A Non-Contact Privacy Protection Bed Angle Estimation Method Based on LiDAR.

Area of Science:

• Biomedical Engineering
• Medical Device Technology
• Sensor Technology

Background:

• Accurate head-of-bed elevation is critical in Intensive Care Units (ICUs) to prevent complications such as ventilator-associated pneumonia.
• Existing camera-based systems raise significant patient privacy concerns.
• There is a need for non-intrusive, real-time bed angle monitoring solutions.

Purpose of the Study:

• To develop and evaluate a non-intrusive bed angle detection system using LiDAR technology.
• To ensure patient privacy by avoiding visual surveillance.
• To improve the accuracy and reliability of head-of-bed angle monitoring in healthcare settings.

Main Methods:

• Utilized the Intel RealSense L515 LiDAR sensor for time-of-flight measurements.
• Implemented coordinate system transformation and plane fitting algorithms.
• Integrated a deep learning framework combining YOLO-X with an enhanced Angle-to-Joint (A2J) algorithm and customized loss functions.

Main Results:

• The LiDAR-based system demonstrated effective, real-time, and privacy-preserving bed angle monitoring.
• Achieved an average angle detection error of less than 3 degrees in experimental ICU environments.
• Validated the system's capability to monitor head-of-bed elevation without direct visual surveillance.

Conclusions:

• LiDAR technology offers a viable, privacy-conscious alternative for critical bed angle monitoring in ICUs.
• The proposed system enhances patient safety by enabling accurate detection of bed positioning.
• This technology has the potential to reduce healthcare-associated complications and improve patient outcomes.