Medical Image Segmentation Using Squeeze-and-Expansion Transformers

20 May 2021  ·  Shaohua Li, Xiuchao Sui, Xiangde Luo, Xinxing Xu, Yong liu, Rick Goh ·

Medical image segmentation is important for computer-aided diagnosis. Good segmentation demands the model to see the big picture and fine details simultaneously, i.e., to learn image features that incorporate large context while keep high spatial resolutions... To approach this goal, the most widely used methods -- U-Net and variants, extract and fuse multi-scale features. However, the fused features still have small "effective receptive fields" with a focus on local image cues, limiting their performance. In this work, we propose Segtran, an alternative segmentation framework based on transformers, which have unlimited "effective receptive fields" even at high feature resolutions. The core of Segtran is a novel Squeeze-and-Expansion transformer: a squeezed attention block regularizes the self attention of transformers, and an expansion block learns diversified representations. Additionally, we propose a new positional encoding scheme for transformers, imposing a continuity inductive bias for images. Experiments were performed on 2D and 3D medical image segmentation tasks: optic disc/cup segmentation in fundus images (REFUGE'20 challenge), polyp segmentation in colonoscopy images, and brain tumor segmentation in MRI scans (BraTS'19 challenge). Compared with representative existing methods, Segtran consistently achieved the highest segmentation accuracy, and exhibited good cross-domain generalization capabilities. The source code of Segtran is released at https://github.com/askerlee/segtran. read more

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Brain Tumor Segmentation BRATS 2019 Bag of tricks ET 0.729 # 2
WT 0.895 # 1
TC 0.802 # 3
Brain Tumor Segmentation BRATS 2019 Segtran (i3d) TC 0.817 # 1
Avg. 0.817 # 1
Brain Tumor Segmentation BRATS 2019 Extension of nnU-Net ET 0.740 # 1
WT 0.894 # 2
TC 0.807 # 2
Avg. 0.812 # 2
Optic Cup Segmentation REFUGE Challenge Segtran (EfficientNet-B4) Dice 0.872 # 2
Optic Disc Segmentation REFUGE Challenge Segtran (EfficientNet-B4) Dice 0.961 # 1

Methods