| Parkinson’s disease (PD) is a progressive neurologic disorder characterized by the cardinal motor signs of bradykinesia, tremors, rigidity, and abnormalities of posture and gait. These motor impairments lead to the varying of the characteristic gait pattern of PD, which frequently accompanied by split step, festinating gait, freezing of gait (FOG),difficulty in turning, postural instability and postural conversion disorder. At present,the analysis and evaluation of gait disorder in Parkinson’s disease mainly depend on the clinician’s experience observation, the response of specific drug, the international general rating scale and the way of the questionnaire. Although these methods are simple and easy to use, and also have been widely utilized in clinical assessments and research studies, its susceptibility are easy to be influenced by the physicians’ clinical experience or the patient’s transient symptoms. So measuring tools and evaluating methods to objectively quantify and assess the severity of a pathological gait pattern are the clinical demands. Several measuring tools and evaluating indexes for quantifying and assessing the severity of a pathological gait were presented in the literature. Due to the complexity and variability of symptoms of gait disorders in patients with PD, although a variety of analytical methods and evaluation systems already exist, they often focus on the analysis of a single signal of gait disorders, the comprehensiveness of quantitative analysis and accuracy can not be effectively guaranteed. In addition, most systems are still used in the laboratory, and have not been generally used in the quantitative analysis of gait disorders in patients with Parkinson’s disease. Therefore, based on the biomechanical characteristics of gait analysis and the existing domestic and international research on the gait acquisition device, a novel U-shaped electronic walkway system (USEWS) was established in this dissertation with the aim of assessing the gait variation in PD persons on the basis of the flexible force-sensitive technology and the 3D force-measuring platform. Furthermore, based on the gait characteristics acquired by the novel USEWS, the objective quantitative evaluation and quantitative classification evaluation method of gait disorder in Patients with PD were studied in this dissertationThe main research contents of this dissertation are presented as follows: design and implementation of a novel type of U-shaped electronic walkway system,comprehensive extraction of gait signal characteristics in patients with PD, the objective quantitative assessment and quantitative classification evaluation of gait disorder in patients with PD, and gait recognition and evaluation in patients with PD.The specific research contents of this dissertation are as follows:(1) Based on the flexible force-sensitive technology and the 3D force-measuring platform, a novel USEWS was established in this dissertation with the aim of assessing the gait variation in persons with Parkinson’s Disease. This system was able to provide comprehensive access to plantar pressure, gait kinematics and kinetic parameters in patients with PD during static standing and dynamic walking. It can also obtain the gait characteristics in the cornerpart of the USEWS.(2) Aimed at the problem of footprints segmentation in plantar pressure images, a phased segmentation algorithm was proposed. By using the difference of the undulation between the inside and outside arches of the foot arch and the outer foot arch, the real-time dynamic recognition of the walking footprints was realized.(3) The differences between the gait in PD patients and normal controls were comprehensively compared from four aspects: comparative analysis of static and dynamic plantar pressure, statistical analysis of gait kinematics and kinetic parameters,evaluation of gait symmetry and bilateral coordination and evaluation of the nonlinear features of gait. On the basis of the above analysis of differences, the objective quantitative evaluation and quantitative classification evaluation method of gait disorder in patients with PD were studied. Finally, according to the characteristics of unbalanced data composed of PD patients and healthy subjects, a cost-sensitive support vector machine (CS-SVM) method was proposed to construct the gait signal classification model. These results further validated the effectiveness and feasibility of the USEWS designed in this dissertation to assess gait disorder in patients with PD.Although the novel U-shaped electronic walkway system constructed in this dissertation is limited by space, it can acquire the gait characteristics of PD patients in static standing, straight and turning situations. The experimental results show that the objective and quantitative evaluation method of gait disorder presented in this dissertation has positive guiding significance for the diagnosis and treatment of abnormal gait in patients with PD. In addition, the multi-classification SVM model constructed in this dissertation has achieved initial success with the error classification concentrated in the two adjacent rating scales. Further, compared with the k-nearest neighbor algorithm (KNN), support vector machine classification (SVM) and Naive Bayesian classifier (NBC), the classification model constructed in this dissertation based on a cost-sensitive support vector machine method has the best comprehensive performance. The CS-SVM model not only effectively improves the recognition accuracy of PD patients, but also effectively reduces the misjudgment cost of clinicians in the misdiagnosis of patients with Parkinson’s disease. At present, the novel U-shaped electronic walkway system and the objective quantitative evaluation and quantitative classification evaluation method of gait disorder in patients with PD have been successfully applied to the clinical practice of the Affiliated Hospital, Institute of Neurology, Anhui University of Chinese Medicine, Hefei and the effect is remarkable. |
PDF Full Text Request