Anthropometric Features Based Gait Pattern Prediction Using Random Forest for Patient-Specific Gait Training
Using lower limb rehabilitation robots to help stroke patients recover their walking ability is becoming more and more popular presently. The natural and personalized gait trajectories designed for robot assisted gait training are very important for improving the therapeutic results. Meanwhile, it has been proved that human gaits are closely related to anthropometric features, which however has not been well researched. Therefore, a method based on anthropometric features for prediction of patient-specific gait trajectories is proposed in this paper. Firstly, Fourier series are used to fit gait trajectories, hence, gait patterns can be represented by the obtained Fourier coefficients. Then, human age, gender and 12 body parameters are used to design the gait prediction model. For the purpose of easy application on lower limb rehabilitation robots, the anthropometric features are simplified by an optimization method based on the minimal-redundancy-maximal-relevance criterion. Moreover, the relationship between the simplified features and human gaits is modeled by using a random forest algorithm, based on which the patient-specific gait trajectories can be predicted. Finally, the performance of the designed gait prediction method is validated on a dataset.
KeywordsPatient-specific gait Anthropometric features Random forest Gait prediction
- 6.Luu, T.P., Lim, H., Qu, X., Hoon, K., Low, K.: Subject-specific lower limb waveforms planning via artificial neural network. In: 2011 IEEE International Conference on Rehabilitation Robotics, pp. 1–6. IEEE (2011)Google Scholar
- 8.Luu, T.P., Lim, H.B., Qu, X., Low, K.: Subject tailored gait pattern planning for robotic gait rehabilitation. In: 2010 IEEE International Conference on Robotics and Biomimetics, pp. 259–264, December 2010Google Scholar
- 11.Peng, H., Long, F., Ding, C.: Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans. Pattern Anal. Mach. Intell. 27, 1226–1238 (2005)Google Scholar
- 13.Refaeilzadeh, P., Tang, L., Liu, H.: Cross-validation. In: Encyclopedia of Database Systems, pp. 532–538. Springer, New York (2009)Google Scholar