Stahl672 bmp

196 WALTER R. STAHL

Function and regulation in the myocardium are the subject of several control theory models. McLeod and Defares (1963) used an analog model to simulate Starling’s Law of the heart, by which the heart responds homeostatically to varying inputs of blood. A digital Computer model has been chosen by Reinboldt et al. (1963) to represent spread of an electrical excitation wave through the myocardium. The system provides a rather convincing imitation of electrocardiographic waveform genera-tion and allows artificial study of damage (infraction) as it infłuences the electrocardiogram (ECG).

B. Respiratory Control Models

Regulatory functions in the mammalian respiratory system have been studied at least as intensively as those of the cardiovascular system. A recent volume (Nahas et al., 1963) reviews the entire problem and pre-sents a variety of partial models, ranging from neuronal to purely phe-nomenological ones. Grodins and James (1963) present a detailed electro-analog Simulator which includes inputs from a variety of chemoreceptors and other transducers. The approach of Grodins has been extended by Defares et al. (1964), who use a complex analog Computer design, based on a dozen or morę basie physiological variables, and they have been able to represent responses of the entire respiratory system to changing carbon dioxide levels in a convincing manner. This type of model can be compared with the prototype both on the basis of numerical invariants pertaining to “control performance,” relational criteria showing “what is connected to what,” and qualitative reactions to various changes of inputs.

Related analog methods are presented in a detailed report by Milhorn and Guyton (1965), who direct their attention to the semipathological control response known as Cheyne-Stokes breathing. In several reports Horgan and Lange (1962, 1964) describe both analog and digital Computer models of respiratory regulation which incorporate multiple Chemical and physical transducer inputs. It is found that tissue phe-nomena, such as carbon dioxide diffusion in the brain and cerebrospinal fluid, must be simulated in the attempt to produce a complete model. Clegg and associates (1964) have also undertaken an analog simulation of the entire respiratory system, with representation of brain and pe-ripheral receptors, physical lung phenomena, metabolism, circulatory effects, etc. Insufficient information is presently available on the numerical value of many contro ller constants in silu to assure that quantitative