Machine learning model for predicting acl failure load

MRI allows non-invasive, quantitative assessment of ACL healing in vivo. Previously, a linear model employing MRI T2* relaxometry has been shown to predict structural properties of the ACL post ACL-repair (ACLR). This study shows that a machine learning model using common MRI features can achieve statistically similar test sample accuracy to the T2* linear model when predicting ACL failure load post-ACLR.