Information Technology Reference
In-Depth Information
nen reported on a number of experiments in measuring
brain activity relating to attention using auditory stimuli. Even though attention
research involving ERPs had been going on for over 50 years at the time, N
In 1990, Risto N
ää
t
ä
ää
t
ä
nen
was keen to distinguish between the brain
'
s automatic responses to stimuli and
responses derived from someone
'
s attention and their interpretation of the heard
stimuli (N
nen
1990
). The idea of a subject being able to shift their attention at
will to auditory stimuli opened up possibilities of BCMI systems controlled by a
user
ää
t
ä
s attention to elements of what they are hearing.
Research into attention and sound has long been investigated even before the use
of EEG, and earlier research observed a phenomenon known as dichotic listening in
regard to how we focus our hearing attention. Dichotic listening is the process of
paying attention to sound from one ear whilst ignoring sound from the opposite ear.
When asked to focus on speech arriving at one ear, subjects were often unable to
recall speech of the same volume from the opposite ear (Cherry
1953
). In N
'
ä-
nen
'
s experiments, he found that the brain reacts to deviations from repetitive
sounds automatically, even when a listener focuses their attention away from what
they are hearing. This was measured with a P300 EEG response, where the potential
begins with a positive de
ää
t
ection and peaks at around 300 ms after the onset of the
stimuli. This
implied that when presented with recurring audio
information, the brain reacts automatically, and predictably, to deviations in audio
patterns.
Throughout the 1990s and early 2000s, further research into how the brain
responds to auditory stimuli shed light on how the brain processes our perceptions
of music. A key area in this
'
oddball paradigm
'
field is the study of meaning held within ERPs, building
upon previous research into how the brain processes language (Besson and Macar
1987
). Here, the term meaning has more depth than mere EEG association to input.
Besson and Fa
ta (
1995
) demonstrated how different responses within ERPs are
elicited when subjects listen to musical phrases that end either congruently or
incongruently in pitch or rhythm. The results also show how differences between
musicians and non-musicians indicate that musical expertise can in
ï
uence aspects
of music processing, aside from mere perception.
In 2003, Besson and Sch
ö
n reported that the P600 ERP response (a positive
de
ection peaking at around 600 ms post-stimuli) is associated with syntactic
violations in language and music such as grammatical errors and incongruously
ending musical phrases. Whereas increases in the N400 (negative de
ection around
400 ms) ERP are associated with unexpected semantic violations in language, such
as
'
The pizza is too hot to cry
'
(Besson and Sch
ö
n
2003
). The amplitude of the ERP
is relative to the degree of the violation; the more abstract the meaning results in a
potential with higher amplitude.
This research indicates that there is a separate mechanism in the brain for pro-
cessing music, and although the P600 is a slower response than the N400, it
nonetheless provided a basis for further research into applying auditory perception
into controlling music. A dif
culty in using ERPs as a control source in BCMIs is
the issue of identifying potentials amongst non-related EEG information. To
address this, epochs of ERPs are summed and averaged from many presentations of
