Page 24 - Control Theory in Biomedical Engineering
P. 24

12    Control theory in biomedical engineering


          many aspects of identifiability are studied, like the role of initial conditions
          (Pia Saccomani et al., 2003) and optimal design of clinical tests (Bezzo and
          Galvanin, 2018). Different application papers of structural identifiability
          have been published (Xia and Moog, 2003; Raue et al., 2009; Miao
          et al., 2011; Eberle and Ament, 2012; Tuncer et al., 2016; Pironet et al.,
          2019). A software tool to test global identifiability of biological and physi-
          ological system is described in Bellu et al. (2007).

          2.5 Practical identifiability

          Practical identifiability concerns parameter and state estimation processes via
          online or offline techniques. Since the 1970s, great efforts have been made to
          describe physiological systems in explicit mathematical models for which
          several online or offline techniques for parameter estimation are developed
          and the entire procedure of estimation, from model formulation to
          computer selection, is re-examined (Rideout and Beneken, 1975). Nowa-
          days, parameter estimation techniques are of ever-increasing interest in
          the fields of medicine and biology for which few books and book
          chapters (Marmarelis and Marmarelis, 1978; Khoo, 1999; Westwick and
          Kearney, 2003; Heldt et al., 2013; Ho, 2019), survey papers (De Nicolao
          et al., 1997; Giannakis and Serpedin, 2001), and application papers
          (Tong, 1976; Misgeld et al., 2016) are available. A software package is also
          available that solves structural/practical identifiability problems (Galvanin
          et al., 2013), as described in Fig. 6. This framework first conducts a thorough
          analysis to identify and classify the nonidentifiable parameters and provides a
          guideline for solving them. If no feasible solution can be found, the frame-
          work instead initializes the filtering technique prior to yield a unique
          solution.


          2.6 Application examples
          There are several well-known examples of mathematical models in physio-
          logical systems, such as the endocrine system, immune system and cardio-
          vascular system. In this subsection, we describe the simplest models of
          these renowned systems and give an extensive bibliography.


          2.6.1 The endocrine system models
          The endocrine system is the set of glands in the body that produce hormones
          directly into the circulatory system in order to regulate physiological and
          behavioral activities (Neave, 2008). Hormones are used to communicate
          between organs and tissues. The endocrine process is primordial for the
   19   20   21   22   23   24   25   26   27   28   29