3066. Applications of Machine Learning to Optimal Facility Design for Healthcare Worker and Patient Safety
Invited abstract in session TD-13: Machine learning in healthcare, stream OR in Healthcare (ORAHS).
Tuesday, 14:30-16:00Room: Clarendon SR 1.01
Authors (first author is the speaker)
| 1. | Hengameh Hosseini
|
| Health Administration, University of Scranton/Penn State University |
Abstract
Designing healthcare facilities to accommodate individuals with disabilities, particularly for workers in physically demanding roles, presents significant challenges. Many healthcare facilities fail to meet the needs of workers with disabilities, especially those with physical disabilities and low vision, who are at increased risk of injury due to poorly designed environments. This study investigates the impact of facility design on healthcare workers with low vision, focusing on the potential safety risks posed by insufficient accommodations.
We conducted surveys, interviews, and in-facility observations across seven healthcare facilities in the U.S., representing various sizes and ownership models. Our analysis identified five major design shortcomings common across all facilities that could endanger physically and visually impaired employees, ranging from minor to severe hazards. A key finding was the lack of design empathy among facility managers and administrators.
To address these issues, we propose three solutions to guide design: (1) leveraging an open-source Augmented Reality (AR) application to simulate the low-vision experience for designers; (2) employing a machine learning model to analyze patient and employee feedback to detect facility shortcomings; and (3) using computer vision to predict hazardous facility features for visually impaired employees. Through comparative experiments, we demonstrate which methods can be effective under specific conditions.
Keywords
- Health Care
- Facilities Planning and Design
- Machine Learning
Status: accepted
Back to the list of papers