Improving Anatomical Plausibility in Medical Image Segmentation via Hybrid Graph Neural Networks: Applications to Chest X-Ray Analysis

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...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on medical imaging Vol. 42; no. 2; pp. 546 - 556
Main Authors: Gaggion, Nicolas, Mansilla, Lucas, Mosquera, Candelaria, Milone, Diego H., Ferrante, Enzo
Format: Journal Article
Language:English
Published: United States IEEE 01.02.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
anatomically plausible segmentation
Artificial neural networks
Biomedical imaging
Chest
Coders
Computational modeling
Convolutional neural networks
Decoding
Domains
Encoders-Decoders
Entropy
Graph convolutional neural networks
graph generative models
Graph neural networks
Graph representations
Graphical representations
Image processing
Image Processing, Computer-Assisted - methods
Image segmentation
landmark based segmentation
localized skip connections
Medical imaging
Neural networks
Neural Networks, Computer
Occlusion
Pixels
Radiography
Shape
Thorax - diagnostic imaging
X ray analysis
X-Rays
ISSN:0278-0062, 1558-254X, 1558-254X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary: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.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0278-0062
1558-254X
1558-254X
DOI:10.1109/TMI.2022.3224660