Information about project titled 'Predicting Outcome of Anterior Cruciate Ligament Reconstruction Through Machine Learning Analysis of National Knee Ligament Registries'
Predicting Outcome of Anterior Cruciate Ligament Reconstruction Through Machine Learning Analysis of National Knee Ligament Registries
| Details about the project - category | Details about the project - value |
|---|---|
| Project status: | Ongoing |
| Project manager: | Kyle Martin, Gilbert Moatshe, Roald Bahr, Lars Engebretsen |
| Coworker(s): | Andreas Persson, Håvard Visnes, Anne Marie Fenstad, Eivind Inderhaug |
Description
Background: Several factors have been identified as being associated with an increased risk of anterior cruciate ligament (ACL) reconstruction failure. However, the ability to accurately translate these factors into a quantifiable risk estimation at a patient-specific level has remained elusive.
Purpose/Aim: To apply machine learning to the Norwegian (NKLR) and Danish (DKRR) knee ligament registries in order to:
1. Evaluate factors for predicting ACLR outcome
2. Predict risk of a poor outcome at the individual and group levels
3. Create clinically useful tools for surgeons
Methods: The NKLR and DKRR record data for all patients undergoing ACL reconstruction in Norway and Denmark, respectively. De-identified data from these registries will be analyzed using both supervised and unsupervised machine learning approaches to develop algorithms capable of estimating risk of revision surgery and inferior patient reported outcome. Algorithms will then undergo external validation and accuracy will be compared with that of human surgeon predictions.
Implications: This project represents the first application of machine learning approaches to produce clinical prediction algorithms based on national registry data. This new information can be used in the clinical setting to guide surgical discussions and set realistic expectations with patients with an ACL tear.