Optimization of CNN for Diagnosis on Lung Disease by Lung Segmentation and Rib Suppression

TL;DR

Two U-Nets for lung segmentation and rib suppression, plus histogram equalization, feed an Xception-based CNN for chest X-ray lung disease classification; the pipeline classifies more accurately than Xception applied directly.

中文简介：用两个 U-Net 分别做肺部分割和肋骨抑制，再经直方图均衡化增强后输入基于 Xception 的 CNN 进行胸片肺病分类，准确率高于直接使用 Xception。

Key points

• One U-Net segments the lungs and another suppresses ribs in chest X-ray images.
• Histogram equalization enhances image contrast before classification.
• An Xception-based CNN classifies the processed images.
• The pipeline reduces interference from regions outside the lungs and from X-ray machine variation, and outperforms direct Xception classification.

Abstract

The discrimination of lung diseases by chest X-ray images is a clinically important tool. How to use artificial intelligence to accurately and quickly help doctors to diagnose different lung diseases is very important in the context of the current COVID-19 global pandemic. In this paper, we propose a model structure, including two U-Net, which implement lung segmentation and rib suppression for chest X-ray images respectively, image enhancement techniques such as histogram equalization, which enhances images contrast, and a Xception-based CNN, which classifies the processed images finally. The model can effectively avoid the interference of regions outside the lung to CNN for feature recognition and the influence of environmental factors such as X-ray machines on the quality of X-ray images and thus on the classification. The experimental results show that the classification accuracy of the model is higher than that of the direct use of the Xception model for classification.

Citation

Xinye Yang, Yuhang Liu, Zhiwei Lin, Lingyu Zhong, Teoh Teik Toe. Optimization of CNN for Diagnosis on Lung Disease by Lung Segmentation and Rib Suppression. 15th International Symposium on Computational Intelligence and Design (ISCID 2022). https://doi.org/10.1109/iscid56505.2022.00063

@inproceedings{yang2022lung,
  title   = {Optimization of CNN for Diagnosis on Lung Disease by Lung Segmentation and Rib Suppression},
  author  = {Yang, Xinye and Liu, Yuhang and Lin, Zhiwei and Zhong, Lingyu and Teoh, Teik Toe},
  booktitle = {2022 15th International Symposium on Computational Intelligence and Design (ISCID)},
  year    = {2022},
  doi     = {10.1109/iscid56505.2022.00063}
}

FAQ

Related work by the authors

• Confidence-gated cloud-edge cascade triage via variational risk minimization for medical imaging. Smart Health 2026. Later chest X-ray triage work built on foundation-model distillation.
• Reliability Stress Tests and Decision-Time Routing for Chest X-ray Vision-Language Models. CHASE 2026 Workshop. Reliability of medical VLMs on chest X-rays.