Enhancing MRI Brain Tumor Segmentation with an Additional Classification Network

Brain tumor segmentation plays an essential role in medical image analysis. In recent studies, deep convolution neural networks (DCNNs) are extremely powerful to tackle tasks. We propose this paper a novel training method that enhances the results by adding additional classification branch network. The whole network was trained end-to-end on Multimodal Tumor Segmentation Challenge (BraTS) 2020 dataset. On BraTS’s test set, it achieved average Dice score of $$80.57\%$$ , $$85.67\%$$ and $$82.00\%$$ as well Hausdorff distances $$(95\%)$$ 14.22, 7.36 23.27, respectively for enhancing tumor, core.