Expert System Based on Rules and Medical Knowledge for the Medical Diagnosis of Typhoid Fever (XperTyph)

Cuteso Matumueni, Humberto and Neto Simbo, Alcides Romualdo (2019) Expert System Based on Rules and Medical Knowledge for the Medical Diagnosis of Typhoid Fever (XperTyph). Asian Journal of Research in Computer Science, 4 (2). pp. 1-14. ISSN 2581-8260

[thumbnail of Matumueni422019AJRCOS52431.pdf] Text
Matumueni422019AJRCOS52431.pdf - Published Version

Download (530kB)

Abstract

Today, common diseases like malaria, typhoid, cholera and others are becoming more dangerous problems for people living in this world. It is simple to use, portable, inexpensive and makes the diagnosis of typhoid faster, more efficient and more accurate, and helps doctors avoid wasting time on the diagnosis of typhoid and the publication of results. The objective is to know how to avoid the queue of patients in the hospital. In this article, the author proposed a model of expert systems using the knowledge of physicians and other health professionals. The expert system Xper Typh is useful for patients who suffer from typhoid fever. This system will give a response similar to that of a doctor or a medical expert. This system is also very useful in rural areas where we have young medical experts or no medical experts. Study cross-sectional between the Cabinda Provincial Hospital and the Center 1st of May Health in Cabinda during June 2015 and July 2018. We included 65 patients (22 men, 28 women and 15 children); age group from 5 to 75 years old. The patients examined presented different symptoms and signs. On 65 patients, 20 patients with confirmed fever, 25 patients without confirmed fever and 20 contraction patients between the expert system and the physician. The results between the doctor and the expert system are as follows: Sixty-five patients were diagnosed with XpertTyph. The XpertTyph correctly identified 20/65 (30%) of patients with typhoid fever and 25/65 (41%) of patients not suffering from typhoid fever. That is 71% concordance and 29% discrepancy. The criterion of validity of the model presented the best content for a certain level of average diagnosis for the diagnosis (k = 0.21 to 0.40) for typhoid fever which corresponds to 69% of confidence.

Item Type: Article
Subjects: Eprints STM archive > Computer Science
Depositing User: Unnamed user with email admin@eprints.stmarchive
Date Deposited: 11 Apr 2023 04:27
Last Modified: 02 Jan 2024 13:06
URI: http://public.paper4promo.com/id/eprint/26

Actions (login required)

View Item
View Item