Improving anatomical plausibility in medical image segmentation via hybrid graph neural networks: applications to chest x-ray analysis

Abstract

Anatomical segmentation is a fundamental task in medical image computing, generally tackled with fully convolutional neural networks which produce dense segmentation masks. These models are often trained with loss functions such as cross-entropy or Dice, which assume pixels to be independent of each other, thus ignoring topological errors and anatomical inconsistencies. We address this limitation by moving from pixel-level to graph representations, which allow to naturally incorporate anatomical constraints by construction. To this end, we introduce HybridGNet, an encoder-decoder neural architecture that leverages standard convolutions for image feature encoding and graph convolutional neural networks (GCNNs) to decode plausible representations of anatomical structures. We also propose a novel image-to-graph skip connection layer which allows localized features to flow from standard convolutional blocks to GCNN blocks, and show that it improves segmentation accuracy. The proposed architecture is extensively evaluated in a variety of domain shift and image occlusion scenarios, and audited considering different types of demographic domain shift. Our comprehensive experimental setup compares HybridGNet with other landmark and pixel-based models for anatomical segmentation in chest x-ray images, and shows that it produces anatomically plausible results in challenging scenarios where other models tend to fail.

arXiv commentSource code at https://github.com/ngaggion/HybridGNet Accepted for publication at IEEE TMI

References in corpus (9)

• Auto-Encoding Variational Bayes
• Automated Design of Deep Learning Methods for Biomedical Image Segmentation
• The Emerging Field of Signal Processing on Graphs: Extending High-Dimensional Data Analysis to Networks and Other Irregular Domains
• Geometric deep learning: going beyond Euclidean data
• Spectral Networks and Locally Connected Networks on Graphs
• Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentation
• PadChest: A large chest x-ray image dataset with multi-label annotated reports
• Common Limitations of Image Processing Metrics: A Picture Story
• Intraoperative Liver Surface Completion with Graph Convolutional VAE

Cited by in corpus (7)

• CheXmask: a large-scale dataset of anatomical segmentation masks for multi-center chest x-ray images
• Multi-center anatomical segmentation with heterogeneous labels via landmark-based models
• Multi-view Hybrid Graph Convolutional Network for Volume-to-mesh Reconstruction in Cardiovascular MRI
• Joint Dense-Point Representation for Contour-Aware Graph Segmentation
• Fitting Skeletal Models via Graph-based Learning
• Combining Image- and Geometric-based Deep Learning for Shape Regression: A Comparison to Pixel-level Methods for Segmentation in Chest X-Ray
• Performance Analysis and Optimization of 3D Generative Diffusion Models across GPU Architectures