Mammograms restoration by using Wiener filter
- University of Montenegro, Faculty of Natural Sciences and Mathematics, Dž. Vašingtona bb, 20000 Podgorica (Montenegro)
- Clinical Center of Montenegro, 20000 Podgorica (Montenegro)
Restoration of digital mammograms, as a pre-processing tool, using deconvolution procedures, is analysed. It implies, knowing Modulation Transfer Function (MTF) of the mammography device and the estimation of the mammogram’s noise. Wiener filter is used, as the most objective in mammograms restoration by deconvolution. Using MATLAB program the deconvolution procedures are conducted in two ways with different level of approximation. The first method approximates the noise/signal power ratio by a constant and the second method uses autocorrelation functions of the noise and signals. Abilities and limitations of the methods are analysed and checked by using the raw images of the bar-pattern due to better visualisation of the obtained results. The raw images are obtained at a Computed Radiography (CR) mammography device in Clinical Centre in Podgorica. It is found that quality of the restored image highly depends of knowledge of type and magnitude of noise. In the both methods spatial resolution of the restored images are improved, but much better in a case where autocorrelation functions of the noise and signal are used. This procedure is proposed as an objective pre-processing tool to put back imperfectness of mammography devices.
- OSTI ID:
- 22591046
- Journal Information:
- AIP Conference Proceedings, Journal Name: AIP Conference Proceedings Journal Issue: 1 Vol. 1722; ISSN APCPCS; ISSN 0094-243X
- Country of Publication:
- United States
- Language:
- English
Similar Records
Practical considerations of the Wiener filtering technique on projection data for PET
Two-dimensional filtering of SPECT images using the Metz and Wiener filters