Page 18 - Control Theory in Biomedical Engineering

Modeling and control in physiology 7

Fig. 2 Basic compartment structure. (The figure has been modified from Brown, R.F., 1980.
Compartmental system analysis: state of the art. IEEE Trans. Biomed. Eng.)

physiological measurements have their analog physical variables (Aidley,
1998). Table 1 describes some of these.
In neurophysiology and cell physiology the use of equivalent circuits is
well known, where the classic examples are circuits simulating the genera-
tion of membrane potential (Aidley, 1998), the cable properties of nerve and
muscle fibers, and action potential and neuromuscular transmission. Using
analogue circuits to model and simulate physiological systems is also consid-
ered as a modern approach for teaching physiology to postgraduate medical
students. The aim is to promote qualitative as well as quantitative analogue
c
thinking about physiological processes (Rupnik et al., 2001; Ribari and
Kordas ˇ, 2011; Sever et al., 2014).
In the last decades, particular physiological systems were modeled by
equivalent electronic circuits such a pulmonary ventilation (Ghafarian
et al., 2016) and blood circulation (Ismail et al., 2018). In Ismail et al.
(2018), for example, to understand the cardiovascular system, the thermo-
dynamics of vessels was represented by an electronic circuit as described in
Fig. 3. In this circuit, the voltage represents the blood pressure when the
current represents the blood flow. C and R are the compliance and the resis-
tance of the systemic arterial tree, respectively. L is the impedance of
proximal aorta.
For the use of mechanical equivalent models, several systems are also pro-
posed. For example, the reader can found a pneumatic model given for the
respiratory system (Shi et al., 2016) and a mass-spring-damper inverted pen-
dulum model for a bipedal-compliant walking system ( Joe and Oh, 2019).
The related model is given in Fig. 4.

2.2.3 Data-driven modeling approach
Data-driven modeling (DDM) approach is an empirical approach that does
not involve mathematical equations derived from physical processes but
instead involves analysis of time series data. Examples include linear