Decision Support System for Case Oriented Diagnosis of Diabetic Cats: A Knowledge Based

A Knowledge Based Decision Support System for Case Oriented Diagnosis of Diabetic Cats

BACKGROUND: Diabetes mellitus is an important endocrinopathy affecting pet felines, thus accurate diagnosis is vital for adequate treatment.

GOAL: Computer based Decision Support Systems (DSSs) are targeted on assisting the diagnostic process. Our DSS novelty regards the way it assists clinical and paraclinical diagnosis for diabetes mellitus associated pathologies. Maximization of accuracy and reliability of clinical decisions represents the main motivation.

METHODS: The DSS design considers the syndrome of polyuria-polydipsia, spotting the accompanying pathologies. It uses a knowledge based approach with production rules managing the successive phases, namely anamnesis, clinician’s input, clinical and paraclinical description, and confirmation diagnostic tests.

RESULTS: Having a user friendly interface, the DSS employs a rule-based software environment (CLIPS) and can determine: (a) diabetes mellitus; (b) diabetes mellitus induced by (b.1) hypersomathotropism, (b.2) hyperthyroidism, (b.3) hyperadrenocorticism, (b.4) diabetogenic medication; (c) diabetes mellitus associated with (c.1) chronic kidney failure, (c.2) heart failure; (d) ketoacidotic diabetes mellitus; (e) pancreatitis. The DSS was used with maximum accuracy in a 27 cases study, revealing: (a)–8, (b.1)–2, (b.2)–1, (b.4)–6, (c.1)–3, (c.2)-3, (d)–4.

CONCLUSIONS: Feline diabetes mellitus treatment protocol requires accurate and complete diagnosis, which can be achieved by advanced computational systems, thus reducing possible expensive medical errors and time consuming tests.


Case hystory, clinical and physical examination

Polyuria / Polydipsia


Yes *







Sustained Hyperglycaemia


Underlying pathologies




Increased appetite with weight gain *

Progressive weight loss with increased appetite

Rapid weight loss with decreased appetite





Conclusive clinical and paraclinical data:

Broad facial features, with inferior prognatia;

Clubbed paws

Organomegaly, with main implication of heart and kidney


Insulin resistance >1.5 UI/KG/administration

Fructosamine > 500 µmol/l


Increased Growth Hormone >5µg/l,

Increased Insulin-like Growth Factor-1 >1000 ng/ml,

Confirmation:  Increased pituitary volume observed on MRI contrast enhanced examination



Diabetes mellitus with underlying hypersomathotropism.


The current research designed an expert system which can provide necessary information and diagnosis indications for all types of diabetes and associated pathologies.

The expert system can be highly helpful for the clinicians and veterinary students.

After data acquisition and designing rule-based decisions, the system was developed and tested in a Veterinary Teaching Hospital University of Medical Sciences and final expert system has been presented.

The acquisitive knowledge was represented in the diagrams, charts and tables. The related source code using of the expert system was given and after testing the system, finally its validation has been done. It has been concluded here the expert system can be used effectively in all areas of medical sciences. In particular, in terms of vast number diabetics throughout the world

Therefore, such a. Since this expert system gathers its knowledge from several medical specialists, the system has a broader scope and can be more helpful to the patients — in comparison to just one physician.

The goal was to design a decision support system which could be applied by veterinary clinicians in the management feline diabetes mellitus diagnosis, with part from polyuric-polydipsic. The present expert system is a computer application that simulates the decision making and reasoning of clinician with expert knowledge in veterinary medicine area.

patients and. for physicians to produce more accurate prescriptions for polyuria patients.

In our paper, we first studied the logical rules that were needed to produce a prescription for polyuria patients and represented them in an antecedent/consequent model of rules that were capable of being used in an expert system.

Afterwards we used a powerful expert system tool, CLIPS, to implement our proposed expert system and discussed parts and fundamentals of our expert system that were designed based on the studied rules.

Since our expert system produced its prescription based on strict medical science instructions, our goal was to make the system be able to cooperate with physicians’ experiences on polyuria patients.

Further in our paper, we used the double-blind technique to evaluate our proposed expert system and compared its results with alternatives (physicians).

As the results indicate, the proposed system presents a high accuracy, which proves the correctness of the implementation. The system could also be useful for veterinary medical students as a decision support system.