Simulation of biomedical systems (e.g. drug metabolism) is an important component of decision support in many medical applications. To cope with the uncertainty of the parameters that are used in the models, one may rely on either applying stochastic methods to quantitative simulation or using qualitative simulation instead, where a number of techniques infer as much as possible in the lack of quantitative information. Many cases, however, rather than lack of quantitative information, are cases of incomplete information, in that numerical bounds can be assigned to the unprecisely known parameters of a model. In such cases, constraint solving techniques might be used to cope with such source of uncertainty and a balance must be sought between the expressive power of the models and the constraint solving capabilities that can be effective applied to these models. This paper presents semi-qualitative modeling and simulation, a new approach aimed at caching such appropriate balance. This new formalism is introduced with a simple example and by means of a more formal presentation, and its application to a multicompartmental model for drug metabolism is discussed.
