2020 Fiscal Year Final Research Report
Development and evaluation of method for preventing Intravenous catheter failer -AI-based care recommmendation system on ultrasound images
Project/Area Number |
19K21424
|
Project/Area Number (Other) |
18H06341 (2018)
|
Research Category |
Grant-in-Aid for Research Activity Start-up
|
Allocation Type | Multi-year Fund (2019) Single-year Grants (2018) |
Review Section |
0908:Society medicine, nursing, and related fields
|
Research Institution | The University of Tokyo |
Principal Investigator |
Takahashi Toshiaki 東京大学, 大学院医学系研究科(医学部), 特任助教 (50824653)
|
Project Period (FY) |
2018-08-24 – 2021-03-31
|
Keywords | 看護理工学 / 超音波検査 / 末梢静脈点滴 |
Outline of Final Research Achievements |
All ultrasound images were collected from patients using PIVC. We evaluated 259 images. The number of False Negative images in which blood vessels could not be detected was 24, and the number of True Positive images in which blood vessels could be detected correctly was 178, resulting in a vessel detection accuracy of 76.4%. The correlation between the 178 images detected and the nurse was calculated to be r = 0.843, and no systematic error was found in the BA plot. The accuracy of the PIVC size recommended by the research nurse and that recommended by the system was 70.2%. The percentage of underestimates was 7.0%, and the percentage of overestimates was 21.9%.
|
Free Research Field |
看護理工学、老年看護学、成人看護学、看護技術
|
Academic Significance and Societal Importance of the Research Achievements |
本研究は人工知能(AI)を用いたエコー画像の自動処理により血管径や深さを識別・重畳表示しエコー画像の情報量と視認性を向上させ、点滴漏れを予防するアプリケーションを開発する。これは看護領域における AI 技術導入の先駆的研究で あり、臨床上多くの患者が経験し潜在的に我慢を強いられている点滴漏れを解決する重要な研究である。エコー画像を自動測定して穿刺針径に見合う太い血管を示す。本研究により、看護におけるエコーによる客観的アセスメントの質を担保し、経験の少ない看護師でも安全なケアが実施可能となる。結果として患者は入院してから疼痛を伴う点滴漏れを経験せず、適切な治療を受ける事が可能になる。
|