Reducing Metal Streak Artifacts in CT Images via Deep Learning: Pilot Results

Lars Gjesteby, Qingsong Yang, Yan Xi, Bernhard Claus, Yannan Jin, Bruno De Man, Ge Wang

DOI:10.12059/Fully3D.2017-11-3202009

Published in:Fully3D 2017 Proceedings

Pages:611-614

Keywords:
computed tomography (CT), deep learning, convolutional neural network (CNN), metal artifact reduction (MAR), proton therapy planning
Machine learning including deep learning is rapidly gaining popularity as a generic solution to problems across many fields. In medical imaging, deep learning has been successfully applied to image processing and analysis. In this paper, we employ a convolutional neural network (CNN) to address the long-standing problem of metal artifacts in CT images. Despite a vast number of metal artifact reduction (MAR) methods developed over the past four decades, there remain clinical areas in need of better results. Specifically, proton therapy planning requires high image quality for accurate tumor volume estimation. Errors in the image reconstruction may lead to treatment failure. In this paper, we merge deep learning with a state-of-the-art normalization-based MAR algorithm, NMAR, to correct metal streaks in critical image regions. Our results show that deep learning is a novel way to address CT reconstruction challenges, yielding images superior to the state of the art. 
Lars Gjesteby
Department of Biomedical Engineering, Rensselaer Polytechnic Institute
Yannan Jin
General Electric Global Research Center
Qingsong Yang
Department of Biomedical Engineering, Rensselaer Polytechnic Institute
Bruno De Man
General Electric Global Research Center
Yan Xi
First Imaging Technology
Bernhard Claus
General Electric Global Research Center
Ge Wang
Department of Biomedical Engineering, Rensselaer Polytechnic Institute
  1. L. Gjesteby, B. De Man, Y. Jin, H. Paganetti, J. Verburg, D. Giantsoudi, and G. Wang, “Metal Artifact Reduction in CT: Where Are We After Four Decades?,” IEEE Access, vol. 4, pp. 5826–5849, 2016.
  2. G. X. Ding and W. Y. Christine, “A study on beams passing through hip prosthesis for pelvic radiation treatment,” Int. J. Radiat. Oncol. Biol. Phys., vol. 51, no. 4, pp. 1167–1175, 2001.
  3. C. Reft, R. Alecu, I. J. Das, B. J. Gerbi, P. Keall, E. Lief, B. J. Mijnheer, N. Papanikolaou, C. Sibata, and J. Van Dyk, “Dosimetric considerations for patients with HIP prostheses undergoing pelvic irradiation. Report of the AAPM Radiation Therapy Committee Task Group 63,” Med. Phys., vol. 30, no. 6, pp. 1162–1182, 2003.
  4. R. M. Lewitt and R. H. T. Bates, “Image-reconstruction from projections. III. Projection completion methods (theory),” Optik (Stuttg)., vol. 50, no. 3, pp. 189–204, 1978.
  5. T. Hinderling, P. Ruegsegger, M. Anliker, and C. Dietschi, “Computed Tomography Reconstruction from Hollow Projections: An Application to In Vivo Evaluation of Artificial Hip Joints,” J. Comput. Assist. Tomogr., vol. 3, no. 1, pp. 52–57, 1979.
  6. G. H. Glover and N. J. Pelc, “An algorithm for the reduction of metal clip artifacts in CT reconstructions.,” Med. Phys., vol. 8, no. 6, pp. 799–807, 1981.
  7. W. A. Kalender, R. Hebel, and J. Ebersberger, “Reduction of CT artifacts caused by metallic implants,” Radiology, vol. 164, no. 2, pp. 576–577, 1987.
  8. C. R. Crawford, J. G. Colsher, N. J. Pelc, and A. H. R. Lonn, “High speed reprojection and its applications,” in SPIE Medical Imaging II, 1988, pp. 311–318.
  9. R. Naidu, I. Bechwati, S. Karimi, S. Simanovsky, and C. Crawford, “Method of and system for reducing metal artifacts in images generated by x-ray scanning devices,” 6,721,387 B1, 2004.
  10. C. S. Olive, M. R. Klaus, V. Pekar, K. Eck, and L. Spies, “Segmentation aided adaptive filtering for metal artifact reduction in radio-therapeutic CT images,” in SPIE Medical Imaging 2004, 2004, vol. 5370, pp. 1991–2002.
  11. C. Lemmens, D. Faul, and J. Nuyts, “Suppression of metal artifacts in CT using a reconstruction procedure that combines MAP and projection completion,” IEEE Trans. Med. Imaging, vol. 28, no. 2, pp. 250–60, 2009.
  12. J. Müller and T. M. Buzug, “Spurious structures created by interpolation-based CT metal artifact reduction,” in SPIE Medical Imaging, 2009, p. 72581Y–72581Y.
  13. E. Meyer, F. Bergner, R. Raupach, T. Flohr, and M. Kachelrieß, “Normalized metal artifact reduction (NMAR) in computed tomography,” Med. Phys., vol. 37, no. 10, pp. 5482–5493, 2010.
  14. E. Meyer, R. Raupach, B. Schmidt, A. H. Mahnken, and M. Kachelriess, “Adaptive Normalized Metal Artifact Reduction (ANMAR) in Computed Tomography,” in 2011 IEEE Nuclear Science Symposium Conference Record, 2011, pp. 2560–2565.
  15. G. Henrich, “A simple computational method for reducing streak artifacts in CT images,” Comput. Tomogr., vol. 4, no. 1, pp. 67–71, 1980.
  16. M. Bal, H. Celik, K. Subramanyan, K. Eck, and L. Spies, “A radial adaptive filter for metal artifact reduction,” in SPIE Medical Imaging 2005, 2005, vol. 5747, pp. 2075–2082.
  17. O. Watzke and W. A. Kalender, “A pragmatic approach to metal artifact reduction in CT: Merging of metal artifact reduced images,” Eur. Radiol., vol. 14, no. 5, pp. 849–856, 2004.
  18. Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, May 2015.
  19. Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel, “Backpropagation applied to handwritten zip code recognition,” Neural Comput., vol. 1, no. 4, pp. 541–551, 1989.
  20. A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,” in Advances in neural information processing systems, 2012, pp. 1097–1105.
  21. C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, “Going deeper with convolutions,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 1–9.
  22. B. De Man, S. Basu, N. Chandra, B. Dunham, P. Edic, M. Iatrou, S. McOlash, P. Sainath, C. Shaughnessy, and B. Tower, “CatSim: a new computer assisted tomography simulation environment,” in Medical Imaging, 2007, p. 65102G–65102G.
  23. C. Dong, C. C. Loy, K. He, and X. Tang, “Image super-resolution using deep convolutional networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 38, no. 2, pp. 295–307, 2016.
  24. Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S. Guadarrama, and T. Darrell, “Caffe: Convolutional architecture for fast feature embedding,” in Proceedings of the 22nd ACM international conference on Multimedia, 2014, pp. 675–678.
  25. N. Haramati, R. B. Staron, K. Mazel-Sperling, K. Freeman, E. L. Nickoloff, C. Barax, and F. Feldman, “CT scans through metal scanning technique versus hardware composition,” Comput. Med. Imaging Graph., vol. 18, no. 6, pp. 429–434, 1994.
  26. K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 770–778.