A Machine Learning Based Clinical Decision Support System for Diagnosis and Treatment of Typhoid Fever

Oguntimilehin, A and Adetunmbi, A.O and Olatunji, K.A. (2014) A Machine Learning Based Clinical Decision Support System for Diagnosis and Treatment of Typhoid Fever. A Machine Learning Based Clinical Decision Support System for Diagnosis and Treatment of Typhoid Fever, 4 (6). pp. 1-9. ISSN 2277 128X

Text V4I6-0359.pdf - Published Version Download (604kB)

Abstract

Health challenges of the world increase daily while medical practitioners and medical facilities ratio to the increasing population is nothing to write home about. As revealed by World Health Organization (WHO), the world is in need of additional over four million medical practitioners. The few available medical facilities and medical personnel are concentrated in the urban centers which tend to make the situation worst in the rural areas. These among other challenges in the health sector make computer based diagnosis systems desirable. Typhoid fever otherwise known as Enteric fever is a trauma to most developing countries of the world with prevalent cases in Africa. It is on the record that more than six hundred thousand deaths occur annually as a result of typhoid fever. This number is very high due to many factors which include insufficient medical facilities, insufficient medical personnel, poor diagnosis and treatment. In this work, a new diagnosis and treatment system was developed to handle typhoid fever cases. A promising machine learning technique-decision tree algorithm was used on labeled set of typhoid fever conditional variables to generate a decision tree and classifiers for the diagnosis of typhoid fever and treatments were provided according to the level of severity of the disease. The accuracy of the system was measured on both the training set and testing set with the detection rates of 100% and 95% respectively. The system was implemented using Visual Basic as front end and MySQL as backend.

Actions (login required)

View Item