MLP-GAN for Brain Vessel Image Segmentation

TL;DR

This paper introduces MLP-GAN, a multi-view 2D cGAN approach for 3D brain vessel segmentation that reduces memory usage while preserving spatial information, outperforming existing methods.

Contribution

The paper proposes a novel multi-view 2D cGAN framework combining U-Net and MLP-Mixer for efficient 3D brain vessel segmentation.

Findings

MLP-GAN outperforms state-of-the-art methods on public datasets.

The multi-view approach alleviates memory issues of 3D networks.

Integration of MLP-Mixer captures global cross-patch information.

Abstract

Brain vessel image segmentation can be used as a promising biomarker for better prevention and treatment of different diseases. One successful approach is to consider the segmentation as an image-to-image translation task and perform a conditional Generative Adversarial Network (cGAN) to learn a transformation between two distributions. In this paper, we present a novel multi-view approach, MLP-GAN, which splits a 3D volumetric brain vessel image into three different dimensional 2D images (i.e., sagittal, coronal, axial) and then feed them into three different 2D cGANs. The proposed MLP-GAN not only alleviates the memory issue which exists in the original 3D neural networks but also retains 3D spatial information. Specifically, we utilize U-Net as the backbone for our generator and redesign the pattern of skip connection integrated with the MLP-Mixer which has attracted lots of attention recently. Our model obtains the ability to capture cross-patch information to learn global information with the MLP-Mixer. Extensive experiments are performed on the public brain vessel dataset that show our MLP-GAN outperforms other state-of-the-art methods. We release our code at https://github.com/bxie9/MLP-GAN