Japanese
English
- 有料閲覧
- Abstract 文献概要
- 1ページ目 Look Inside
- 参考文献 Reference
◆要旨:医療法施行規則では,術後速やかに手術記録を作成することが義務付けられている.現状,手術記録の作成には,外科医の教育要素が含まれるため,多くの時間と労力が費やされることがある.一方,近年の内視鏡外科手術では術野映像を保存することが一般的となった.深層学習を用いた最新の画像認識技術を用いれば,内視鏡映像上の手術器具,解剖構造などを精度よく認識することができる.つまり,術中の客観的事実をデータとして抽出することが技術的に可能となったといえる.本研究は,腹腔鏡下胆囊摘出術を対象術式として,その手術工程をリスト化,深層学習による画像認識情報との関連付けを通して,手術記録の自動化に関する基礎検討を行った.
A surgical report must be promptly prepared after surgery, as required by the Enforcement Regulations on the Medical Care Act. Preparing a surgical record involves educational components for surgeons, thus demanding considerable amount of time and effort. Recently, a high-resolusion video recording of the operation field during endoscopic surgery has become popular. Furthermore, recent image recognition technology based on deep learning can accurately identify operation phases, instruments, and anatomical structures in endoscopic images. This means that objective facts from the surgery can be technically obtained as digial data. In this study, we conducted a pilot study aimed at automating the surgical record process. We achieved this by creating a comprehensive list of surgical procedures for laparoscopic cholecystectomy and then correlating this list with the recording information provided by deep learning models.
Copyright © 2024, JAPAN SOCIETY FOR ENDOSCOPIC SURGERY All rights reserved.