消化管疾患におけるAI診断開発

Artificial Intelligence-assisted Diagnostic Systems for Gastrointestinal Diseases

要旨●人口の高齢化を背景とした癌の罹患数・死亡数の増加や，医療費の増大，慢性的な医師不足などが問題視されている．これに対し，オリンパスは内視鏡の開発を通じて，早期診断，低侵襲治療という価値を提供し続け，近年はAI技術の活用も検討している．病変の発見や鑑別を支援するCADはその一つである．筆者らは内視鏡CADの課題として，“内視鏡操作に違和感を与えない1モニタ仕様”，“さまざまな撮影環境下での性能確保”，“ユーザビリティを考慮したシステム連携”を重視し，これを考慮した大腸内視鏡CADを開発した．また，今後の多臓器展開に向け，自社AI技術開発に加え，AI技術に強みを持つ研究機関・企業との研究開発や，プラットフォームの構築も進めている．

　As society ages, pressing issues worldwide include the increased number of cancer cases and cancer-related deaths, rising medical costs, and chronic shortages of healthcare providers. To solve these issues, Olympus has provided customer value through its development of medical devices such as endoscopes that contribute to early diagnosis and minimally-invasive treatment. Recently, Olympus has been developing endoscopic systems for the application of AI（artificial intelligence）technology. CAD（Computer-aided detection/diagnosis）is an example of an application of AI technology. Three challenges of CAD for endoscopy have been our focus：（1）a single-monitor display of CAD results prevents a feeling of discomfort during endoscopy ; （2）stable high-level performance under various conditions during endoscopy ; and（3）seamless communication between the endoscopic system and the CAD device ensures excellent usability. Olympus has successfully developed a CAD for colonoscopy to address all these challenges. Aiming for the continued expansion of AI technology, Olympus is conducting both in-house and joint development of CAD for other organs with institutions and companies specialized in AI technology. These partnerships will result in the development of the Olympus platform for multiple CAD software.