Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

A Deep Learning Model for Judging Presence or Absence of Lesions in the Chest X-ray Images

흉부 디지털 영상의 병변 유무 판단을 위한 딥러닝 모델

  • Lee, Jong-Keun (Department of Information and Communication Engineering, Chung-buk National University) ;
  • Kim, Seon-Jin (Department of Information and Communication Engineering, Chung-buk National University) ;
  • Kwak, Nae-Joung (Department of Information and Communication Engineering, Chung-buk National University) ;
  • Kim, Dong-Woo (CELLGENTEK CO. LTD) ;
  • Ahn, Jae-Hyeong (Department of Information and Communication Engineering, Chung-buk National University)
  • Received : 2019.12.13
  • Accepted : 2019.12.26
  • Published : 2020.02.29
Download PDF
⟨ Previous Next ⟩

Abstract

There are dozens of different types of lesions that can be diagnosed through chest X-ray images, including Atelectasis, Cardiomegaly, Mass, Pneumothorax, and Effusion. Computed tomography(CT) test is generally necessary to determine the exact diagnosis and location and size of thoracic lesions, however computed tomography has disadvantages such as expensive cost and a lot of radiation exposure. Therefore, in this paper, we propose a deep learning algorithm for judging the presence or absence of lesions in chest X-ray images as the primary screening tool for the diagnosis of thoracic lesions. The proposed algorithm was designed by comparing various configuration methods to optimize the judgment of presence of lesions from chest X-ray. As a result, the evaluation rate of lesion presence of the proposed algorithm is about 1% better than the existing algorithm.

흉부 영상을 통해 진단 가능한 병변은 무기폐, 심비대, 덩어리, 기흉, 삼출 등 그 종류가 수십 가지에 이른다. 흉부 병변의 정확한 진단과 위치 및 크기를 판단하기 위해 일반적으로 전산화단층촬영(CT) 검사가 필요하지만, 전산화단층촬영은 검사 비용과 방사선 피폭 등의 단점이 있다. 따라서 본 논문에서는 흉부 병변 진단의 일차적 선별도구로서 방사선검사(X-ray) 영상에서 병변 유무 판단을 위한 딥러닝 알고리즘을 제안한다. 제안하는 알고리즘은 병변의 유무 판단에 최적화하기 위해 다양한 구성 방법들을 비교하여 설계하였다. 실험 결과, 기존 알고리즘보다 병변 유무 판단률이 약 1% 정도 향상되었다.

Keywords

References

  1. J. H. Yong, J. H. Kim, K. P. Cho, Human Anatomy and Physiology, Jeong-Dam, 1998.
  2. L. Yao, E. Poblenz, D. Dagunts, B. Covington, D. Bernard, and K. Lyman, "Learning to diagnose from scratch by exploiting dependencies among labels," arXiv preprint arXiv:1710.10501, 2017.
  3. B. R. Park, and D. W. Sung, "A comparative study of image quality and radiation dose with changes in tube voltage and current for a digital chest radiography," Journal of the Korean Society of Radiology, vol. 62, pp. 131-137, 2010. https://doi.org/10.3348/jksr.2010.62.2.131
  4. X. Wang, Y. Peng, L. Lu, Z. Lu, M. Bagheri, and R. M. Summers, "ChestX-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases," CVPR, 2017.
  5. P. Rajpurkar, J. Irvin, K. Zhu, B. Yang, H. Mehta, T. Duan, A. Bagul, R. L. Ball, C. Langlotz, K. Shpanskaya, M. P. Lungren, and A. Y. Ng, "CheXNet: Radiologist-level pneumonia detection on chest x-rays with deep learning," arXiv preprint arXiv:1711.05225, 2017.
  6. K. Simonyan, and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," ICLR, 2014.
  7. C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, "Going deeper with convolutions," CVPR, 2015.
  8. Y. Bengio, P. Simard, and P. Frasconi, "Learning long-term dependencies with gradient descent is difficult," Institute of Electrical and Electronics Engineers Transactions on Neural Networks, vol. 5, no. 2, pp. 157-166, Mar. 1994.
  9. X. Glorot, and Y. Bengio, "Understanding the difficulty of training deep feed-forward neural networks," in Proceeding of machine learning research, pp. 249-256, 2010.
  10. K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," CVPR, 2016.
  11. A. P. Bradley, "The use of the area under the ROC curve in the evaluation of machine learning algorithms," Pattern Recognition, vol. 30, no. 7, pp. 1145-1159, Jul. 1997. https://doi.org/10.1016/S0031-3203(96)00142-2
  12. D. P. Kingma, and J. Ba, "Adam: A method for stochastic optimization," ICLR, 2015.

Cited by

  1. 심장비대증 환자의 흉부 X선 영상에 대한 Inception V3 알고리즘의 분류 성능평가 vol.15, pp.4, 2020, https://doi.org/10.7742/jksr.2021.15.4.455