AUTHOR=Zhao Liang , Ma Jiajun , Shao Yu , Jia Chaoran , Zhao Jingyuan , Yuan Hong 

TITLE=MM-UNet: A multimodality brain tumor segmentation network in MRI images

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.950706

DOI=10.3389/fonc.2022.950706

ISSN=2234-943X

ABSTRACT=The brain tumor is a kind of disease that does great harm to human health. Therefore, the localization and segmentation of brain tumor images have always been an active field of medical research. The traditional manual segmentation method is time-consuming, laborious, and subjective. In addition, the information provided by a single image modality is often limited, which can not meet the needs of clinical application. Therefore, we propose a multi-modality feature fusion network for brain tumor segmentation which adopts a multi-encoder and single-decoder structure. Each encoder independently extracts low-level features from the corresponding modality and connects the hybrid attention block to strengthen the features. After fusion with the high-level semantic of the decoder path through skip connection, the decoder restores the pixel level segmentation results. We evaluate our proposed model on the BraTS 2020 dataset. The experimental results show that the MM-UNet achieves the mean Dice score of 79.2% and mean Hausdorff distance of 8.466 respectively, which is a consistent performance improvement over U-Net, Attention U-Net, and ResUNet baseline and proves the effectiveness of our proposed model.