Development of endoscopic treatment assistance system based on relationship modeling between recognition and action

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K12723
• 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: Nagoya University
• Principal Investigator: Oda Masahiro 名古屋大学, 情報基盤センター, 准教授 (30554810)
• Co-Investigator(Kenkyū-buntansha): 古川 和宏 名古屋大学, 医学部附属病院, 病院講師 (70624310)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 認知と行動の関係解析 / 内視鏡支援システム

Outline of Final Research Achievements

The purpose of this study is to realize an endoscopic operation support system based on the analysis of the relationship between cognition and behavior. To achieve this, we first developed a method to measure the cognition and behavior of physicians during endoscopic operations. Subsequently, we developed a cognition-behavior linkage model that can estimate behavior from cognitive information. This model was developed using a data-driven approach with deep learning. Additionally, we developed and evaluated a system to provide support information to physicians. With the cooperation of gastroenterologists, we conducted measurements and evaluations using the system, which demonstrated the ability to present useful treatment support information. On the other hand, it became clear that further adaptation to individual differences in endoscopic operation by physicians is necessary.

Free Research Field

医用画像処理

Academic Significance and Societal Importance of the Research Achievements

従来の治療支援システムは、画像情報などの「認知」情報のみを使って支援情報生成を行うものが主流であった。本研究はそれだけでなく、治療操作つまり「行動」情報までも扱い、認知と行動の関連性をモデル化し治療支援システムに利用する。本研究を行う中で、治療操作を計測しデータ化する方法、形状データを解析する深層学習モデル構築を行い、その成果が国際会議でのOutstanding Paper Award受賞という形で認められた。本研究を行う中で、医師が行う治療操作には医師の個人差が強く影響し、個人差を考慮した方法の開発が必要との知見が得られた。本研究の成果は、今後の治療支援システムの高性能化に大きく貢献する。