The model of health status diagnosis and evaluation of critical patients with the method of artificial neural network

This paper presents a model for diagnosis and evaluation of ICU patient health status based on artificial neural network. It takes the physiological parameters and seriousness grading of APACHE II scoring system as the sample data. Trainings and tests with the method of back propagation neural network classification algorithm are made to the neural network, and the result indicates that the accuracy of the model is over 85%. Contrast to the traditional methods, this model shortens the time to build a health status evaluation system of ICU and decreases the sample size demanded for the evaluation system.