Biomedical Engineering Reference
In-Depth Information
versely with the resistance, or, directly with the volume. The values in COPD and
KS were lower, indicating lower volumes, which corresponds to the clinical pathol-
ogy. In children, the values of healthy and asthma were again close to each other,
while in cystic fibrosis patients the values were higher (i.e. higher values). We con-
clude that the values for the CF group might be subject to the weight and height
biometric parameters.
One of the limitations of the study is that the groups of subjects and patients were
not equally balanced in terms of male/female distribution, hence dependency with
gender was not determined. Similarly, dependency may exist for the adult groups in
terms of age. The healthy adult group had an averaged age value significantly lower
than those diagnosed with COPD and KS. The dependency with height and weight
was not consistently observed in all groups, hence we could not conclude whether
or not these biometric parameters will influence our results. However, even if this is
the case, the seminal ideas presented in this work still hold, in the sense that there
exists a link between changes in the structure and dynamic patterns in the breathing.
We have also concluded that the PV loop provides better results than PPP loop in
terms of separation of the groups. However, the PV loop requires the recording of
two signals: pressure and flow, or similarly, pressure, and volume. In the context of
FOT, there is a 25-fold difference in the cost of pressure sensors and flow sensors.
Therefore, measuring only one signal may be an interesting approach to support the
use of PPP plots instead of PV plots. The advantage would be that if the pressure
is available (i.e. the cheapest measurement), then the delayed derivative of pressure
can be used to plot the PPP.
From a clinical standpoint, it is clear that one of the proposed parameters (i.e.
A
)
is related to the resistive components of the breathing dynamics as extracted from
the PPP loops. However, we do not yet have a parameter which characterizes the
elastic components. Also, there is no information as to how the inhomogeneities in
the lung affect the results of the PPP loops. We conclude that in order to provide a
concise interpretation and mapping the short-term breathing dynamics by means of
PPP plots, a bigger database of patients should be analyzed.
8.4 Summary
Following the theoretical basis laid in the first chapters of this topic and the fre-
quency domain identification from previous chapter, it was only natural to apply
our knowledge in the time domain. First, the link between fractional order paramet-
ric models and time response has been achieved by calculating the impulse response
of the respiratory system. It was shown that this response varies in healthy volun-
teers and in patients diagnosed with breathing impairment. Next, we employed the
notions of fractal dimension and pseudo-phase plot and correlate them to the dy-
namics of the breathing. Again, the link to power-law and implicit fractal dynamics
has been made and results were analyzed over several groups of patients. All these
investigations have been performed in the context of linear systems. The next chap-
ter in this topic will tackle the problem from the nonlinear dynamics point of view
in order to excerpt new information from the system.
