# ImageGCN: Multi-Relational Image Graph Convolutional Networks for Disease Identification with Chest X-rays

## TL;DR

This paper introduces ImageGCN, a graph convolutional network that models relationships between chest X-ray images to improve disease detection and localization, outperforming existing methods on multiple datasets.

## Contribution

The paper proposes ImageGCN, a novel multi-relational image modeling framework that incorporates image relationships into representation learning for disease identification.

## Key findings

- Outperforms baseline models in disease classification accuracy
- Achieves comparable or better results than state-of-the-art methods
- Effective in weakly supervised object detection in X-ray images

## Abstract

Image representation is a fundamental task in computer vision. However, most of the existing approaches for image representation ignore the relations between images and consider each input image independently. Intuitively, relations between images can help to understand the images and maintain model consistency over related images, leading to better explainability. In this paper, we consider modeling the image-level relations to generate more informative image representations, and propose ImageGCN, an end-to-end graph convolutional network framework for inductive multi-relational image modeling. We apply ImageGCN to chest X-ray images where rich relational information is available for disease identification. Unlike previous image representation models, ImageGCN learns the representation of an image using both its original pixel features and its relationship with other images. Besides learning informative representations for images, ImageGCN can also be used for object detection in a weakly supervised manner. The experimental results on 3 open-source x-ray datasets, ChestX-ray14, CheXpert and MIMIC-CXR demonstrate that ImageGCN can outperform respective baselines in both disease identification and localization tasks and can achieve comparable and often better results than the state-of-the-art methods.