Recurrent Neural Networks for Predicting Outcomes after Liver Transplantation: Representing Temporal Sequence of Clinical Observations

Summary

Objectives: This paper investigates a version of recurrent neural network with the backpropagation through time (BPTT) algorithm for predicting liver transplant graft failure based on a time series sequence of clinical observations. The objective is to improve upon the current approaches to liver transplant outcome prediction by developing a more complete model that takes into account not only the preoperative risk assessment, but also the early postoperative history.

Methods: A 6-fold cross-validation procedure was used to measure the performance of the networks. The data set was divided into a learning set and a test set by maintaining the same proportion of positive and negative cases in the original set. The effects of network complexity on overfitting were investigated by constructing two types of networks with different numbers of hidden units. For each type of network, 10 individual networks were trained on the learning set and used to form a committee. The performance of the networks was measured exhaustively with respect to both the entire training and test sets.

Results: The networks were capable of learning the time series problem and achieved good performances of 90% correct classification on the learning set and 78% on the test set. The prediction accuracy increases as more information becomes progressively available after the operation with the daily improvement of 10% on the learning set and 5% on the test set.

Conclusions: Recurrent neural networks trained with BPTT algorithm are capable of learning to represent temporal behavior of the time series prediction task. This model is an improvement upon the current model that does not take into account postoperative temporal information.

Keywords

Neural Networks (Computer) - Liver Transplantation - Decision Support Techniques - Algorithms - Monte Carlo Method - Nonlinear Dynamics

• References

• 1 Gershenfeld NA. et al. The future of time series: Learning and understanding. In: Time Series Prediction: Forecasting the Future and Understanding the Past. Gershenfeld NA, Weigend AS. Addison-Wessley: Reading, MA; 1994: 1-70.
• 2 Rumelhart DE. et al. Learning representations by back-propagating errors. Nature 1986; 323: 533-6.
• 3 Ploeg RJ. et al. Risk factors for primary dys-function after liver transplantation – A multivariate analysis. Transplantation 1993; 55: 807-13.
• 4 Strasberg SS. et al. Selecting the donor liver: Risk factors for poor function after orthotopic liver transplantation. Hepatology 1994; 20: 829-38.
• 5 Rosen HR. et al. Significance of early amino-transferase elevation after liver transplantation. Transplantation 1998; 65: 68-72.
• 6 Deschenes M. et al. Early allograft dysfunction after liver transplantation. A definition and predictors of outcome. Transplantation 1998; 66: 302-10.
• 7 Marino IR, Doyle HR. et al. Effect of donor age and sex on the outcome of liver transplantation. Hepatology 1995; 22: 1754-62.
• 8 Doyle HR. et al. Hepatic retransplantation – An analysis of risk factor associated with outcome. Transplantation 1996; 61 (Suppl. 10) 1-10.
• 9 Doyle HR, Marino IR. et al. Early death and retransplantation in adults after orthotopic liver transplantation: Can the outcome be predicted?. Transplantation 1994; 57 (Suppl. 07) 1028-2036.
• 10 Doyle HR. et al. Predicting outcome after liver transplantation: a connectionist approach. Ann Surg 1994; 219: 408-15.
• 11 Parmanto B, Munro PW, Doyle HR. et al. Neural network classifier for hepatoma detection. Proceedings of the Fourth World Congress on Neural Networks. 1994: I285-90.
• 12 Doyle HR, Parmanto B, Munro PW. et al. Building clinical classifiers using incomplete observations – A neural network ensemble for hepatoma detection in patients with cirrhosis. Methods Inf Med 1995; 34: 253-8.
• 13 Williams R, Zipser D. Gradient-based learning algorithms for recurrent networks and their computational complexity. In: Backpropagation: Theory, Architectures, and Applications. Chauvin Y, Rumelhart D. Hillsdale, NJ: Lawrence Erlbaum; 1993: 433-86.
• 14 Caompolucci P. A Circuit Theory Approach to Recurrent Networks Architectures and Learning Methods. Unpublished doctoral dissertation. University of Bologna, Italy 1998 Available at: http://nnsp.eealab.unian.it/Campolucci_P/PhDthesis.htm
• 15 Werbos PJ. Backpropagation through time: what it does and how to do it. Proceedings of IEEE, Special issue on neural networks 1990; 78 (Suppl. 10) 1550-60.
• 16 Elman JL. Finding structure in time. Cognitive Science 1988; 14: 179-212.
• 17 Parmanto B, Munro PW, Doyle HR. Reducing Variance of A Committee Prediction with Resampling Techniques, Connection Science Journal. 1996 8(3/4) 405-27.
• 18 Parmanto B, Munro PW, Doyle HR. Improving Committee Diagnosis with Resampling Techniques. In: Touretzky D, Mozer M, Hasselmo M. (eds.). Advances in Neural Information Processing 8, MIT Press Advances in Neural Information Processing 8, MIT Press; 1996: 882.
• 19 Efron B, Tibshirani RJ. Cross-validation and other estimates of prediction error. In: An Introduction to the Bootstrap. Efron B, Tibshirani RJ. New York: Chapman & Hall; 1993: 237-57.
• 20 Van Bemmel JH, Musen MA. Handbook of Medical Informatics. Houten/Diagem: Springer; 1997: 233-56.