Biomedical Engineering Reference
In-Depth Information
Fig. 3.7
The performance of
the RLCES model (
star line
),
the CP4 model (
dashed line
)
and the CP5 model (
dotted
line
), against measured data
(
continuous line
) in healthy
subjects
Fig. 3.8
The performance of
the RLCES model (
star line
),
the CP4 model (
dashed line
)
and the CP5 model (
dotted
line
), against measured data
(
continuous line
)inasthmatic
patients
The error values calculated with (
3.12
) for each case are also reported, in terms
of their averaged values and a standard deviation
>
5 %. For those model parame-
ters for which no standard deviation is reported, the standard deviation varied with
<
5%.
In the remainder of this chapter, CP4 and CP5 will denote the constant-phase
model in four parameters from (
3.9
), respectively, the constant-phase model in five
parameters from (
3.10
).
The performance of the models from (
3.9
), (
3.10
) and RLCES on the impedance
complex data is depicted in Figs.
3.7
,
3.8
and
3.9
. It can be observed that these
models characterize sufficiently well the frequency-dependent behavior of the
impedance. It is also clear that the FO model in four parameters from the literature,
givenby(
3.9
), is unable to capture the real part of impedance, which is increasing
with frequency. This model is then only valid in the low frequency range where the
real part of the impedance is decreasing as frequency is increasing.
As observed from the results given in Tables
3.2
,
3.3
,
3.4
,
3.5
,
3.6
,
3.7
and
3.8
,
the viscoelastic model has the poorest performance in terms of total error, explained
by the absence of inductance in the model structure. Within the integer-order mod-
els, Mead's model has the least total error results in all subject groups. Notice that
Extended RLC is a (simplified) special case of Mead's model, and therefore it will
never provide better results. For the case of a healthy subject, peripheral resistance
