Biomedical Engineering Reference
In-Depth Information
Table 3. Pseudo-code of threshold determination
/*Program: Threshold determination*/
CONST:
G(N,E);
VAR:
Threshold[i,j : 1…4n] : real;
Action:
for
(
i
:
=
1
i
≤
4
n
;
i
+
+
)
for
(
j
:
=
1
j
≤
4
n
;
j
+
+
)
if
(((
i
,
j
)
∈
E
)
∧
(
i
<
j
))
begin
threshold[i,j] := 1.0;
threshold[j,i] := 0.0;
end
else
begin
threshold[i,j] := 0.0;
threshold[j,i] := 0.0;
end
end
end
end
DISCUSSIONS AND PROSPECT
Model Assessment
We present a general approach from the macro-
scopic point of view, based on large-scale, paral-
lel
computation exercised on legged biological
specimens. The significance of this macroscopic
method and related model is twofold. First, it is
possible to apply the method and related model
directly on retrieving single gait of an animal
independently. Second, the method emphasizes on
providing a general, whole spectrum simulation of
any gait types without limitations of continuous
mathematical
models. In addition, the method
is amenable to circuit implementation due to its
digital and scalar nature. The sole presumption
is that the spatio-temporal structure of an object
is known.
The prominent characteristic of the asymmetric
Hopfield network under SMER is that every bool-
ean input of a macroneuron can influence its output
because of the multiplication operation between
input and output of the neuron. The model has
been shown to be able to retrieve all gait patterns
proposed by Golubitsky et al.'s general pattern
generation theory through calculation of large-
scale matrices in discrete equations environment.
A
mathematical description to construct the oscil-
latory building blocks is provided, together with
a detailed method
on how to design
a
rhythmic
system and its critical parameter matrices. The
main contributions of this work can be sum-
marised as follows,
1.
A general neural network
model embed-
ding the SMER algorithm and capable of
Search WWH ::
Custom Search