2023 Fiscal Year Final Research Report
Predictive models for sedation and analgesia scales using artificial intelligence.
| Project/Area Number |
18K08896
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 55060:Emergency medicine-related
|
|---|
| Research Institution | Yokohama City University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
高木 俊介 横浜市立大学, 附属病院, 准教授 (90644823)
|
|---|
| Project Period (FY) |
2018-04-01 – 2024-03-31
|
| Keywords | 人工知能 / 表情認識 / 遠隔集中治療 / 鎮静スケール |
|---|
| Outline of Final Research Achievements |
The purpose of this study was to create a model for the continuous calculation of the sedation scale based on facial images of patients' expressions acquired by a camera installed in the ceiling of a hospital room. The goal was to develop a model to discriminate between the two groups of A and VPU in the AVPU score [Alert (clear consciousness) Verbal (responsive to voice prompts) Pain (responsive to pain stimuli) Unresponsive (no response)]. The agreement of the model to discriminate the two groups, A and VPU, based on the percentage of eyes open and closed, with the discrimination by health care providers was approximately 50-80%. Based on these results, future projects will be directed toward implementation of the model in real-world clinical practice.
|
| Free Research Field |
遠隔集中治療
|
| Academic Significance and Societal Importance of the Research Achievements |
高い精度を持つ開閉眼予測モデルの作成が達成された。また、意識レベル評価スケールであるAVPUにおけるAとVPUとを識別するモデルについても比較的高い精度を持ったモデルの作成が達成された。当初はRichmond Agitation-Sedation Scale等のより一般的なスケール評価が目標であったが、体動の検出などより高度なモデルの作成が必要であった。今後は、体動の検知モデルの開発やバイタルサインのトレンドを評価に加える等、モデル精度を発展的に高めることが目標となる。開閉眼評価は一般的な鎮静・意識スケールに含まれる項目であり、本研究はこれらの評価モデル開発の足掛かりとしての意義があった。
|
