Towards Contactless Patient Positioning.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Public Health

Background:

• The COVID-19 pandemic caused healthcare system strain and personal protective equipment shortages, increasing infection risk for medical professionals.
• Understaffed hospitals struggle to manage patient influx due to healthcare worker infections.
• There is a critical need for solutions that minimize physical contact during patient examinations.

Purpose of the Study:

• To design and develop a contactless patient positioning system for remote scanning.
• To reduce physical contact time between healthcare workers and patients.
• To alleviate the strain on healthcare systems during the pandemic.

Main Methods:

• Development of a contactless workflow including automated calibration, positioning, and multi-view synthesis.
• Implementation of a novel robust dynamic fusion (RDF) algorithm for accurate 3D patient body modeling.
• Utilizing multi-modal inference for adaptable training across various sensor types and applications.

Main Results:

• The system enables patient scanning without physical proximity, ensuring automated calibration and technician verification of positioning.
• The robust dynamic fusion algorithm provides accurate 3D patient body models.
• Extensive experiments demonstrated the system's efficacy using diverse datasets.

Conclusions:

• The contactless patient positioning system effectively reduces physical contact, enhancing safety for healthcare professionals during the COVID-19 pandemic.
• The system's multi-modal inference capability allows for scalable deployment across different healthcare settings and applications.
• The technology has shown positive impact in frontline hospitals and is expected to see increased global adoption.