Email this article EFFICIENT BONE SUPPRESSION IN CHEST X-RAY WITH SIMPLIFIED LOSS FUNCTION
About this article
Received: 19/04/25 Revised: 30/06/25 Published: 30/06/25Authors
1. Nguyen Quang Duy, University of Transport and Communications2. Nguyen Duc Manh, University of Transport and Communications
3. Pham Huy Hoang, University of Transport and Communications
4. Pham Thanh An, University of Transport and Communications
5. Cao Thi Luyen
, University of Transport and Communications
Abstract
Chest X-ray imaging is a vital tool for diagnosing thoracic diseases such as pneumonia, lung cancer, and rib fractures. However, rib shadows often obscure or mimic pulmonary lesions - especially in posterior and axillary regions - thereby reducing diagnostic accuracy. Traditional techniques like dual-energy subtraction can separate bone and soft tissue structures but require specialized equipment and may increase radiation exposure. In this paper, we survey representative deep learning architectures applied to the task of bone suppression in standard chest X-ray images, aiming to improve the visibility of lung abnormalities without the need for additional hardware. Our main contribution lies in the use of a simplified loss function that reduces computational complexity while maintaining high inference accuracy. The effectiveness of this loss function is evaluated across various models to demonstrate its performance in suppressing bone shadows.
Keywords
Full Text:
PDFReferences
[1] P. Vock and Z. Szucs-Farkas, “Dual energy subtraction: Principles and clinical applications,” Eur. J. Radiol., vol. 72, no. 2, pp. 231–237, 2009.
[2] M. Gusarev, R. Kuleev, A. Khan, A. R. Rivera, and A. M. Khattak, “Deep learning models for bone suppression in chest radiographs,” Proc. IEEE Conf. Comput. Intell. Bioinf. Comput. Biol. (CIBCB), 2017, pp. 1–7.
[3] A. Zarshenas, J. Liu, P. Forti, and K. Suzuki, “Separation of bones from soft tissue in chest radiographs: Anatomy-specific orientation-frequency-specific deep neural network convolution,” Med. Phys., vol. 46, no. 5, pp. 2232–2242, 2019.
[4] Y. Chen et al., “Bone suppression of chest radiographs with cascaded convolutional networks in wavelet domain,” IEEE access, vol. 7, pp. 8346–8357, 2019.
[5] M.-C. Huynh, T.-H. Nguyen, and M.-T. Tran, “Context learning for bone shadow exclusion in CheXNet accuracy improvement,” in Proc. 10th Int. Conf. Knowl. Syst. Eng. (KSE), 2018, pp. 135–140.
[6] P. Rajpurkar et al., “CheXNet: Radiologist-level pneumonia detection on chest X-rays with deep learning,” arXiv preprint, arXiv:1711.05225, 2017.
[7] P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” arXiv preprint, arXiv:1611.07004, 2016.
[8] G. Rani, A. Misra, V. S. Dhaka, E. Zumpano, and E. Vocaturo, “Spatial feature and resolution maximization GAN for bone suppression in chest radiographs,” Comput. Methods Programs Biomed., vol. 224, 2022, Art. no. 107024.
[9] Z. Zhou, L. Zhou, and K. Shen, “Dilated conditional GAN for bone suppression in chest radiographs with enforced semantic features,” Med. Phys., vol. 47, no. 12, pp. 6207–6215, 2020.
[10] O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional networks for biomedical image segmentation,” Proc. Med. Image Comput. Comput.-Assist. Intervent. (MICCAI), 2015, pp. 234–241.
[11] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016, pp. 770–778.
[12] S. Arvind, J. V. Tembhurne, T. Diwan, and P. Sahare, “Evaluation of deep learning methods for bone suppression from dual energy chest radiography,” Artificial Neural Networks and Machine Learning – ICANN 2020, 2020, pp. 247–257.
[13] T.-Y. Lin, P. Dollár, R. B. Girshick, K. He, B. Hariharan, and S. J. Belongie, “Feature pyramid networks for object detection,” arXiv preprint, arXiv:1612.03144, 2016.
[14] S. Kalisz and M. Marczyk, “Autoencoder-based bone removal algorithm from x-ray images of the lung,” Proc. IEEE 21st Int. Conf. Bioinf. Bioeng. (BIBE), 2021, pp. 1–6.
[15] S. Rajaraman, G. Zamzmi, L. Folio, P. Alderson, and S. Antani, “Chest x-ray bone suppression for improving classification of tuberculosis-consistent findings,” Diagnostics, vol. 11, no. 5, 2021, Art. no. 840.
[16] S. Arvind, J. V. Tembhurne, T. Diwan, and P. Sahare, “Improvised light weight deep CNN based U-Net for the semantic segmentation of lungs from chest X-rays,” Results Eng., vol. 17, 2023, Art. no. 100929.
[17] O. Oktay et al., “Attention U-Net: Learning where to look for the pancreas,” arXiv preprint, arXiv:1804.03999, 2018.
[18] P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2017, pp. 1125–1134.
[19] H. M. Chuong, “X-ray bone shadow suppression,” Kaggle, 2022. [Online]. Available: https://www.kaggle.com/datasets/hmchuong/xray-bone-shadow-supression. [Accessed March 30, 2025].
DOI: https://doi.org/10.34238/tnu-jst.12628
Refbacks
- There are currently no refbacks.