In order to evaluate one's risk of tripping easily in near future, this study proposed a method to estimate the minimum toe clearance (MTC) during walking from Floor Reaction Force (FRF) of the users and a model that made from the gait data of third party, and evaluate its robustness and accuracy. First, we measured the FRF and the MTC of one gait cycle from 26 health young and 6 healthy elderly participants. Second, we build 10 different models to estimate the MTC from the FRF, by using randomly chosen 20 young participants' data. The robustness of the proposed method has evaluated with the contribution ratio and independent variables that used to build the models. Moreover, the accuracy of the proposed method has evaluated with the root mean squared error (RMSE) and coefficient of correlation between the estimated MTC and the actual MTC that derived from the remaining data of 6 young and 6 elderly participants. The contribution ratio of the models were 0.67±0.03 and 5 independent variables were commonly used to build the models. RMSE between the estimated MTC and the actual MTC were 4.6±0.7[mm] for young participants and 6.4±0.5[mm] for elderly participants, which is similar to the results of previous study, and the estimated MTC and the actual MTC show a signifi cant correlation. From these results, we confi rmed that proposed method can build robust model regardless of the participants group, and can estimate the MTC as accurate as previous study.
View full abstract