
This study aimed to create a scalable and feasible knee osteoarthritis (OA) radiographic database using artificial intelligence (AI) tools. The researchers applied two commercially available and two custom-built AI tools to analyze six years of consecutive knee radiographs from a hospital in Denmark. The tools provided various assessments, including Kellgren-Lawrence grades, joint space widths, and implant detection. The study demonstrated the potential for building clinical knee OA databases efficiently with limited human reading time, saving approximately 800 hours of radiologist time. The approach is scalable across time and regions, enabling diverse inclusion of radiographic knee OA data for global research.
Like
Save
Share