validation analysis of usefulness for the diagnosis of pancreatic diseases using artificial intelligence

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K15938
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 53010:Gastroenterology-related
• Research Institution: Aichi Cancer Center Research Institute
• Principal Investigator: Takamichi Kuwahara 愛知県がんセンター(研究所), システム解析学分野, 研究員 (10816408)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 消化器内科 / 人工知能 / deep learning

Outline of Final Research Achievements

1)Multicenter Retrospective Study on the Diagnosis of Benign and Malignant IPMN:
We collected EUS images from approximately 600 IPMN cases and developed an AI using a supercomputer to differentiate between benign and malignant IPMN. The AI's accuracy was determined to be 87%. A report is currently being drafted on this topic.2) Multicenter Retrospective Study on the Diagnosis of Pancreatic Tumors 3) Multicenter Prospective Study on the Differential Diagnosis of Pancreatic Diseases: Research plans for 2) and 3) were formulated, and discussions with companies regarding commercialization were conducted. However, due to the low prevalence of pancreatic diseases, the market was deemed too narrow to recoup development costs, leading to the abandonment of commercialization efforts. Consequently, the data collection methods needed revision, and both studies 2) and 3) were canceled.

Free Research Field

消化器内科

Academic Significance and Societal Importance of the Research Achievements

超音波内視鏡（EUS）画像をAIによって解析することによって、膵疾患診断（膵腫瘍、膵嚢胞）において既存の画像診断技術より高度に診断することが可能であったことが示された。この技術を内視鏡診断機器に搭載することで、今まで診断困難であった病変を場所を選ばず診断できる可能性が示唆された。膵疾患の診断は医師によりかなり差異があるため、医療の質の均てん化できる可能性がある。しかしながら、現状市場規模が小さく製品化への道筋が立っていない状況である。今後製品化へのコストを下げる研究を新たに立ち上げ社会実装を試みる予定である。