178

Bioinformatics of the Brain

can also mix two or more different kinds of BCI types in one application and

develop hybrid BCI systems to provide more robust control.

c) Synchronization: The categorizing of the BCI systems can also be done

based on operating approaches which are synchronous or asynchronous proto-

cols. Synchronous (cue-paced) systems direct the user via cognitive tasks at a

desired time. Conversely, in asynchronous (self-paced) systems, the users can

think freely at any time without any cue from the system.

d) Dependability: The other classification type of BCI systems is based on

the interaction between the user and the system. If the users need their muscles

for system control, then this system is called a dependent BCI system whereas

in independent BCI systems, muscle activity is not required to generate the

brain signals that the system uses.

e) Timing: Another aspect of using BCI systems is related to implementing

the system in real-time or offline.

7.2.4

Hardware and Software Technology of EEG-based BCI

Systems

According to Nicolas-Alonso and Gomez-Gil [21], BCI is a system where the

hardware and software communicate with each other to provide user-machine

interaction. The diversity of both software and hardware has increased with

the development of technology.

Multiple hardware technologies developed by various companies are used

in the current investigations, especially in the EEG signal acquisition phase

of BCI. Generally, hardware technologies consist of electrodes, amplifiers, an

A/D converter, and a recording device. There are many technologies currently

in use and they mostly vary based on the type of transmission (wireless or

wired), wearability, sensor type (wet and dry sensors), number of electrodes

(from 1 to 280), and sampling frequency property (128, 256, 512 Hz, and so

on) [6]. Also, one company could manufacture more than one device with

different characteristics (please see the comprehensive review: [22]).

The acquisition unit (e.g., headsets) and the translation unit (e.g., ampli-

fier) could be connected via wired technologies as well as wireless technologies

such as Wi-Fi or Bluetooth. Both of them have several advantages and disad-

vantages [17]. In the EEG signal acquisition, wired and wireless devices use

different sensor technology that is used to capture the brain activities of the

users [23]. Wired devices use wet sensor technology that requires a conductive

gel as an interface between sensors and scalps whereas wireless devices use

dry sensor technology without the need for the use of gels. Moreover, EEG

devices may contain different numbers of sensors (electrodes) starting from 1

and up to 280, which are placed according to various types of montage with

the standard electrode positions [24, 25]. In the literature, the number of elec-

trodes is associated with the sampling frequency. While the devices with a low

number of electrodes have a general sampling rate of 500 Hz, the devices with

high electrode numbers archive a sampling rate of higher than 1,000 Hz. After