深度学习用于量化新生儿重症监护病房的护理操纵活动
Abrar Majeedi1, Ryan M McAdams2, Ravneet Kaur3
1Department of Biostatistics and Medical Informatics, School of Medicine and Public Health, University of Wisconsin-Madison, Madison, WI, USA.
NPJ digital medicine
|June 27, 2024
概括
这项研究开发了一种深度学习方法,用于监测新生儿在新生儿重症监护室 (NICU) 的新生儿压力. 人工智能准确地量化了护理操纵活动和生理反应,有助于婴儿压力评估.
科学领域:
- 新生儿护理 新生儿护理
- 神经科学是一个神经科学.
- 人工智能的人工智能
背景情况:
- 早期的压力对神经发育障碍构成重大风险.
- 在NICU监测新生儿压力对婴儿福祉至关重要.
研究的目的:
- 量化新生儿护理操纵活动的持续时间,频率和生理反应.
- 开发和评估用于自动化压力监控的深度学习方法.
主要方法:
- 利用2个NICU的27名新生儿的289小时的床边视频和生理数据.
- 开发了一个深度学习模型来检测操纵活动,估计持续时间/频率,并整合生理信号.
- 与人类注释对比验证了该方法,并评估了其在估计生理反应方面的有效性.
主要成果:
- 深度学习方法在活动持续时间和频率 (13.8%的相对容错率) 方面取得了与人类注释相等的统计结果.
- 该方法有效估计了对护理活动的短期生理反应.
- 成功检测出具有显著生理偏差的活动,并量化了新生儿婴儿压力因子评分.
结论:
- 开发的深度学习方法为NICU中客观的新生儿压力监测提供了可靠的工具.
- 这项技术可以帮助评估护理活动对婴儿压力水平的影响.
- 提高与新生儿压力相关的神经发育风险的早期检测和管理的潜力.
相关概念视频
Nursing Implementation
5.0K
Implementation is the execution of the nursing care plan developed during the planning phase.
The five steps to implementing effective nursing care include reassessing the patient, reviewing and revising the existing nursing care plan, organizing the resources and care delivery, anticipating and preventing complications, and implementing nursing interventions.
The five steps to implementing effective nursing care include reassessing the patient, reviewing and revising the existing nursing care plan, organizing the resources and care delivery, anticipating and preventing complications, and implementing nursing interventions.
5.0K
Documentation in Long-Term and Home Healthcare Setting
878
Documentation in long-term care facilities and home healthcare settings is crucial for ensuring continuous, coordinated, and comprehensive care for patients. Each setting has its specific documentation processes and tools:
Long-Term Care Facilities
Long-Term Care Facilities
878


