023)
  • Author(s): 名取京太郎，林賢一
  • Organizer: 2023年度日本分類学会大会
• [Presentation] 無作為化比較試験における半教師あり学習 を用いたリスク差の推定 (2022)
  • Author(s): 林賢一，吉牟田迪弥
  • Organizer: 日本計量生物学会年会
• [Presentation] k近傍法を用いたリチウムイオン電池の微小内部短絡検出 (2022)
  • Author(s): 志村 重輔，林 沙織，岡安 悟志，板垣 昌幸，林 賢一
  • Organizer: 日本分類学会
• [Presentation] Odds-based predictive improvement index for binary regression models (2022)
  • Author(s): Kenichi Hayashi, Shinto Eguchi
  • Organizer: The 11th Conference of the IASC-ARS The Asian Regional Section of the International Association for Statistical Computing
  • Int'l Joint Research / Invited
• [Presentation] A new integrated discriminant improvement index via odds (2021)
  • Author(s): 林賢一，江口真透
  • Organizer: 日本計量生物学会年会
• [Presentation] 二値回帰モデルに対するオッズに基づいた予測改善指標 (2021)
  • Author(s): 林賢一，江口真透
  • Organizer: 日本計算機統計学会 第35回シンポジウム
• [Presentation] シフトガンマフレイルティを伴う加速時間治癒モデルとその疫学研究への応用 (2020)
  • Author(s): 会田晴郎，林賢一
  • Organizer: 統計関連学会連合大会
• [Presentation] Estimation of causal effects in the presence of noncom- pliance without actual treatment information (2019)
  • Author(s): Hayashi, K., Taguri, M.
  • Organizer: The 40th Annual Conference of the International Society for Clinical Biostatistics
  • Int'l Joint Research
• [Presentation] A model selection criterion for missing data with MAR mechanism but NMAR for its subsets (2019)
  • Author(s): Hayashi, K., Takai, K.
  • Organizer: International conference Data Science, Statistics & Visualisation (A Satellite Conference of the 62nd World Statistics Congress promoted by IASC)
  • Int'l Joint Research
• [Presentation] A variable selection criterion for competing risk data with pseudo-observations (2019)
  • Author(s): Tajima, F., Hayashi, K.
  • Organizer: The 6th International Society for Biopharmaceutical Statistics
  • Int'l Joint Research
• [Presentation] A new composite estimand for regulatory clinical trials with dropouts (2018)
  • Author(s): Taguri, M., Hayashi, K.
  • Organizer: The 29th International Biometric Conference
• [Book] Rで学ぶ統計的データ解析 (2020)
  • Author(s): 林賢一
  • Total Pages: 352
  • Publisher: 講談社サイエンティフィク
  • ISBN: 9784065186190