Behavioral analysis using laser scanners in the intensive care unit of a university hospital

Abstract

In hospitals, intensive care units (ICUs) are places for severely ill patients requiring intensive care, such as after major surgeries or in critical conditions. The ICUs’ complex flow of people and goods involves multidisciplinary medical staff and devices which frequently move in and out, presenting challenges in monitoring and management. This research aimed to develop a method of analyzing medical staff behavior in ICUs using 2D point cloud data from cost-effective and compact 2D laser scanners, prioritizing privacy and safety. The method involves clustering the 2D point cloud data by point distances. The trajectories of people within the ICU were then extracted by tracking these clusters using the SORT (Simple Online and Real-time Tracking) algorithm. Subsequent validation of these extracted trajectories illustrated their effectiveness in analyzing medical staff behavior. This approach offers insights into ICU operations, potentially enhancing efficiency and patient care by optimizing staff movement and resource allocation.