Research Communications of the 24th ECVIM-CA Congress

The design of our feline diabetes mellitus DSS emerges from the syndrome of polyuria-polydipsia, with the possibility of spotting the accompanying pathologies. Fuzzy logic is used for dealing with knowledge representation and uncertainty. The fuzzy rules proposed to represent this knowledge emerge from anamnesis, clinician's input, clinical and paraclinical description, and confirmation diagnostic tests. Clinical signs such as polyuriapolydipsia, persistent hyperglycemia, polyphagia, weight fluctuations, administration of drugs with a diabetogenic potential, were considered decisive in the pattern of diagnosis establishment. Registered medical records of 29 cats, 16 males and 13 females, whit ages from 7 to 18 years old, were analyzed in order to validate the DSS. Using Matlab software, the DSS was implemented and tested. For any case with polyuria-polydipsia the system provides, via a friendly graphical user interface, the diagnosis with the highest probability. The set of diagnoses which can be generated by the DSS consists in: a) diabetes mellitus; b) diabetes mellitus induced by (b.1) hypersomathotropism, (b.2) hyperthyroidism, (b.3) hyperadrenocorticism and (b.4) diabetogenic medication; c) diabetes mellitus in association with (c.1) chronic kidney failure and (c.2) heart failure; d) ketoacidodic diabetes mellitus; e) pancreatitis. The DSS was applied with success on all 29 cases, revealing the following diagnoses / no of cases: (a) -8, (b.1) -2, (b.2) -1, (b.4) -8, (c.1) -3, (c.2) -3, (d) -4.