Model-Based Analysis of Physiological Systems

Abstract

Physiological modeling involves the development of mathematical, electrical, chemical, or other analogs whose behavior closely approximates the behavior of a particular physiological system. If the model components correspond to features of the physiological system under consideration, then we call it a biophysical model. In most physiological systems, only a few of the features are observable, and it is often only possible to construct a model based on empirical relations between these observable features. Such models are termed black box models since they make no attempt to describe the internal mechanisms of these systems. The earliest models of physiological system were physical analogies. Analogies extend intuitive knowledge in one area to another. Building physical models is cumbersome and expensive in material cost. In mathematical models, physical quantities represented by mathematical functions are related to each other by algebraic equations, and changes in these functions over time and space resemble that of the modeled physiological process. Mathematical models are limited only by our ability to solve sets of equations, but this limitation is usually far less confining than the physical construction of analogous models. A model is built using experimental observations, and the model is validated by making testable predictions. Once a model is validated, it can enable us to conduct virtual experiments.

Sometimes a computing machine does do something rather weird that we hadn’t expected. In principle one could have predicted it, but in practice it’s usually too much trouble.

– Alan Turing