Biomedical Engineering Reference
In-Depth Information
LAD
ligature
pH-56
pH-53
Ischemic
region
K-52
K-47
(a)
10
7.50
pH-53
9
7.25
pH-56
8
7
7.00
6
5
K-47
6.75
K-52
4
Occlusion
12
6.50
3
0
2
4
6
8
10
(b)
FIGURE 4.15
Two pH and two K
planar sensors in the
in-situ
porcine heart preparation at midmyocar-
dial depth (a) and the recorded fall in the pH and increase in K
activity, respectively, during the course of
coronary artery occlusion (b). (From [18].)
Time (Min)
to the analytical interpretation [105]. Furthermore, when no highly selective electrodes
are available, this approach allows one to use only sparingly selective electrodes.
A number of mathematical methods for multivariate and multi-way analysis of sen-
sor array complex response are available nowadays, including various regression and
pattern recognition techniques. Multivariate regression methods based on classical sta-
tistics usually matched against more sophisticated “black box” methods as fuzzy logic
and artifi cial neural networks. In the latter the signal processing is often performed in
a non-transparent way, hindering critical evaluation of the results. However, all these
methods, while quite diverse, are intended to provide higher signal-to-noise ratio for
Search WWH ::
Custom Search