2023 Fiscal Year Final Research Report
Development of a novel algorithm for drug dose settings using machine learning method
Project/Area Number |
20K16035
|
Research Category |
Grant-in-Aid for Early-Career Scientists
|
Allocation Type | Multi-year Fund |
Review Section |
Basic Section 47060:Clinical pharmacy-related
|
Research Institution | Keio University (2022-2023) Hokkaido University (2020-2021) |
Principal Investigator |
Imai Shungo 慶應義塾大学, 薬学部(芝共立), 講師 (40845070)
|
Project Period (FY) |
2020-04-01 – 2024-03-31
|
Keywords | バンコマイシン / 機械学習 / 治療薬物モニタリング / 薬剤投与設計 |
Outline of Final Research Achievements |
In this study, we constructed an initial dose setting algorithm for vancomycin using the Decision Tree model, which is a typical machine learning method, and validated the its usefulness. We obtained clinical information on 822 patients from two medical facilities and constructed a model using the Classification And Regression Tree (CART) algorithm. The accuracy of the developed model was better than the conventional dose setting algorithm, suggesting the usefulness of this approach.
|
Free Research Field |
医療薬学
|
Academic Significance and Societal Importance of the Research Achievements |
本研究では、機械学習で予測する変数を従来の「副作用発現あり/なし」の名義変数から、連続変数の「薬剤投与量」に置き換えることで、バンコマイシンの初回投与量を精度高く予測できることを見いだした。今回得られた知見は、様々な薬剤の投与設計アルゴリズム構築に応用可能であり、高い発展性を有することから、医療薬学研究の発展に寄与する新手法としての価値があると考える。
|