Abstract
The cirrhosis disease represents the final stage of hepatitis, characterized by the death of liver cells and irreversible liver damage. Although there are some methods used in the prediction of cirrhosis, especially those utilizing various artificial intelligence techniques, it is still difficult to accurately predict cirrhosis. The aim of this research is to detect cirrhosis by focusing on deep learning methods. In addition to analyzing the performance of deep learning methods for cirrhosis prediction, the study also compares the performance of traditional machine learning algorithms with deep learning techniques. Decision Tree (DT), k-Nearest Neighbors (kNN), Random Forest (RF) and Logistic Regression (LR) algorithms are used in order to achieve these goals. Considering the relatively lower performance of some of these algorithms, Deep Neural Networks performed the classification accurately. In the dataset used in the study, there were 362 patients with cirrhosis and 1023 without cirrhosis. Model performance showed that deep neural networks achieved high classification performance with metrics such as 95.96% accuracy. According to the results, deep learning methods showed strong performance, providing high accuracy and sensitivity for cirrhosis prediction alongside traditional machine learning methods.