| Project/Area Number |
18K12049
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 90110:Biomedical engineering-related
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology |
|---|
Principal Investigator |
Kosaka Ryo 国立研究開発法人産業技術総合研究所, 生命工学領域, 主任研究員 (10415680)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Project Status |
Completed (Fiscal Year 2020)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2020: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2019: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
|---|
| Keywords | 血液ポンプ / 人工心臓 / 人工知能 / ニューラルネットワーク / 動圧軸受 / 流体軸受 / 最適化技術 / 軸受剛性 / AI / 実験計画法 / 多目的最適化手法 / 革新的実験計画法 / 機械学習 |
|---|
| Outline of Final Research Achievements |
A hydrodynamically levitated centrifugal blood pump has been developed. In this study, design of the hydrodynamic thrust bearing for the blood pump was optimized using artificial intelligence (AI) to improve bearing performance and hemocompatibility. Input parameters for neural network (NN) were number of grooves, groove angle, inner and outer groove depths. Output parameters were bearing levitation force and damage index (DI) of red blood cell calculated by computational fluid dynamics analysis. By these parameters, a NN was constructed. Then, bearing levitation force and DI for 450 models were calculated, and three optimal candidate models were selected. In the validation tests, one of the optimal candidate models had better levitation performance and hemolysis performance compared to other models including the conventional models. In conclusion, a hydrodynamic bearing could be optimized using AI to improve both bearing performance and hemocompatibility.
|
| Academic Significance and Societal Importance of the Research Achievements |
本課題では、人工知能（AI）を用いた革新的実験計画法により、長期耐久性と血液適合性に優れた体外循環用動圧浮上遠心血液ポンプの動圧軸受の最適設計を実施した。本手法は、多入力多目的最適化手法と機械学習を統合した最適化手法である。本最適化手法は、従来は試行錯誤で最適化していた複雑システムの最適解を簡易に探索可能である。本手法は、血液ポンプだけでなく、他の医療機器や産業機器の最適設計に応用可能であるため、学術的意義や社会的意義は高い。
|
