Data Mining Applications in Liver Transplantation: A Review and Extension

Document Type : Research Paper


1 Health policy research center, Shiraz University of medical sciences, Shiraz, Iran

2 Department of industrial engineering, Iran University of Science and Technology, Tehran, Iran

3 Department of pediatrics, Shiraz University of medical sciences, Shiraz, Iran


Background and Objective: Liver transplantation is an accepted treatment for patients who present with end-stage liver disease. However, LT is restricted due to a lack of suitable donors, and further on this imbalance between supply and demand leads to death for those who are awaiting in the waiting list. Specialists use MELD to assign the donor organ to the recipient and predicting the survival of patients after liver transplantation, but this index alone will not be suitable for this task and has some weaknesses. Therefore, other indicators need to be selected so that they can have a more appropriate allocation and better predictive power.The optimal allocation in waiting time and predicting the survival of patients after liver transplantation is a problem that might be answered using data mining techniques. The purpose of the present study is to review and compare the different data mining, machine learning, and deep learning techniques in the articles published in this area.
Method: Using relevant keywords, international databases relevant materials were investigated. After limiting the search strategy and deleting the duplications, the rest of the valid papers were screened by examining the title and abstract. In order to increase the sensitivity of the searching procedure, reference lists of papers were also examined. Finally, 42 articles related to the subject of research were selected from 1994 to 2020.
Results: Since this process has some complications for further implementation, the whole process must be created and developed as an LT decision support system which might amend to current HIS or medicine 2.0 as a new module. The latter has to be monitored and further enhanced through feedback given during the development phase
Conclusion:By reviewing the literature, we found that artificial neural network (ANN), ensemble models such as random forest (RF) and Gradient boosting machine (GBM) and their combinations among other data mining models, have shown the best results in the allocation problem and forecasting graft survival after LT.