Radiomics and Radiogenomics Analysis Using Deep Learning in Ovarian Cancer

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K09466
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 56040:Obstetrics and gynecology-related
• Research Institution: The University of Tokyo
• Principal Investigator: Miyamoto Yuichiro 東京大学, 医学部附属病院, 講師 (70634955)
• Co-Investigator(Kenkyū-buntansha): 曾根 献文 東京大学, 医学部附属病院, 准教授 (90598872)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 卵巣癌 / 子宮肉腫 / 深層学習

Outline of Final Research Achievements

Since ovarian cancer has a large number of cases, we first aim to establish an AI-based diagnostic system from MRI images of uterine sarcoma, and then apply the development flow to the ovarian cancer diagnostic model. We developed an automatic diagnosis system by retrospectively entering 63 uterine sarcoma cases and 200 uterine myoma cases. The development of an ovarian cancer diagnostic model was also conducted in parallel with the development of the system. The correct diagnosis rate of the uterine sarcoma MRI imaging model was comparable to that of radiologists. In the AI-assisted diagnosis, we were able to raise the diagnostic level of radiology residents to that of specialists. Preliminary experiments on the ovarian cancer diagnosis model showed relatively satisfactory results, but the number of cases is still small.

Free Research Field

婦人科腫瘍

Academic Significance and Societal Importance of the Research Achievements

子宮肉腫と子宮筋腫を鑑別する深層学習モデルを開発した。このモデルの臨床応用を目指すことにより子宮肉腫の正確な診断、最適な治療方針を提供できる事になる。この開発フローを卵巣癌診断モデルの開発に応用する事ができる。