한국방사선학회논문지 (Journal of the Korean Society of Radiology)

  • 제18권5호
  • /
  • Pages.531-540
  • /
  • 2024
  • /
  • 1976-0620(pISSN)
  • /
  • 2384-0633(eISSN)
한국방사선학회 (The Korean Society of Radiology)
권호 표지
DOI QR코드

DOI QR Code

최적화 알고리즘과 학습률 적용에 따른 흉부 X선 영상 딥러닝 분류 모델 성능평가

Performance Evaluation of Chest X-ray Image Deep Learning Classification Model according to Application of Optimization Algorithm and Learning Rate

  • 김지율 (대우병원 영상의학과) ;
  • 정봉재 (방사선과학기술연구소)
  • Ji-Yul Kim (Daewoo general hospital) ;
  • Bong-Jae Jeong (Radiation Science and Technology Research Institute)
  • 투고 : 2024.10.14
  • 심사 : 2024.10.31
  • 발행 : 2024.10.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

최근에는 딥러닝을 이용한 의료영상 분야의 자동진단 솔루션에 대한 연구 및 개발이 활발하게 진행되고 있다. 본 연구에서는 컨볼루션 인공 신경망 기반의 딥러닝 모델인 Inception V3를 이용하여 흉부 X선 영상의 폐렴 유무 분류에 대한 신속하면서도 정확한 분류 딥러닝 모델링을 찾고자 하였다. 이러한 이유로 딥러닝 모델링에 최적화알고리즘 AdaGrad, RMS Prop, Adam을 적용한 후 학습률을 0.01과 0.001로 선택적으로 적용하여 딥러닝 모델링을 구현한 후 흉부 X선 영상 폐렴 유무 분류에 대한 성능을 비교 평가하였다. 연구결과 분류 모델의 성능과 인공신경망의 학습상태를 평가할 수 있는 검증 모델링에서는 학습률 0.001과 최적화 알고리즘으로 Adam을 적용한 경우 흉부 X 선 영상의 폐렴 유무 분류에 대한 딥러닝 모델링의 성능이 가장 우수하다는 것을 알 수 있었다. 그리고 최근 딥러닝 모델링의 설계 시 최적화 알고리즘으로 주로 적용이 되는 Adam의 경우 학습률 0.01과 0.001의 선택적인 적용에서 우수한 성능 및 우수한 Metric 결과를 나타내었다. 테스트 모델링에 대한 Metric 평가에서는 학습률 0.1을 적용한 AdaGrad 가 가장 우수한 결과를 나타내었다. 이러한 결과를 통하여 이진법 기반의 의료영상 분류 딥러닝 모델링의 설계 시, 신속하면서도 정확한 성능을 기대하기 위해서는 최적화 알고리즘으로 Adam을 적용하는 경우에는 학습률 0.01, AdaGrad를 적용하는 경우에는 학습률은 0.01을 우선적으로 적용할 것을 권고한다. 그리고 향후 유사 연구 시, 본 연구 결과는 기초자료로 제시될 것이라 사료되며 딥러닝을 이용한 의료영상의 자동 진단 목적의 헬스·바이오 산업에서 유용한 자료로 활용되기를 기대한다.

Recently, research and development on automatic diagnosis solutions in the medical imaging field using deep learning are actively underway. In this study, we sought to find a fast and accurate classification deep learning modeling for classification of pneumonia in chest images using Inception V3, a deep learning model based on a convolutional artificial neural network. For this reason, after applying the optimization algorithms AdaGrad, RMS Prop, and Adam to deep learning modeling, deep learning modeling was implemented by selectively applying learning rates of 0.01 and 0.001, and then the performance of chest X-ray image pneumonia classification was compared and evaluated. As a result of the study, in verification modeling that can evaluate the performance of the classification model and the learning state of the artificial neural network, it was found that the performance of deep learning modeling for classification of the presence or absence of pneumonia in chest X-ray images was the best when applying Adam as the optimization algorithm with a learning rate of 0.001. I was able to. And in the case of Adam, which is mainly applied as an optimization algorithm when designing deep learning modeling, it showed excellent performance and excellent metric results when selectively applying learning rates of 0.01 and 0.001. In the metric evaluation of test modeling, AdaGrad, which applied a learning rate of 0.1, showed the best results. Based on these results, when designing deep learning modeling for binary-based medical image classification, in order to expect quick and accurate performance, a learning rate of 0.01 is preferentially applied when applying Adam as an optimization algorithm, and a learning rate of 0.01 is preferentially applied when applying AdaGrad. I recommend doing this. In addition, it is expected that the results of this study will be presented as basic data during similar research in the future, and it is expected to be used as useful data in the health and bio industries for the purpose of automatic diagnosis of medical images using deep learning.

키워드

참고문헌

  1. J. Y. Hong, S. H. Park, Y. J. Jung, "Artificial Intelligence based medical imaging: An overview", Journal of Radiological Science and Technology, Vol. 43, No. 3, pp. 195-208, 2020. https://doi.org/10.17946/JRST.2020.43.3.195 
  2. Y. J. Kim, K. G. Kim, "Development of an optimized deep learning model for medical imaging", Journal of the Korean Society of Radiology, Vol. 81, No. 6, pp. 1274-1289. 2020. https://doi.org/10.3348/jksr.2020.0171 
  3. K. H. Ann, S. Y. Oum, "A COVID-19 Chest X-ray Reading Technique based on Deep Learning", The International Promotion Agency of Culture Technology, Vol. 6, No. 4, pp. 789-795, 2020. http://dx.doi.org/10.17703/JCCT.2020.6.4.789 
  4. T. Rahman, M. E. Hoque Chowdhury, A. Khandakar, K. R. Islam, K. F. Islam, Z. B. Mahbub, M. A. Kadir, S. B. A. Kashem, "Transfer Learning with Deep Convolutional Neural Netꠓwork (CNN) for Pneumonia Detection Using Chest X-ray", MDPI, Vol. 10, No. 3233, 2020. https://doi.org/10.3390/app10093233 
  5. D. S. Al-Dulaimi, A. G. Mahmoud, N. M. Hassan, A. Alkhayyat, S. A. Majeed, G. Singh, "Development of Pneumonia Disease Detection Model Based on Deep Learning Algorithm", Wireless Communications and Mobile Computing, Vol. 2022, 2022. https://doi.org/10.1155/2022/2951168 
  6. A. K. Jaiswal, P. Tiwari, S. Kumar, D. Gupta, A. Khanna, J. J.P.C. Rodrigues, "Identifying pneumonia in chest X-rays: A deep learning approach", Elsevier, Vol. 145, pp. 511-518, 2019. https://doi.org/10.1016/j.measurement.2019.05.076 
  7. H. J. Song, E. B. Lee, H. J. Jo, "Evaluation of Classification and Accuracy in Chest X-ray Images using Deep Learning with Convolution Neural Network", The Korean Society of Radiology, Vol 14, No. 1, pp. 39-44, 2020. https://doi.org/10.7742/jksr.2019.14.1.39 
  8. B. Hu, H. Guo, P. Zhou, Z. L. Shi, "Characteristics of SARS-CoV-2 and COVID-19", Nature Reviews Microbiology, Vol. 19, No. 3, pp. 141-154, 2021. https://doi.org/10.1038/s41579-020-00459-7 
  9. G. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Setio, F. Ciompi, M. Ghafoorian, J. A.W.M. van der Laak, B. van Ginneken, C. I. Sanchez, "A survey on deep learning in medical image analysis", Medical Image Analysis, Vol. 42, pp. 60-88, 2017. https://doi.org/10.1016/j.media.2017.07.005 
  10. S. Y. Kim, W. K. Chung, S. R. Shin, "Acoustic Full-waveform Inversion using Adam Optimizer", Korean Society of Earth and Exploration Geophysicists, Vol. 22, No. 4, pp. 202-209, 2019. https://doi.org/10.7582/GGE.2019.22.4.202 
  11. J. H. Lee, B. S. Kang, "Improving Performance of Machine Learning-Based Algorithms with Adaptive Learning Rate", The Journal of Korean Institute of Information Technology, Vol. 18, No. 10, pp. 9-14, 2020. https://doi.org/10.14801/jkiit.2020.18.10.9 
  12. Ch. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna, "Rethinking the Inception Architecture for Computer Vision", IEEE, 2016. https://doi.org/10.48550/arXiv.1512.00567 
  13. S. J. Kim, J. C. Yu, "COVID-19 Chest X-ray reading Technique based on Deep Learning", Proceedings of the Korean Society of Computer Information Conference, Vol. 29, No. 1, pp. 31-32. 2021. 
  14. J. Y. Kim, S. Y. Ye, "Diagnostic Classification of Chest X-ray Pneumonia using Inception V3 Modeling", Journal of The Korean Society of Radiology, Vol. 14, No. 6, pp. 773-780, 2020. https://doi.org/10.7742/jksr.2020.14.6.773 
  15. C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna, "Rethinking the Inception Architecture for Computer Vision", IEEE Xplore 2016. https://doi.org/10.1109/CVPR.2016.308 
  16. N. Lin, Y. I. Chen, H. Liu, H. Liu, "A Comparative Study of Machine Learning Models with Hyperparameter Optimization Algorithm for Mapping Mineral Prospectivity", Minerals, Vol. 11, No. 2, pp. 159, 2021. https://doi.org/10.3390/min11020159 
  17. Y. H. Kim, J. Y. Hong, B. J. Kim, "Performance comparison of machine learning classification methods for decision of disc cutter replacement of shield TBM", Journal of Korean Tunnelling and Underground Space Association, Vol. 22, No. 5, pp. 575-589, 2020. https://doi.org/10.9711/KTAJ.2020.22.5.575 
  18. J. H. Lee. S. Y. Ye, "A Comparative Study of Deep Learning Models for Pneumonia Detection : CNN, VUNO, LUIT Models", Journal of Radiation Industry, Vol. 81, No. 6, pp. 1274-1289, 2020. https://doi.org/2024.10.23042/radin.2024.18.3.177 
  19. W. Y. Jung, J. H. Kim, J. E. Park, "Evaluation of Classification Performance of Inception V3 Algorithm for Chest X-ray Images of Patients with Cardiomegaly", Journal of the Korean Society of Radiology, Vol. 15, No. 4, pp. 455-461, 2021. https://doi.org/10.7742/jksr.2021.15.4.455 
  20. J. K. Lee, S. J. Kim, N. J. Kwak, "A Deep Learning Model for Judging Presence or Absence of Lesions in the Chest X-ray Images", Journal of the Korea Institute of Information and Communication Engineering, Vol. 24, No. 2, pp. 212-218, 2020. https://doi.org/10.6109/jkiice.2020.24.2.21