Use of artificial intelligence and simple mathematics to analyze a physiological model
John Kunz · 1984
The objective of this research is to demonstrate a methodology for design and use of a physiological model in a computer program that suggests medical decisions. This methodology uses a physiological model based on first principles and facts of physiology and anatomy. The model includes inference rules for analysis of causal relations between physiological events. The model is used to analyze physiological behavior, identify the effects of abnormalities, identify appropriate therapies, and predict the results of therapy. This methodology integrates heuristic knowledge traditionally used in artificial intelligence programs with mathematical knowledge traditionally used in mathematical modeling programs. A vocabulary for representing a physiological model is proposed. Analysis and explanation of physiological function is based ultimately on causal relations. This project distinguishes between two kinds of causal relations. Type-1 causal relations are empirical, based on definitions or on repeated observation. Type-2 causal relations have a basis in physical law, represented mathematically. Inference rules are proposed for making valid qualitative causal arguments with both kinds of causal basis. Such analysis allows qualitative causal inferences to be based ultimately either on empirical observation or on physical laws. In both cases, causal effects are considered to propagate along an anatomical network through which physiological processes can function. A knowledge base in the field of renal physiology has been built. A computer program analyzes this knowledge base and demonstrates the methodology. The program shows that a computer system exploiting physiological knowledge is able to make decisions accurately. The proposed methodology has general applicability to problems which can be analyzed from first principles and basic facts of structure and function. Results of this research suggest that there is more power for problem solving in a methodology exploiting both domain-specific heuristics and basic principles than there is in methodologies emphasizing one of these kinds of knowledge to the exclusion of the other. Further, research results suggest that artificial intelligence techniques provide an appropriate framework to which simple mathematical analysis can be added effectively.