In a study of 86 people with no discernible symptoms of osteoarthritis, retrospective analysis of cartilage texture maps by a machine learning classifier found that the classifier detected the beginning stages of osteoarthritis progression with 78% accuracy up to 3 years before symptom onset, suggesting that early detection may enable treatment at a stage when damage is reversible, according to the authors.
Article #19-17405: "Enabling early detection of osteoarthritis from presymptomatic cartilage texture maps via transport-based learning," by Shinjini Kundu et al.
MEDIA CONTACT: Shinjini Kundu, Johns Hopkins Hospital, Baltimore, MD; e-mail: skundu2@jhmi.edu
###
Proceedings of the National Academy of Sciences