EEG-based BCI Systems in Neuropsychiatric Diseases

175

(e.g., functional resonance imaging (fMRI), electroencephalography (EEG)),

acquisition techniques, feature extraction methods, and decoding algorithms

(Machine Learning (ML), Deep Learning (DL) techniques) are utilized. The

main aim of this chapter is to provide an overview of EEG-based BCI applica-

tions to detect neuropsychiatric diseases, including Alzheimer’s Disease (AD),

Parkinson’s Disease (PD), Mood Disorders, Schizophrenia (SCZ) Spectrum,

and other psychotic disorders. In the meanwhile, the fundamental technologies

for acquiring, encoding, and decoding phases of BCI are briefly described.

This chapter is organized as follows: Section 7.2 introduces the background

information about EEG-based BCI including its definition, history, categories,

and technologies. Section 7.3 presents the application steps of EEG-based

BCI, starting with the acquisition of EEG signals followed by pre-processing,

feature extraction/selection phases, and application of artificial intelligence

techniques. Section 7.4 illustrates current BCI applications. Section 7.5 dis-

cusses the challenges and future perspectives of BCI technologies. Section 7.6

concludes the chapter.

7.2

Understanding the Brain-Computer Interface (BCI)

In this section, the background information related to BCI technology is given

starting with the definition of BCI and followed by the idea of emergence, the

current types and the hardware and software technologies to implement BCI.

7.2.1

What is BCI?

The Brain-Computer Interface (BCI) term, also sometimes called brain-

machine interface or Human-Machine Interface in the recent literature, was

originally used in the study by Vidal [1] and described as a direct link between

man and machine (in particular, a computer) to provide a dialog by utilizing

brain signals. However, due to the rapid improvement of technology in the last

decades, there has been an exponential growth of BCI applications [2] and var-

ious types of BCI systems emerged, such as active BCI, and passive BCI (see

Section 7.2.3). Therefore, multiple terms have been started to be used for the

description of BCI systems due to the high heterogeneity of devices, protocols,

applications, and disciplines. According to Antonietti [3], 34 definitions of the

different BCI types have been used in the current literature of various fields,

including neuroscience, psychology, clinical neurology, computer science, and

engineering.

The common goal of the earliest BCI systems is to provide an alternative

way of controlling peripheral movements without using neural pathways, es-

pecially for people with motor impairment or paralysis (see the review by Xu

[4]). These kinds of systems were accepted as promising tools that translate