Optimization and Performance Evaluation of the Median Modified Wiener Filter Algorithm in Breast Electromagnetic X-ray Image
- Authors
- Mo, Yunhee; Kang, Seong-Hyeon; Lee, Youngjin
- Issue Date
- Dec-2023
- Publisher
- KOREAN MAGNETICS SOC
- Keywords
- Breast X-ray image; Mammography; Median modified Wiener filter (MMWF); Optimization of mask size; Quantitative evaluation of image qualities
- Citation
- JOURNAL OF MAGNETICS, v.28, no.4, pp 392 - 400
- Pages
- 9
- Journal Title
- JOURNAL OF MAGNETICS
- Volume
- 28
- Number
- 4
- Start Page
- 392
- End Page
- 400
- URI
- https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/91457
- DOI
- 10.4283/JMAG.2023.28.4.392
- ISSN
- 1226-1750
2233-6656
- Abstract
- The mask size of the median modified Wiener filter (MMWF) algorithm was optimized for application to digital mammographic phantom images obtained under different tube currents to obtain an optimal image by removing noise with a minimal dose of radiation. A simulation was conducted using the Female Adult Mesh (FASH) phantom to optimize the mask size of the MMWF algorithm. The optimal mask size was determined by measuring various evaluation factors and applying different mask sizes to the noisy image of the various evaluation phantom. An American College of Radiology (ACR) phantom image was obtained per milliamperes using digital mammography equipment to evaluate the MMWF algorithm with the optimal mask size. Based on the noise level and similarity evaluation factors, the optimal mask size for the MMWF algorithm was 7 x 7. The results of the experiment indicated that improvements in the noise level evaluation factors were most noticeable for the MMWF, followed by the median, Gaussian, and Wiener filters. Overall, we confirmed that the optimal mask size for the MMWF algorithm was 7 x 7 for application to digital mammographic images, and its effectiveness was proven through a comparative evaluation with conventional filters.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - ETC > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.