ScholarWorks@동국대학교

초록

With aging populations worldwide, the estimated number of people living with dementia is expected to reach 150 million by 2050. Detecting dementia in its early stages is a cost-effective way to reduce costs, both personally and socially. It also improves the quality of life of elderly people from a healthcare perspective. Early dementia detection methods mainly rely on paper questionnaires or cognitive tests; however, these methods are time-consuming and expensive as they require expert evaluation. Analyzing Wartegg-Zeichen Test (WZT) images drawn by the elderly to predict early dementia using deep learning provides a technique that can support the development of novel, efficient, and accessible methods to help detect the early stages of dementia. This study aimed to predict early dementia by analyzing WZT images drawn by elderly people at a welfare center using a deep learning model and automated techniques. The collected data were analyzed by deep learning using a convoluted neural network (CNN), computer vision, and SRGAN for artificial intelligence analysis, where CNN showed the best results in predicting an early dementia diagnosis. In an aging society, the early detection of dementia in elderly individuals can help improve their personal quality of life from a healthcare perspective. The analytical models of deep learning that will be developed in the future will create many opportunities for various applications along with IT convergence.

키워드