# SegAN: Adversarial Network with Multi-scale $L_1$ Loss for Medical Image Segmentation

## TL;DR

SegAN introduces an adversarial network with a multi-scale $L_1$ loss for improved medical image segmentation, demonstrating superior stability and performance over existing methods like U-net on brain tumor datasets.

## Contribution

The paper proposes a novel SegAN framework with a multi-scale loss and critic network, enhancing segmentation accuracy and stability compared to traditional GANs and U-net.

## Key findings

- SegAN achieves comparable results to state-of-the-art methods on BRATS datasets.
- SegAN shows improved precision and sensitivity for tumor core segmentation.
- SegAN outperforms U-net in dice score and precision on brain tumor segmentation tasks.

## Abstract

Inspired by classic generative adversarial networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale $L_1$ loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. In our SegAN framework, the segmentor and critic networks are trained in an alternating fashion in a min-max game: The critic takes as input a pair of images, (original_image $*$ predicted_label_map, original_image $*$ ground_truth_label_map), and then is trained by maximizing a multi-scale loss function; The segmentor is trained with only gradients passed along by the critic, with the aim to minimize the multi-scale loss function. We show that such a SegAN framework is more effective and stable for the segmentation task, and it leads to better performance than the state-of-the-art U-net segmentation method. We tested our SegAN method using datasets from the MICCAI BRATS brain tumor segmentation challenge. Extensive experimental results demonstrate the effectiveness of the proposed SegAN with multi-scale loss: on BRATS 2013 SegAN gives performance comparable to the state-of-the-art for whole tumor and tumor core segmentation while achieves better precision and sensitivity for Gd-enhance tumor core segmentation; on BRATS 2015 SegAN achieves better performance than the state-of-the-art in both dice score and precision.