Development of Computer Techniques Designed to Aid Tests of Digital Mammography Systems Quality Evaluation with the Phantom CDMAM 3.4

摘要

Image quality in digital mammography can be achieved by conducting periodic tests. It is recommended that some quality parameters be measured from images acquired by exposing specific phantoms, as CDMAM 3.4, in such systems. Whereas this task is hard-working and time consuming, this study has attempted to develop and compare two computational methods in order to assist the technical professional in performing the tests with such phantom images, reducing the subjectivity due to the observers. Tests used 27 phantom images obtained from six different digital mammography systems -five CR-type and one DR-type. Both methods proved to be effective in detecting structures. However, the first one allowed the complete image processing and not just a region of interest, which makes it quite advantageous.