碩士 / 國立雲林科技大學 / 工業工程與管理系 / 105 / Alzheimer's disease mainly occurs in the elderly over the age of 65. Since the global aging of population is increasingly serious, there is a trend that the dementia patients growing faster and faster.
Alzheimer is a kind of progressive degenerative brain disease.The current diagnostic method is the use of simple intelligent state inspection such as Mini-Mental State Examination、Clinical Dementia Rating and Global Deterioration Scale to evaluate patient behavior and cognitive ability. Sometimes we will also use the Computed Tomography、Positron Emission Tomography、Magnetic Resonance Imaging and other medical equipment to do auxiliary diagnosis.
Therefore, this study would like to expand the traditional measurement methods to figure out a set of objective and effective way to identify the Alzheimer, hoping that we can find a decision that can provide doctors with a more discernible indicator of the diagnosis of normal and the Alzheimer.
The methods are MRI images segmentation of brain regions, the left half, right half, the integrity of the white matter of the whole brain seven regions of the brain (Frontal、Temporal、Parietal 、Occipital、Limbic、midbrain、Sub lobar) and the use of diffusion tensor imaging in λ_1, λ_2, λ_3, Fractional Anisotropy, Mean Diffusivity, C_l, C_p, C_s, Radial Diffusivity. By using the attribute ordering algorithm, these feature indices are added to the feature combination according to the attribute importance, then placed in GRNN and Random Forests for analysis. Finally, the performance of decision rules is governed by the receiver operating characteristic curve.The aim of this study was to add a new measure of the way to find effective identification of dementia disease and normal classification.
Identifer | oai:union.ndltd.org:TW/105YUNT0031047 |
Date | January 2017 |
Creators | CHEN, YI-CI, 陳宜琪 |
Contributors | FU, JA-CHIH, 傅家啟 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 91 |
Page generated in 0.0128 seconds