•  
  •  
 

Abstract

Liver transplantation is the last life-saving solution for patients with end stage liver disease. The discrepancy between waiting list and available organs has led to the appearance of extended donation criteria and the development of several scores (Child-Pugh score, MELD score, DRI score, SOFT score), in order to find the most suitable donor-recipient match. But none of these scores can predict survival after transplantation. Artificial Intelligence (AI) has recently been shown as an excellent tool for the study of the liver and comes in this aid with its various methods (random forest, artificial neural networks, decision tree, Bayesian networks, and support vector machine). Materials and Methods. By reviewing the literature (mostly retrospective multicenter studies), we aimed to establish if the AI is a proper or even a more accurate method of predicting posttransplant survival, in comparison with the existing linear statistical models. Results. Machine learning showed better results than several current scoring systems that use either isolated donor/recipient scores or combined donor/recipient factors. The advantages of this model are its capacity for analyzing both linear and nonlinear relationships between features and outcomes, its robustness of overfitting by design, and built-in insights into feature importance aiding model explainability. Nevertheless, machine learning has its limitations because it requires large amounts of data, which can be difficult to obtain, it also requires high levels of technical skill, can be difficult to engineer and it’s expensive. Conclusion. AI may have significant potential in aiding clinical decision-making during liver transplantation, including donor-recipient matching.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Download

Included in

Surgery Commons

Share

COinS