Automatic Diagnosis of Alzheimer Disease from Electroencephalograms

摘要

Preventive medicine requires early detection of diseases in order to begin a treatment that would allow a better life quality for the patients. In the Alzheimer Disease (AD) early diagnosis is key for taking actions that prevent the disease from advancing too quickly. People with AD are normally not aware of the problem when they are in the early stages of the disease. For these reasons, it is important for everybody to have easy access to cheap and fast diagnosis systems. In this paper, the implementation of a simple diagnosis system for AD is described, the system uses Electroencephalograms (EEG) since they are obtained in a non invasive way and have been proved to be useful for the diagnosis of AD. The system extracts features from the EEG to build a feature vector and implements a simple dichotomizer. The tests included 16 individuals suffering from AD and 34 healthy individuals, the results show that the system is functional and has an accuracy similar to equivalent systems in the state-of-the-art.