Deep-learning artificial intelligence tool may accurately measure posterior tibial slope
Key takeaways:
- A deep-learning computer vision algorithm may accurately measure posterior tibial slope on lateral knee radiographs.
- The algorithm provided accurate measurements within 5 seconds.
WASHINGTON — According to presented results, a deep-learning computer vision algorithm may accurately measure posterior tibial slope on lateral knee radiographs.
“Posterior tibial slope has significant influences on biomechanics and kinematics, specifically with how it affects the tibia translating anteriorly,” Yining Lu, MD, said in his presentation at the American Orthopaedic Society for Sports Medicine Annual Meeting. “Its importance as a risk factor for failure after primary or revision ACL reconstruction has been well-reviewed in the literature,” he added.
Lu and colleagues used a deep-learning computer vision algorithm to automatically measure posterior tibial slope on short leg lateral radiographs from 300 patients who underwent ACL reconstruction. The algorithm factored in tibial shaft, tibial joint surface and tibial tuberosity segments. Researchers compared the performance of the deep learning algorithm to manual human measurements.
According to the abstract, the deep learning algorithm achieved a mean Dice similarity coefficient of 0.885 on the validation cohort. Mean difference between the deep learning algorithm and manual human measurements was 1.92°. Researchers incorporated the algorithm into a digital application, which provided accurate measurements of posterior tibial slope within 5 seconds, according to Lu.
“[Artificial intelligence] is not the ‘end-all be-all.’ It’s not a ‘cure-all.’ So, it’s only as good as your best data. We likely need to validate this with comparisons to measurements on the gold standard, which are long leg radiographs,” Yu said. “It still remains to be seen whether or not these measurements enable clinical prediction in terms of outcomes, such as graft failure or revision,” he concluded.
Collapse
Lu Y, et al. Paper 44. Presented at: American Orthopaedic Society for Sports Medicine Annual Meeting; July 12-16, 2023; Washington, D.C.