Abstract:Purpose Automatic segmentation of colon, small intestine, and duodenum is a challenging task because of the great variability in the scale of the target organs. Multi-scale features are the key to alleviating this problem. Previous works focused on extracting discriminative multi-scale features through a hierarchical structure. Instead, the purpose of this work is to exploit these powerful multi-scale features more efficiently. Methods A Scale Attention Module (SAM) was proposed to recalibrate multi-scale features by explicitly modeling their importance score adaptively. The SAM was introduced into the segmentation model to construct the Scale Attention Network (SANet). The multi-scale features extracted from the encoder were first re-extracted to obtain more specific multi-scale features. Then the SAM was applied to recalibrate the features. Specifically, for the feature of each scale, a summation of Global Average Pooling and Global Max Pooling was used to create scale-wise feature representations. According to the representations, a lightweight network was used to generate the importance score of each scale. The features were recalibrated based on the scores, and a simple pixel-by-pixel summation was used to fuse the multi-scale features. The fused multi-scale feature was fed into a segmentation head to complete the task. Results The models were evaluated using fivefold cross-validation on 70 upper abdominal computed tomography scans of patients in a volume manner. The results showed that SANet could effectively alleviate the scale-variability problem and achieve better performance compared with UNet, Attention UNet, UNet++, Deeplabv3p, and CascadedUNet. The Dice similarity coefficients (DSCs) of colon, small intestine, and duodenum were (84.06 +/- 3.66)%, (76.79 +/- 5.12)%, and (61.68 +/- 4.32)%, respectively. The HD95 were (7.51 +/- 2.45) mm, (11.08 +/- 2.45) mm, and (12.21 +/- 1.95) mm, respectively. The values of relative volume difference were (3.4 +/- 0.8)%, (11.6 +/- 11.81)%, and (6.2 +/- 3.71)%, respectively. The values of center-of-mass distance were 7.85 +/- 2.82, 9.89 +/- 2.70, and 9.94 +/- 1.58, respectively. Compared with other attention modules and multi-scale feature exploitation approaches, SAM could obtain a 0.83-2.71 points improvement in terms of DSC with a comparable or even less number of parameters. The extensive experiments confirmed the effectiveness of SAM. Conclusions The SANet can efficiently exploit multi-scale features to alleviate the scale-variability problem and improve the segmentation performance on colon, small intestine, and duodenum of the upper abdomen.

SUnet: A Multi-Organ Segmentation Network Based on Multiple Attention

Multi-organ segmentation network for abdominal CT images based on spatial attention and deformable convolution

Many Birds, One Stone: Medical Image Segmentation with Multiple Partially Labeled Datasets

DASGC-Unet: An Attention Network for Accurate Segmentation of Liver CT Images

U-Net combined with multi-scale attention mechanism for liver segmentation in CT images

SADSNet: A robust 3D synchronous segmentation network for liver and liver tumors based on spatial attention mechanism and deep supervision

UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation

GASA-UNet: Global Axial Self-Attention U-Net for 3D Medical Image Segmentation

Variance‐aware attention U‐Net for multi‐organ segmentation

MTSegNet: Semi-supervised Abdominal Organ Segmentation in CT

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Hybrid dilation and attention residual U-Net for medical image segmentation

UNet based on dynamic convolution decomposition and triplet attention

DSGA-Net: Deeply Separable Gated Transformer and Attention Strategy for Medical Image Segmentation Network

Automatic Segmentation of Colon, Small Intestine, and Duodenum Based on Scale Attention Network

A Medical Image Segmentation Method Based on Improved UNet 3+ Network

Adaptive Attention Convolutional Neural Network for Liver Tumor Segmentation

CAT-Unet: An Enhanced U-Net Architecture with Coordinate Attention and Skip-Neighborhood Attention Transformer for Medical Image Segmentation

Sequential grafting of the right gastroepiploic artery in coronary artery bypass surgery.

N-Net: an UNet architecture with dual encoder for medical image segmentation

mm3DSNet: multi-scale and multi-feedforward self-attention 3D segmentation network for CT scans of hepatobiliary ducts