Lung motion prediction by static neural networks

The paper presents a study and comparison of static feedforward neural network performance for prediction of lung motion. A feedforward neural network with local and global optimization to predict human lung respiration is presented. Applicability of the Levenberg-Marquardt algorithm and the backpropagation learning rule during the batch training are discussed. Sliding window learning for retraining static neural network is applied as a more efficient learning prediction method. Prediction results are presented and compared to demonstrate the effectiveness of the applied neural network method.