Extraction Key Factors of Medical Process by Text Mining

2017 Fiscal Year Final Research Report

• Project/Area Number: 15H02778
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Life / Health / Medical informatics
• Research Institution: Kyushu University
• Principal Investigator: Hirokawa Sachio 九州大学, 情報基盤研究開発センター, 教授 (40126785)
• Co-Investigator(Kenkyū-buntansha): 中藤 哲也 九州大学, 情報基盤研究開発センター, 助教 (20253502) ; 中島 直樹 九州大学, 大学病院, 教授 (60325529)
• Research Collaborator: SOEJIMA Hidehisa 済世会熊本病院, 院長 ; YAMASHITA Takanori 九州大学, 大学病院, 助教
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: クリニカルパス / 機械学習 / テキストマイニング / Support Vector Machine / 属性選択 / 可視化

Outline of Final Research Achievements

We focused on systematic clinical records (clinical paths) of clinical practice and patient condition, aiming at discovering problem factors and applied machine learning and visualization. For example, machine learning was applied by taking records of long-term inpatients as positive cases and other records as negative cases, and feature words and examination values indicating important factors were extracted, and the discrimination performance was evaluated. In the visualization, we realized an interactive factor time series evolution graph that facilitates interpretation and a system that supports creation of template of medical record.

Free Research Field

情報科学