| Summary: | 碩士 === 國立成功大學 === 電腦與通信工程研究所 === 101 === The problem of aging population is getting worse in the world. It is not only huge medical burden but also social burden. Dementia is one of the important problems and it makes deep fading of memory, attention, and incapacity, etc. Therefore, the health care becomes more important. How to effectively detect dementia needs to be proposed. Accordingly, most research focus on the relationship between the cognition and behavior using video or inertial sensors. This thesis proposed a series of methods to analyze the association using an accelerometer and a gyroscope and to find the features using Hidden Markov model (HMM). Analyze the normal and abnormal walking via sensor mounted on the foot, mapping to the neuropsychological tests. Help medical staff to detect the mild Alzheimer's disease which is common one of the dementia, and classify a participant. Results show that the score of the HMM is significant differences between normal control and Alzheimer’s disease in our walking tests. In the future, wish a simple walking detection method by subjects themselves using inertial sensors could be widely used.
|
|---|
