Biomedical Engineering Reference
In-Depth Information
Wave peak
Amplitude
Sleep spindle
Awake
Wave
trough
1 cycle
Stage I
REM
1s = 13 cycles
Stage II
Stage III
Stage IV
FIGURE 20.2
EEG patterns in humans during wake and different sleep stages. (Data from Pace-Schott, E.F.
and Hobson, J.A.,
Nat. Rev. Neurosci
., 3, 591, 2002.)
21,000
17,000
14,000
10,500
7,000
3,500
0
0
1
2
3
4
5
6
7
8
9
4
3
2
1
W
REM
0
1
2
3
4
5
6
7
8
9
Elapsed time (h)
FIGURE 20.3
SWS and SWA. The bottom panel illustrates a hypnogram showing the amount of time spent
in the different sleep stages over the course of the night, using traditional visual scoring criteria. Each progres-
sion through the i ve stages (stages 1-4 and REM) constitutes a sleep cycle. The top panel represents a fast
Fourier transformation of EEG data. The frequency band selected for power analysis was 0.75-4.5 Hz. This
representation illustrates that SWA is present throughout the night, even when the sleeper does not reach stage
3 or 4 according to traditional visual scoring criteria. (From Ebert, B. et al.,
Pharmacol. Ther
., 112, 612, 2006.
With permission.)
light sleep, while stages 3 and 4 are often described as deep or slow wave sleep (SWS). The hallmark
waveform of SWS consists of rhythmic, low frequency waves (~0.5-4.5 Hz) with large amplitude.
The amount of slow waves can be quantii ed by a Fourier transformation, whereby any pattern
can be described by a combination of a series of sinusoidal waves with different frequencies and
amplitudes. The numeric amplitude squared is a rel ection of the amount of a certain frequency and
is termed power. The power spectrum for slow waves (slow wave activity or SWA) during NREM
sleep (Figure 20.3 top) is particular strong during stages 3 and 4, but are present throughout the
sleep period, although to a much smaller extent.
