3D convolutional neural networks for brain tumor segmentation

This paper presents our work on applying 3D Convolutional Networks for brain tumor segmentation for the BRATS challenge. We are currently experimenting with different 3D fully convolutional architectures. We present preliminary results using these architectures and outline our future steps and experiments, which involve hyperparameter optimization, comparison of the models’ performance and implementation of a post-processing stage to eliminate false positive predictions.