In an attempt to quantitate the physical behavior of biological systems, Fourier analysis has been applied to the respiratory and circulatory systems by a number of investigators. The validity of this application has been questioned on the basis that these systems are nonlinear and not strictly periodic. If these objections were valid much of the more recent work in this field would have to be re-evaluated. The applicability of Fourier analysis to these two systems was therefore investigated, both theoretically and experimentally, using on-line analysis on a LINC (laboratory instrument computer) digital computer. In normal anesthetized dogs errors introduced by deviations from periodicity and linearity were found to be within the range of measurement errors. In sinusoidally perfused aortas the amount of second harmonic produced by the vessel was less than 5%. In addition, the magnitude of errors due to faulty determination of cycle length, sampling techniques, aliasing, and A-D (analogue to digital) conversion were evaluated and found to be within the noise level of the measuring equipment when appropriate techniques were employed. Utmost care has to be used in the coupling between a transducer and the system to be measured, and dynamic calibration before each experiment is a prerequisite for successful analysis. With presently available equipment the static measurement errors can be reduced to +/-0.2 cm H(2)O for pressure transducers, 0.1 cm(3)/sec for electromagnetic flowmeters, and 5 x 10(-4) cm for measurement of radius changes. The frequency response of this equipment once properly coupled to the system is flat to at least 20 cycle/sec.
Reglage autom., EPFL, Lausanne/Switzerland INTRGDUCTIONThe dynamic behaviour of physiological systems-is usually determined by applying sinusoidal, step, impulsive or pseudorandom inputs. The input/output responses are then commonly represented by impulse or frequency responses. However, a mathematical representation of minimal dimensionality has less often been attempted. Well known techniques exist for estimating difference equation parameters. Yet the physiologist is rather interested in differential equations for describing a physiological System. The linear relationships between difference and differential equation parameters are only approximative, and unwanted oscillations occasionally appear on the output of a difference equation model. Curve fitting techniques in the time domain can involve lengthy iterations, even for a simple model (e.g. two exponentials). Curve fitting techniques in the frequency domain have been proposed which involved linearization Jjl/j or which considered only the gain characteristic \2J. Therefore, a noniterative method was derived for estimating the parameters and the parameter variances for a differential equation or transfer function of any order. METHGDSThe initial steps of this new method are analogous to those mentioned by Astroem and Eykhoff [3] for the case of difference equation models. The method is based an the minimization of an error function V( n) = ,/q dt where n= System order, q(t)= equation error, and t=time. For a 4th order system:(1) (t) + q(t) Eq. I V(n) is differentiated with respect to the parameters P., where Thus il = 2jq(t) i2-LLi dt =0 Eq. II 1 o P±
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.