Journal of Magnetics

The Korean Magnetics Society (한국자기학회)
권호 표지
DOI QR코드

DOI QR Code

Feasibility Study of Improved Patch Group Prior Based Denoising (PGPD) Technique with Medical Ultrasound Imaging System

  • Kim, Seung Hun (Department of Radiological Science, Eulji University) ;
  • Seo, Kanghyen (Department of Radiological Science, Eulji University) ;
  • Kang, Seong Hyeon (Department of Radiological Science, Eulji University) ;
  • Kim, Jong Hun (Department of Radiological Science, Eulji University) ;
  • Choi, Won Ho (Department of Radiological Science, Eulji University) ;
  • Lee, Youngjin (Department of Radiological Science, Eulji University)
  • Received : 2017.02.02
  • Accepted : 2017.02.22
  • Published : 2017.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

The purpose of this study was to quantitatively evaluate image quality using intensity profile, coefficient of variation (COV), and peak signal to noise ratio (PSNR) with respect to noise reduction techniques in the ultrasound images. For that purpose, we compared with the median filter, Rudin-Osher-Fatemi (ROF), Anscombe and proposed patch group prior based denoising (PGPD) techniques. To evaluate image quality, the Shepp-Logan phantom and the ultrasound image were acquired using simulation and experiment, respectively. According to the results, the difference of intensity profile using PGPD technique is lowest compared with original Shepp-Logan phantom. In simulation, the measured COV was 0.249, 0.198, 0.198, 0.177, and 0.080 using noisy, median, ROF, Anscombe and PGPD technique, respectively. Also, in experimental image, the measured COV was 0.245, 0.230, 0.231, 0.242 and 0.187 using noisy, median, ROF, Anscombe and PGPD technique, respectively. Especially, when we used PGPD technique, the PSNR has highest value in both simulation and experiment. In this study, we performed simulation and experiment study to compare various denoising techniques in the ultrasound image. We can expect the PGPD technique to improve in medical diagnosis with excellent noise reduction.

Keywords

References

  1. G. J. Czarnotta and M. C. Kolios, Imaging in Medicine 2, 17 (1020).
  2. P. Hellier, C. Kervrann, and C. Barillot, IEEE Transactions Image Processing 18, 2221 (2009). https://doi.org/10.1109/TIP.2009.2024064
  3. K. Djemal, IEEE International Conference on Image Processing 3, 357 (2005).
  4. L. I. Rudin, S. Osher, and E. Fatemi, Physica D 60, 259 (1992). https://doi.org/10.1016/0167-2789(92)90242-F
  5. S. Suresh, G. R. Suresh, and R. Sukanesh, International Journal of Computer Theory and Engineering 1, 7 (2009).
  6. A. Buades, B. Coll, and J. M. Morel, Multiscale Modeling & Simulation 4, 490 (2005). https://doi.org/10.1137/040616024
  7. P. Chatterjee and P. Milanfar, IEEE Transactions on Image Processing 19, 895 (2010). https://doi.org/10.1109/TIP.2009.2037087
  8. A. Buades, B. Coll, and J. M. Morel, IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2, 60 (2005).
  9. J. Xu, L. Zhang, W. Zuo, D. Zhang, and X. Feng, The IEEE International Conference on Computer Vision 244 (2015).
  10. C. M. Bishop, Springer, New York (2006).
  11. CF Jeff Wu, The Annals of Statistics 11, 95 (1983). https://doi.org/10.1214/aos/1176346060
  12. D. Isa, Lam H. Lee, V. P. Kallimani, and R. RajKumar, IEEE Transactions on Knowledge and Data Engineering 20, 1264 (2008). https://doi.org/10.1109/TKDE.2008.76
  13. P. Burt and E. Adelson, IEEE Transactions on Communications 31, 532 (1983). https://doi.org/10.1109/TCOM.1983.1095851
  14. L. A. Shepp and B. F. Logan, IEEE Transactions on Nuclear Science 21, 21 (1974).