Knowledge representation formalisms as methods to test pathophysiological hypotheses

After having built and tested an expert system for renal function monitoring, the authors developed a knowledge representation formalism that is being used as a formal method to generate and verify new and unknown pathophysiological hypotheses. In particular, they have studied if and when hypoperfusion determines a renal injury in a patient submitted to cardiac surgery. Using these methods, a description of hypoperfusion was generated and tested sufficiently deep and powerful to determine and distinguish between the hypoperfusive states that produce renal injury and the ones that never generate renal damage.<
>