Personalized classification using inter-limb movement variability in acl reconstructed knees using wearable sensors

We evaluated inter-limb movement variability from wireless sensor data captured during treadmill walking and jogging gait. Using machine learning algorithms, gait features were extracted and analysed to classify patients with Anterior Cruciate Ligament reconstructed (ACLR)knees from healthy subjects with >90% accuracy. All patients were studied at a time following surgery when decisions are made regarding return to unrestricted physical activity. This study is a step towards using sensor-based gait features to guide healthcare decisions.