A new technique for survival prediction in trauma care using a neural network

Artificial neural networks have been proven very effective at extrapolation and prediction. They are able to fit surface features to data points more closely than statistical regression. For these reasons, an artificial neural network was chosen as the basis for a new survival prediction system for traumatic injury in both children and adults. Traditional mathematical models for survival prediction in trauma care (TRISS and ASCOT) were compared to a new system based upon a multilayer perceptron neural network