Skip to main content

Real-Time Video Denoising to Reduce Ionizing Radiation Exposure in Fluoroscopic Imaging

Part of the Lecture Notes in Computer Science book series (LNIP,volume 12964)

Abstract

Fluoroscopic imaging relies on ionizing radiation to provide physicians with high quality video feedback during a surgical operation. Radiation exposure is harmful for both the physician and patient, but reducing dosage results in a much noisier video. We hence propose an algorithm that delivers the same quality video with \(4{\times }\) reduction in radiation dose. Our method is a deep learning approximation to VBM4D, a state-of-the-art video denoiser. Neither VBM4D nor previous deep learning methods are clinically feasible, however, as their high inference runtimes prohibit live display on an operating room monitor. On the other hand, we present a video denoising method which executes orders of magnitude faster while achieving state-of-the-art performance. This provides compelling potential for real-time clinical application in fluoroscopic imaging.

Keywords

  • Fluoroscopic imaging
  • Video denoising
  • Real-time

This is a preview of subscription content, access via your institution.

Buying options

Chapter
USD   29.95
Price excludes VAT (USA)
  • DOI: 10.1007/978-3-030-88552-6_11
  • Chapter length: 11 pages
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
eBook
USD   44.99
Price excludes VAT (USA)
  • ISBN: 978-3-030-88552-6
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
Softcover Book
USD   59.99
Price excludes VAT (USA)
Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.

References

  1. Arias, P., Morel, J.M.: Video denoising via empirical bayesian estimation of space-time patches. J. Math. Imag. Vis. 60(1), 70–93 (2018)

    MathSciNet  CrossRef  Google Scholar 

  2. Balter, S., Hopewell, J.W., Miller, D.L., Wagner, L.K., Zelefsky, M.J.: Fluoroscopically guided interventional procedures: a review of radiation effects on patients’ skin and hair. Radiology 254(2), 326–341 (2010)

    CrossRef  Google Scholar 

  3. Cesarelli, M., Bifulco, P., Cerciello, T., Romano, M., Paura, L.: X-ray fluoroscopy noise modeling for filter design. Int. J. Comput. Assist. Radiol. Surg. 8(2), 269–278 (2013)

    CrossRef  Google Scholar 

  4. Dabov, K., Foi, A., Katkovnik, V., Egiazarian, K.: Image denoising by sparse 3-d transform-domain collaborative filtering. IEEE Trans. Image Process. 16(8), 2080–2095 (2007)

    MathSciNet  CrossRef  Google Scholar 

  5. Davy, A., Ehret, T., Morel, J.M., Arias, P., Facciolo, G.: Non-local video denoising by CNN. arXiv preprint arXiv:1811.12758 (2018)

  6. Dosovitskiy, A., et al.: FlowNet: learning optical flow with convolutional networks. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2758–2766 (2015)

    Google Scholar 

  7. Hoffman, D.A., Lonstein, J.E., Morin, M.M., Visscher, W., Harris III, B.S., Boice, J.D., Jr.: Breast cancer in women with scoliosis exposed to multiple diagnostic X rays. JNCI J. Natl. Cancer Inst. 81(17), 1307–1312 (1989)

    Google Scholar 

  8. Huda, W.: Kerma-area product in diagnostic radiology. Am. J. Roentgenol. 203(6), W565–W569 (2014)

    CrossRef  Google Scholar 

  9. Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)

  10. Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. In: Advances in Neural Information Processing Systems, pp. 1097–1105 (2012)

    Google Scholar 

  11. Maggioni, M., Boracchi, G., Foi, A., Egiazarian, K.: Video denoising, deblocking, and enhancement through separable 4-d nonlocal spatiotemporal transforms. IEEE Trans. Image Process. 21(9), 3952–3966 (2012)

    MathSciNet  CrossRef  Google Scholar 

  12. Maggioni, M., Huang, Y., Li, C., Xiao, S., Fu, Z., Song, F.: Efficient multi-stage video denoising with recurrent spatio-temporal fusion. arXiv preprint arXiv:2103.05407 (2021)

  13. Mastrangelo, G., Fedeli, U., Fadda, E., Giovanazzi, A., Scoizzato, L., Saia, B.: Increased cancer risk among surgeons in an orthopaedic hospital. Occup. Med. 55(6), 498–500 (2005)

    CrossRef  Google Scholar 

  14. NVIDIA: Tensorrt open source software (2018). https://developer.nvidia.com/tensorrt

  15. Paszke, A., et al.: Automatic differentiation in PyTorch (2017)

    Google Scholar 

  16. Primak, A.N., McCollough, C.H., Bruesewitz, M.R., Zhang, J., Fletcher, J.G.: Relationship between noise, dose, and pitch in cardiac multi-detector row CT. Radiographics 26(6), 1785–1794 (2006)

    CrossRef  Google Scholar 

  17. Prince, J.L., Links, J.M.: Medical Imaging Signals and Systems. Pearson Prentice Hall, Upper Saddle River (2006)

    Google Scholar 

  18. Rampersaud, Y.R., Foley, K.T., Shen, A.C., Williams, S., Solomito, M.: Radiation exposure to the spine surgeon during fluoroscopically assisted pedicle screw insertion. Spine 25(20), 2637–2645 (2000)

    CrossRef  Google Scholar 

  19. Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28

    CrossRef  Google Scholar 

  20. Sampat, M.P., Wang, Z., Gupta, S., Bovik, A.C., Markey, M.K.: Complex wavelet structural similarity: a new image similarity index. IEEE Trans. Image Process. 18(11), 2385–2401 (2009)

    MathSciNet  CrossRef  Google Scholar 

  21. Sarno, A., et al.: Real-time algorithm for poissonian noise reduction in low-dose fluoroscopy: performance evaluation. Biomed. Eng. Online 18(1), 1–21 (2019)

    CrossRef  Google Scholar 

  22. Slovis, T.L.: Children, computed tomography radiation dose, and the as low as reasonably achievable (ALARA) concept. Pediatrics 112(4), 971–972 (2003)

    CrossRef  Google Scholar 

  23. Tang, X., Zhen, P., Kang, M., Yi, H., Wang, W., Chen, H.B.: Learning enriched features for video denoising with convolutional neural network. In: 2020 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS), pp. 248–251. IEEE (2020)

    Google Scholar 

  24. Tassano, M., Delon, J., Veit, T.: DVDNet: a fast network for deep video denoising. In: 2019 IEEE International Conference on Image Processing (ICIP), pp. 1805–1809. IEEE (2019)

    Google Scholar 

  25. Tassano, M., Delon, J., Veit, T.: FastDVDNet: towards real-time deep video denoising without flow estimation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 1354–1363 (2020)

    Google Scholar 

  26. Wu, S., Xu, J., Tai, Y.-W., Tang, C.-K.: Deep high dynamic range imaging with large foreground motions. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11206, pp. 120–135. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01216-8_8

    CrossRef  Google Scholar 

  27. Zhou, K.-H., Luo, C.-F., Chen, N., Hu, C.-F., Pan, F.-G.: Minimally invasive surgery under fluoro-navigation for anterior pelvic ring fractures. Indian J. Orthopaedics 50(3), 250–255 (2016). https://doi.org/10.4103/0019-5413.181791

    CrossRef  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Dave Van Veen .

Editor information

Editors and Affiliations

Appendix

Appendix

Fig. 5.
figure 5

Training pair simulation using the method outlined in Fig. 1, which inputs the normal dose video and outputs both a ground truth and a low dose video. We note the noise map is free of structure and artifacts.

Table 3. Five-point Liekert scale used for reader study clinical evaluation.
Table 4. Reader study scores. Positive mean scores denote our output is superior, which holds true for each category. We use this data to create a power analysis (95% confidence, 80% power) to estimate the number of paired low/normal dose samples required for a superiority test, n = 61 in the worst case. While clinical acquisition was beyond the scope of this preliminary work, we plan to collect this data in the future to pursue further clinical validation.
Fig. 6.
figure 6

Single-frame results from a fluoroscopy video. Metrics are calculated with respect to ground truth.

Rights and permissions

Reprints and Permissions

Copyright information

© 2021 Springer Nature Switzerland AG

About this paper

Verify currency and authenticity via CrossMark

Cite this paper

Van Veen, D. et al. (2021). Real-Time Video Denoising to Reduce Ionizing Radiation Exposure in Fluoroscopic Imaging. In: Haq, N., Johnson, P., Maier, A., Würfl, T., Yoo, J. (eds) Machine Learning for Medical Image Reconstruction. MLMIR 2021. Lecture Notes in Computer Science(), vol 12964. Springer, Cham. https://doi.org/10.1007/978-3-030-88552-6_11

Download citation

  • DOI: https://doi.org/10.1007/978-3-030-88552-6_11

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-030-88551-9

  • Online ISBN: 978-3-030-88552-6

  • eBook Packages: Computer ScienceComputer Science (R0)