Estimation of knee flexion features using foot motion data from in-shoe motion sensors

Deterioration of knee joint function greatly affects quality of life. Therefore, technology is needed that accurately, inexpensively, and easily measures knee joint function. As the first step for knee deterioration assessment in daily walking, we developed estimation models for six parameters that capture the features of healthy knee movement by using an in-shoe motion sensor (IMS). We recruited 72 healthy participants to construct the estimation models. Using intra-class correlation coefficients (ICCs), we showed that five of the six models achieved “fair” or “good” agreement as characteristics of walking behavior by healthy people. We demonstrated the possibility of estimating knee behaviors as the specific flexion knee angles in only using foot-motion via single IMS for the first time. It will be helpful to make knee condition monitoring in daily living available easier.