Page 176 - Computational Modeling in Biomedical Engineering and Medical Physics

Bioimpedance methods 165

Figure 5.17 The hemodynamic flow and the EMF in the BCVI (left), non-dimensional time deriva-
tive of the BCVI (right).

5.7 Some comments on numerical modeling results
The values of the cardiovascular found through numerical modeling may differ from
those of their counterparts measured using the ECM because it seems reasonable to
expect discrepancies when using impedance measurements on different arterial
branches, situated farther to the heart.
Details such as the A-wave, revealed experimentally and numerically by the ECM,
Fig. 5.11C, are not identified in the numerical simulation of the BCVI. It remains to inves-
tigate experimentally if this is a limitation concerning the BCVI with respect to the ECM.
As expected, the usage of non-deformable computational domains discards the
direct influence of “artifacts” in the bioimpedance signal that are produced by the dis-
placements caused by factors such as respiration and heart mechanics. It is not possible
to correlate it with the respiration. The “good” part is that Z(t) does not need specific
filtering, signal-processing. In what concerns the heart mechanics, the inlet velocity
profiles used in the numerical modeling account for their effects upon the blood flow,
therefore Z(t) numerical simulation results are close to experimental data.
On closing this discussion, a major problem in passing from idealized laboratory
experiments performed to characterize the blood electrical conduction or from analytic
solutions to the real conditions of the EBI procedure is their consistency with the
computational domains. Anatomically realistic computational domains used to com-
pute arterial flows may raise concerns in using experimental results on the electrical
conductivity of the arterial blood. This difficulty may be overcome by using an equiv-
alent electrical conductivity of the arterial blood (function of flow rate), based on
known analytic and experimental results. Several nonlinear electrophoresis effects (e.g.,
the delayed, hysteretic change of the electrical conductivity of blood for decelerating
and accelerating flows), not addressed here, are approachable too, but—within the