Study On The Measurement And Feature Extraction Of Gait Kinematics For Functional Evaluation

Functional evaluation of gait is the study of the biomechanics of human movement aimed at quantifying factors governing the functionality of the lower extremities, which is crucial for the detection of gait disorders, establishment of clinical decisions, identification of balance factors, and assessment of clinical gait interventions and rehabilitation programs. In addition, gait studies have contributed significantly to the design of artificial limbs(prostheses) and artificial locomotor controllers used in exoskeletons and robotics. However, current technologies of gait kinematic parameters measurement and gait feature extraction still cannot fully meet the requirements of clinical use, it still has a number of challenging issues to be further studied.Although the gait motion measurement systems based on optical 3D motion capture instrument have been widely applied in gait studies, their marker set models need more markers which are required to be placed accurately with respect to specific anatomical landmarks(ALs). The positions of the ALs are usually identified by specialists through touch. Therefore, failure to place markers accurately is seen as a major contributor to the reductions in measurement accuracy and repeatability. Because of the long preparation time caused by the involvement of more markers and the manual process of markers locating, in pathological gait measurement, the patients can get tired and their gait may be changed, the authenticity of the results will be affected accordingly.Most studies of gait function evaluation using kinematic parameters have been confined to evaluation of spatiotemporal gait parameters or a limited number of characteristic kinematic values of joint angle curves(e.g. maximum extension and flexion, total sagittal plane excursions, etc.). This is important for a rapid overall evaluation but is difficult to provide more detailed information about the abnormal gait. Moreover, the complex relationships between parameters to be investigated make the measurement system more complex, and reduce the interpretability of measurement outcomes.Aimed at improving the gait motion measurement techniques and exploring the underlying information of the gait characteristics contained in foot motions, the key technologies of human gait measurement and feature extraction were deeply studied in this dissertation. The main content could be presented as follows:1. By the analysis of anatomic structure of the main lower limb joints, The multi-body kinematics model of human lower limb was established based on reasonable simplifications to the body segments and joints of lower limbs. The motion equations of foot and lower limb joints were derived. Through kinematic analysis, a new marker set configuration principle and a new measuring structure based on Optotrak Certus? motion capture instrument are presented, which simplified the preparation time and reduce the space occupation.2. According to the kinematic model, the constraint equations of lower limb motion were constructed, an algorithm for joint centers, geometric parameters of the model, and gait kinematics was established. The feasibility of the algorithm and the gait kinematics measurement system was verified by experiment. The algorithm is the key technology of the gait measurement system, which overcome the limitation of the traditional systems. It reduce the number of measure points, and make the markers not need to be accurately placed and not depend on the ALs.3. Combination with the General Tau Theory in cognitive science, this dissertation innovative put forward an improved intrinsic tau guidance strategy, namely intrinsic Tau-J guidance strategy. Through theoretical analysis and experimental research on the foot movement by using Tau strategies, a simple, sensitive and interpretable single parameter index was proposed to evaluate the performance of posture control during walking, which is of great significance to explore the intrinsic nerve and cognitive mechanisms of abnormal gait.4. Through the experiment and the analysis of the gaits of healthy elderly and patients with Parkinson’s disease(PD), a gait feature extraction and evaluation method based on frequency band energy density and wavelet entropy was studied in order to reflect the dynamic feature and complexity of the gait, and thus to further explore gait characteristic information contained in foot movement. The experimental results demonstrate that this method can not only capture the underlying information associated with gait changes sensitively, but also reflect the characteristics of PD tremor. For the gait evaluation of PD patients, this method has shown its particular advantage.In summary, key techniques were presented in this study to measure the gait kinematics, a new measurement scheme and a new algorithm was proposed. Through the analysis of the time series of foot movement, the underlying information of the gait abnormalities and gait performance in the foot movement were also explored, several new gait features and their extraction and evaluation methods are proposed, which may provide theoretical support for further simplifying the gait measurement method and improving the gait evaluation effect. The results of the study not only provided effective means for the measurement and quantitative assessment of human gait, but also provided the method and clue for the human movement mechanism analysis.