A MODEL FOR DIAGNOSING LIVER DISEASES USING “MACHINE LEARNING” TECHNIQUES

Background & Aims:

 Early diagnosis of liver diseases has a significant effect on the prevention of its complications as well as control and treatment of the disease. “Machine learning” is one of the branches of artificial intelligence that has many applications in the field of medical diagnosis. This study aimed to provide a model with high accuracy and reliability for diagnosing liver diseases using machine learning methods that can help physicians in the early diagnosis and control of liver diseases.

Materials & Methods:

 This applied-developmental study used the dataset of 583 liver patients. In order to more accurately diagnose of the people with liver diseases, the results of the three classifiers including: Random Forest, Support Vector Machine, and Artificial Neural Network were combined using Dempster-Shafer theory. Weka data mining tool and Python programming language were used to implement the model. The k-fold cross-validation method was applied to evaluate efficiency of the model.

The results showed that accuracy, sensitivity, and specificity in the proposed model were 91.47%, 89.52%, and 93.03%, respectively, which had a better performance than similar studies.

The proposed model in the studied statistical population has a better performance in diagnosing liver diseases, and can help physicians in early diagnosis of the disease and appropriate treatment of it in the early stages of it and thus prevent development of the disease.