IOSR Journal of Computer Engineering (IOSR-JCE) e-ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 22, Issue 3, Ser. III (May - June 2020), PP 01-12 www.iosrjournals.org DOI: 10.9790/0661-2203030112 www.iosrjournals.org 1 | Page High Accuracy of Brain Signals Predictions Using Electroencephalography Carlos J. O. Santos 1 , Rodrigo M. de Figueiredo 2 , Sandro J. Rigo 3 1 (Graduation on Automation Engineering, Unisinos University, Brazil) 2 (Graduation on Electrical Engineering, Unisinos University, Brazil) 3 (Graduation on Applied Computing, Unisinos University, Brazil) Abstract: Background: To accurately interpret the human brain has been a desire of humankind since the creation of the first device capable of reading its electrical impulses. The applications resulting from this knowledge can represent a significant advance in current technology for brain-computer interaction. A device called Electroencephalogram can read electrical signals from the scalp, and that signals can be treated and comprehended. Nevertheless, the literature in this field has some gaps to be fulfilled, mostly related to the signals processing and its efficient classification regarding human movements. Materials and Methods: The objective of this research was to develop a method capable of receiving, processing, and understanding the electrical impulses of the brain related to different body movements, using an electroencephalography device with a non-invasive approach. To this end, a commercial electroencephalography device called Emotiv Epoc+® was used to capture brain signals. The obtained data is used in an artificial neural network, which can determine which movement was performed. We developed the pre-processing of the data using the Parseval Theorem as a contribution to this research, regarding the precision increase in data interpretation. Results: The device created was able to determine, given a certain period, the movements with a high accuracy and performance. Conclusion: This research also contributes by providing a method capable of predict with 100% accuracy movements from signals obtained with an electroencephalography device and how to manipulate the data to achieve such high accuracy. Key Word: Brain; Electroencephalography; Artificial Neural Networks. --------------------------------------------------------------------------------------------------------------------------------------- Date of Submission: 22-05-2020 Date of Acceptance: 09-06-2020 --------------------------------------------------------------------------------------------------------------------------------------- I. Introduction Living in a connected world in which concepts such as the Internet of Things (IoT) and AI (Artificial Intelligence) are becoming part of our everyday life, the human being perceives himself within a computerized environment, and his connection with the machines becomes more intimate in conjunction with the industry 4.0 revolution. The objective proposed in this project is to develop a stable system, which allows the reading and connection of the human brain to computers - developing a brain-computer interface - and, later, connecting this interface to other programs or applications that make use of signals collected. The development of this method allows higher speed in processes performed usually with a mouse and keyboard and make available to impaired people the convenience of using computer systems. Some techniques already used today allow this brain signal reading operations and among them we find the electroencephalogram, a non-invasive exam widely used in the medical field for the diagnosis of brain diseases or abnormal neurological behavior. The electroencephalogram is one of the bases used in this work because, in addition to enabling the reading of brain signals by a non-invasive method, it also has an acceptable compatibility for the subsequent treatment of these signals. As the general process of electroencephalography (signal capture - signal treatment - application development with the signal) is quite extensive, a commercial data capture device will be used to validate the development carried out so that, in this way, it is possible to direct the focus of this project to the improvements in the treatment of the signal itself. Once treated, it is possible to use this signal, for example, to send commands to external equipment such as mechanical arms and legs (examples of evolution in accessibility), to communicate the program with various other programs that perform direct contact with human beings (reading the level of emotions like happiness) or integrating it with digital games using movement readings. To read brain signals and send these data to the computer, an Emotiv Epoc+® will be used, a commercial EEG device