•  
  •  
 

Abstract

Our work focuses on the usefulness of previously stored correct extracted results, which form a sort of stored knowledge got from previous experiences, from enhancing Toulmin's argument model that deals with drug conflict problems in therapeutic diagnostics. New patients are entered using friendly user interface to store in files and then they are matched with the records of previous results, patients’ symptoms and histories datasets which also contain the correct best drugs extracted results. If the new entered record of a patient is matching with any previous record then the correct result of drug will be found immediately and displayed. Otherwise, it will enter for processing by the argument improvement of Toulmin's model that deals with conflicting problems in medicine based on Naive Bayes' theory. The symptoms of each disease are linked to its relevant treatment by using the inference rules which at last give rise to diagnosis of the appropriate treatment. Many competent features of each drug will either support or attack the drug and then a decision is made by employing the Naive Bayes technique based on the features of both the treatment and the patient as extracting results which will be stored to be validated by human experts. Datasets are gathered from some educational hospitals in Iraq and they have been approved by experts from the medical sector. The samples used in the proposed system cover 325 cases with two kinds of diseases and the average percentages of accuracy with them were 93.03% (hypertension) 95.125% (angina pectoris).

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS