Graph Neural Network and Superpixel Based Brain Tissue Segmentation (Corrected Version)

TL;DR

This paper introduces GNN-SEG, a novel brain tissue segmentation method using graph neural networks and superpixels, overcoming CNN limitations and demonstrating superior performance on MRI datasets.

Contribution

The paper presents a GNN-based segmentation approach that uses superpixels and interaction modules, offering a new perspective beyond traditional CNN methods.

Findings

GNN-SEG outperforms CNN-based methods on four MRI datasets.

Superpixel-based GNN effectively captures brain tissue structures.

Interaction modules enhance feature learning in segmentation.

Abstract

Convolutional neural networks (CNNs) are usually used as a backbone to design methods in biomedical image segmentation. However, the limitation of receptive field and large number of parameters limit the performance of these methods. In this paper, we propose a graph neural network (GNN) based method named GNN-SEG for the segmentation of brain tissues. Different to conventional CNN based methods, GNN-SEG takes superpixels as basic processing units and uses GNNs to learn the structure of brain tissues. Besides, inspired by the interaction mechanism in biological vision systems, we propose two kinds of interaction modules for feature enhancement and integration. In the experiments, we compared GNN-SEG with state-of-the-art CNN based methods on four datasets of brain magnetic resonance images. The experimental results show the superiority of GNN-SEG.