Assessment of Visual Patient Re-Identification in a Live Emergency Department Waiting Room
2023 IEEE 14th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)(2023)
摘要
The re-identification of individuals through computer vision has been extensively studied. As “smart hospitals” come online, the use of computer vision to identify and monitor patients could become a core element of healthcare. Investigators therefore studied an experimental patient re-identification system in an Emergency Department (ED) ambulatory waiting room setting. The study investigated the feasibility of deploying a vision-based patient deterioration monitoring system with a patient re-identification subsystem in an ED waiting room. After implementing appropriate privacy, confidentiality, and security measures, investigators studied 440 consented patients and companions who were enrolled over a 12-month period to determine the performance characteristics of the experimental system’s video-based person re-identification solution. Several commonly used convolutional neural network algorithms, from lightweight MobileNet to larger ResNet models, were investigated in the study. We report on the system’s exhibited state-of-the-art accuracy and remaining challenges.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要