Prediction of hepatic encephalopathy complication in liver transplant patients using support vector machine algorithm in active middle-aged women

Abstract

Background and Aim: In liver transplant patients, the occurrence of postoperative complications increases the length of hospitalization, care of patients and the costs of treatment. The aim of this study was to predict the complications of hepatic encephalopathy in liver transplant patients using the support vector machine (SVM) algorithm in active middle-aged women. Methods: The statistical population included 652 patients, among them, 165 active middle-aged women with encephalopathy symptoms who underwent liver transplantation during 2011-2022 were included. SVM algorithm was used to predict the complications of hepatic encephalopathy in liver transplant patients and MATLAB software was used for data analysis. Results: Using 14 features related to laboratory, anthropometry and lifestyle data, the SVM algorithm can predict people with and without hepatic encephalopathy complications with 81.2% accuracy and 74.6% precision. Conclusion: According to the accuracy of the SVM algorithm on the data, it seems that this system may help physicians predict the risk of hepatic encephalopathy complications after transplantation with high accuracy and the lowest cost. Computer-based decision support systems can reduce poor clinical decisions and also minimize costs associated with unnecessary clinical trials.

Keywords

• liver transplantation
• encephalopathy
• support vector machine algorithm
• middle-aged wome