AMSUnet: A neural network using atrous multi-scale convolution for medical image segmentation

In recent years, Unet and its variants have gained astounding success in the realm of medical image processing. However, some Unet variant networks enhance their performance while increasing the number of parameters tremendously. For lightweight and performance enhancement jointly considerations, inspired by SegNeXt, we develop a medical image segmentation network model using atrous multi-scale (AMS) convolution, named AMSUnet. In particular, we construct a convolutional attention block AMS using atrous and multi-scale convolution, and redesign the downsampling encoder based on this block, called AMSE. To enhance feature fusion, we design a residual attention mechanism module (i.e., RSC) and apply it to the skip connection. Compared with existing models, our model only needs 2.62 M parameters to achieve the purpose of lightweight. According to experimental results on various datasets, the segmentation performance of the designed model is superior for small, medium, and large-scale targets. Code will be available at https://github.com/llluochen/ AMSUnet.

Automated summary

In recent years, Unet and its variants have achieved remarkable success in medical image processing. However, some Unet variants increase their performance by significantly increasing the number of parameters. To address this issue, we propose a lightweight medical image segmentation network called AMSUnet, which utilizes atrous multi-scale (AMS) convolution. Our model only requires 2.62 M parameters while achieving excellent segmentation performance for small, medium, and large-scale targets.