Abstract:
Gait analysis is the process of collecting and analyzing the data of human walking movement. This kind of information is useful and widely used in many areas, especially in clinical research study. Doctors and physicians use it to identify a type of movement disorder diseases. Different illnesses require different types of doctors and treatments. This thesis presents an automated diagnosis system using gait data to classify patients into three groups: Normal, Sick/Knee Osteoarthritis (OA), and Sick/Parkinsons disease. In the study, there are 88 samples (patients): 27 elder normal, 34 osteoarthritis, and 27 Parkinson. The best classification scheme is based on a feature set of four major positions in the gait cycle and SVM using the One-VS-One strategy. The experimental results show that the proposed system achieved 70% accuracy. To primarily assist the diagnosis, the prototype system was implemented showing a comparison of gait cycle graphs between a patient and normal people.