Development of radiotherapy treatment planning system using deep learning

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K08093
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 52040:Radiological sciences-related
• Research Institution: Kindai University
• Principal Investigator: Doi Hiroshi 近畿大学, 医学部, 講師 (50529047)
• Co-Investigator(Kenkyū-buntansha): 門前 一 近畿大学, 医学部, 准教授 (10611593)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 知識ベース治療計画

Outline of Final Research Achievements

Based on 561 intensity-modulated radiation therapy plans used in clinical practice across multiple facilities, a machine learning model was developed to estimate the doses of radiation that each structure within a patient's body would receive, and plans were created automatically without any human intervention. The plan quality was found to be equivalent to those created using traditional methods. Furthermore, the use of this model allowed for standardization of treatment plan quality by demonstrating consistency in the doses of radiation delivered to the rectum and bladder, independent of the treatment facility.

Free Research Field

放射線治療

Academic Significance and Societal Importance of the Research Achievements

強度変調放射線治療を用いることによって病変部周囲の放射線量をより低減できるようになるため放射線治療の適応症例が拡大されている一方で、治療計画作成には計画者の知識や経験に依存するところが大きい。本研究によって人が介在しない治療計画作成が可能になれば医療の均てん化につながる。また臨床試験に用いた場合には一定水準の放射線治療を実施された患者の臨床成績を基に新しい治療法の有効性を比較できる。