Classifying Motor Imagery Actions in Electroencephalogram Signals of Older Adults Using Artificial Neural Networks

Brain-computer interfaces (BCIs) have enabled users to control computing applications by measuring brain activity, such as through electroencephalograms (EEGs). These BCIs have the potential to assist individuals with cognitive and motor impairments. However, research on BCIs has shown that their performance can vary among study subjects, particularly in the case of older adults. Cognitive impairment is a natural condition that occurs in older adults, and if left untreated, it can lead to the development of dementia, which, in turn, results in dependency for performing basic daily life activities. This paper presents the results of a preliminary analysis on the use of BCIs to recognize motor imagery of up to five right-hand movements, using a supervised machine learning algorithm. This approach aims to facilitate the interaction of older adults with cognitive stimulation applications, mobility devices, and other tools that improve their quality of life.