Attention U-Net: Learning Where to Look for the Pancreas

TL;DR

This paper introduces Attention U-Net, an enhanced neural network model that automatically learns to focus on relevant regions in medical images, improving segmentation accuracy without external localization modules.

Contribution

It presents a novel attention gate mechanism integrated into U-Net, enabling better focus on target structures with minimal added computational cost.

Findings

Attention U-Net improves segmentation accuracy on CT datasets.

Attention gates enhance model sensitivity and robustness.

The architecture maintains computational efficiency.

Abstract

We propose a novel attention gate (AG) model for medical imaging that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using explicit external tissue/organ localisation modules of cascaded convolutional neural networks (CNNs). AGs can be easily integrated into standard CNN architectures such as the U-Net model with minimal computational overhead while increasing the model sensitivity and prediction accuracy. The proposed Attention U-Net architecture is evaluated on two large CT abdominal datasets for multi-class image segmentation. Experimental results show that AGs consistently improve the prediction performance of U-Net across different datasets and training sizes while preserving computational efficiency. The code for the proposed architecture is publicly available.