Biomedical Engineering Reference
In-Depth Information
50
Region meeting
suppression
criteria (< 5
μ
V,
> 0.5 sec)
0
−
50
0
2
4
6
8
10
12
14
16
Time (s)
Figure 9.4
The BSR algorithm [42, 43].
able (nonstationary) nature of burst suppression, the BSR should be averaged over
at least 60 seconds.
At present there are about 40 publications referring to the use of burst suppres-
sion in EEG monitoring during anesthesia or critical care.
9.5
Frequency-Domain EEG Algorithms
Like all complex time-varying voltage waveforms, EEGs can be viewed as many
simple, harmonically related sine waves superimposed on each other. An important
alternative approach to time-domain analysis examines signal activity as a function
of frequency. So-called frequency-domain analysis has evolved from the study of
simple sine and cosine waves by Jean Baptiste Joseph Fourier in 1822. Fourier anal-
ysis is covered in detail in Chapter 3. Here we concentrate on its applications.
9.5.1 Fast Fourier Transform
The original integral-based approach to computing a Fourier transform is
computationally laborious, even for a computer. In 1965, Cooley and Tukey pub-
lished an algorithm for efficient computation of Fourier series from digitized data
[44]. This algorithm is known as the fast Fourier transform. More information
about the implementation of FFT algorithms can be found in the text by Brigham
[45] or in any current text on digital signal processing. Whereas the original calcula-
tion of the discrete Fourier transform of a sequence of
N
data points requires
N
2
complex multiplications (a relatively time-consuming operation for a microproces-
sor), the FFT requires only
N
(log2
N
)/2 complex multiplications. When the number
of points is large, the difference in computation time is significant, for example, if
N
=
1,024, the FFT is faster by a factor of about 200.
Search WWH ::
Custom Search