Fully-automatic brain tumor segmentation based on effective receptive field and attention fusion mechanism(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2024年第5期
- Page:
- 563-570
- Research Field:
- 医学影像物理
- Publishing date:
Info
- Title:
- Fully-automatic brain tumor segmentation based on effective receptive field and attention fusion mechanism
- Author(s):
- ZOU Xiang1; WANG Yu1; XIAO Hongbing1; YANG Di2
- 1. School of Computer and Artificial Intelligence, Beijing Technology and Business University, Beijing 100048, China 2. China CAMC Engineering Co., Ltd., Beijing 100048, China
- Keywords:
- Keywords: brain tumor segmentation EAU-Net effective receptive field expansion block attention fusion module inverted residual structure
- PACS:
- R318;TP181
- DOI:
- DOI:10.3969/j.issn.1005-202X.2024.05.006
- Abstract:
- Abstract: Despite significant achievements of deep learning in medical image segmentation, there are challenges for brain tumor segmentation using deep learning, such as insufficient receptive field, excessive redundant information, and information loss. To address these issues, a novel brain tumor segmentation network model (EAU-Net) is proposed based on encoder-decoder structure. EAU-Net incorporates an effective receptive field expansion block and an attention fusion module to minimize the adverse effects caused by insufficient receptive field and excessive redundant information which often occurred in the current brain tumor segmentation network. Additionally, a bottleneck resampling module based on inverted residual structure is introduced to effectively avoid information loss during upsampling and downsampling, while deep convolutions are used to reduce computational complexity. Experimental results on the BraTS2020 dataset reveal that EAU-Net achieves the highest segmentation accuracy, demonstrating its feasibility and effectiveness for brain tumor segmentation.
Last Update: 2024-05-24