2020 Fiscal Year Final Research Report
CFD analysis on MR enhanced wall for investigation of mechanism of aneurysm instabiity
Project/Area Number |
18K18355
|
Research Category |
Grant-in-Aid for Early-Career Scientists
|
Allocation Type | Multi-year Fund |
Review Section |
Basic Section 90110:Biomedical engineering-related
|
Research Institution | Tohoku University |
Principal Investigator |
ANZAI HITOMI 東北大学, 流体科学研究所, 助教 (50736981)
|
Project Period (FY) |
2018-04-01 – 2021-03-31
|
Keywords | 脳動脈瘤 / MR血管造影 / ディープラーニング |
Outline of Final Research Achievements |
This study aims to correlate the wall enhancement in MR vessel wall imaging and hemodynamic stress in small-middle cerebral aneurysms. In the case of 10-years follow-up, the extension of aneurysm wall occurs near the breb where the wall enhancement observed. For further study to establish the threshold of hemodynamic parameters, analysis for big data will be required. Therefore, deep learning technique was applied for predicting the flow field inside blood vessels. This network can predict the flow from the geometry directly and achieved the speedup of analysis 600 times than conventional CFD.
|
Free Research Field |
数値流体力学解析
|
Academic Significance and Societal Importance of the Research Achievements |
脳動脈瘤は人口の3-5%に発生し、1%/年で破裂の可能性を持つ。しかし破裂メカニズムは不明なことも多く、未だ破裂の予測は難しい。本研究では破裂の前段階に存在すると考えられている動脈瘤壁の炎症と血流分布を照らし合わせ、血流ストレスが壁面の縮退に与える影響を調べることで、動脈瘤の成長メカニズムの解明に貢献した。さらに、疫学的エビデンスを得るためには大規模な患者群に対する解析が不可欠であることから、深層学習ネットワークの構築により血流解析を即時化し、大規模な患者群への血流解析を可能とした。
|