AOSSM 2023 Annual Meeting Recordings no CME-Deep Learning Artificial Intelligence Tool for Automated Radiographic Determination of Posterior Tibial Slope in Patients with Anterior Cruciate Ligament Injuries

Deep Learning Artificial Intelligence Tool for Automated Radiographic Determination of Posterior Tibial Slope in Patients with Anterior Cruciate Ligament Injuries

Back to course

Video Transcription

Video Summary

The video transcript discusses the significant influence of posterior tibial slope on knee biomechanics and its importance as a risk factor for failure after anterior cruciate ligament reconstruction. Various methods of measurement are mentioned, with the full-length standing knee to ankle being the most accurate but short-leg knee radiographs being the standard in most institutions. The study aimed to develop a deep learning tool to automatically measure posterior tibial slope on short-leg knee radiographs. The tool achieved a good concordance with measurements made by surgeons, with a mean difference of 1.9 degrees. The algorithm demonstrated efficiency, completing measurements for 90 images in under a minute. Further validation through gold standard measurements and external data is required, as well as determining the clinical prediction outcomes based on these measurements.

Asset Caption

Yining Lu, MD

Keywords

posterior tibial slope

knee biomechanics

anterior cruciate ligament reconstruction

measurement methods

deep learning tool