Development of a simulation model to predict the therapeutic effect of stereotactic radiosurgery using radiomics analysis

2020 Fiscal Year Final Research Report

• Project/Area Number: 19K24042
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0906:Surgery related to the biological and sensory functions and related fields
• Research Institution: The University of Tokyo
• Principal Investigator: SHINYA YUKI 東京大学, 医学部附属病院, 助教 (20844616)
• Project Period (FY): 2019-08-30 – 2021-03-31
• Keywords: 脳腫瘍 / radiomics / 機械学習 / 定位放射線治療 / ガンマナイフ

Outline of Final Research Achievements

In this study, we developed a simulation model for predicting the response to stereotactic radiosurgery (SRS) using radiomics. We collected a total of 1000 cases of metastatic brain tumors and 500 cases of meningiomas from 3800 cases treated with SRS at the Department of Neurosurgery, the University of Tokyo. Then, we analyzed the treatment imaging data, dose planning data, and clinical data before and after treatment. Also, we input these data and output treatment response as a binary variable between controlled or uncontrolled. We identified a total of 695 imaging features and created a database of these imaging features and clinical data. Using these databases, we developed a prediction model for SRS treatment response.

Free Research Field

脳腫瘍

Academic Significance and Societal Importance of the Research Achievements

定位放射線治療は低侵襲的かつ汎用性の高い脳腫瘍治療である。一方、治療効果を予測する高精度modelは確立していない。治療奏効率は概して70-90%と高いが、同一組織型腫瘍であっても治療反応性は異なり得る。ガンマナイフ後の再発例には手術が選択されるが放射線性変化として腫瘍硬化や癒着など手術成績悪化に繋がる変化が起きる。治療前にガンマナイフへの反応性を予測することが出来れば回避可能な未来である。治療反応性予測モデルの構築は『大規模データに裏打ちされた客観的かつ頑健なdecision making』を可能にし、脳腫瘍医療におけるprecision medicine実現に向けた大きな一歩となる。