The Development of Computer-Aided Detection System for Alzheimer''s Disease

• Main Authors: Bo-Kai Huang, 黃柏凱
• Other Authors: Jenn-Lung Su
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/22661939722750684762

碩士 === 中原大學 === 生物醫學工程研究所 === 104 === It has been predicted by the Taiwan Alzheimer Disease Association that the amount of people suffering from Alzheimer’s will increase from 2%, in 2031, to 3%, in 2041. Current drug treatments for Alzheimer’s disease (AD) aim to retain cognitive function as well as delay disease progression, and the sooner the treatment is started the greater the effectiveness. We can assess the level of AD by Clinical Dementia Rating (CDR) and obtain the brain atrophy ratio via MRI and CT imaging. However, the manual assessment of the brain atrophy ratio is not objective and time-consuming. In this thesis, a computer-aided detection (CAD) system was presented based on the correlation between CDR and the brain atrophy ratio to provide relevant data for physicians. This CAD system was developed by calculating the ratio of white matter (WM), gray matter (GM) and brain ratio. The steps of this study include: (1) removing noise and separate the full area of brain through image pre-processing; (2) detecting white matter, gray matter and cerebrospinal fluid (CSF) by edge detection; (3) calculating the volume of WM, GM and CSF; (4) determining the volume, feature, brain ratio and atrophy level on the CAD system.; and finally, (5) using phantom image to vilify developed algorithm and 50 set CT images (3 images /set), where 20 are used as the training group and 30 as the test group, and 30 MRI images to evaluate the effectiveness of the system. Preliminary results and statistical analysis showed that the atrophy ratio and CDR scores are proportional directly. The CT image diagnostic system demonstrates an accuracy of 87.5%, sensitivity of 90%, specificity of 83.3%, and kappa of 0.53. The MRI image diagnostic system demonstrates an accuracy of 80%, sensitivity of 80%, specificity of 80%, and kappa of 0.50. These indicate that the system for the diagnosis of Alzheimer''s disease is effective and accurate. In addition, the average segmentation time for brain images in manual and developed system method are 297.2 and 7.6 seconds, respectively. The system also provides ease of use interface for physician faster diagnosis. The results showed that the atrophy ratio of CT and MRI images and CDR scores are highly relatively. In the future, we can make use of CT imaging and MRI imaging as a primary diagnostic tool for early detection and reduce the patients’ need for other diagnostic procedures, thereby reducing medical waste.