Development of computer-aided diagnostic system for early detection of dementia

点击翻译按钮获取中文摘要

We studied the possibility of making an objective early detection of dementia. Subjects consisted of 30 patients with Alzheimer-type dementia, and 30 healthy elderly individuals, with no statistically significant difference in age or sex between the two populations. Texture analysis was performed on cerebral parenchyma images extracted from magnetic resonance images of the head for all the subjects. The texture analysis confirmed Alzheimer-type dementia 24 (80.0%) of the 30 patients with the disease, and the normal brain in 22 (73.3%) of the 30 healthy individuals. However, neural network confirmed Alzheimer-type dementia 30 (100%) of the 30 patients with the disease, and the normal brain in 28 (93.3%) of the 30 healthy individuals. These results indicate that this diagnostic technique provides a useful system for early detection of dementia.