Share Email Print

Proceedings Paper

Contrast improvements in digital radiography using a scatter-reduction processing algorithm
Author(s): Kent M. Ogden; Charles R. Wilson; Robert W. Cox
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

An expectation maximization (EM) algorithm has been developed based on a model of radiographic imaging that accounts for scatter radiation and resolution degradation. Digital radiographs of a chest phantom were acquired, and the amount of scatter in several regions was computed using the known radiographic exposure and the known material properties. Contrast and noise were measured in a step wedge in the phantom. The phantom images were processed by the EM algorithm for up to 8 iterations, and the image intensity and noise values were measured at each iteration step. These values were used to compute the scatter reduction properties of the algorithm and its effect on the contrast-to-noise ratio. The algorithm removed over 90% of the scattered radiation in the image. Image noise values were reduced an average of 50% in the first iteration, but then increased to values equal or above unprocessed images. The contrast-to- noise ratio was initially increased substantially, but gradually decreased as further iterations caused the image noise to increase. With proper selection of processing parameters, this algorithm could provide considerable qualitative enhancement of clinical images with a single iteration as well as numerically accurate scatter reduction.

Paper Details

Date Published: 9 May 2002
PDF: 14 pages
Proc. SPIE 4684, Medical Imaging 2002: Image Processing, (9 May 2002);
Show Author Affiliations
Kent M. Ogden, SUNY/Upstate Medical Univ. (United States)
Charles R. Wilson, Medical College of Wisconsin (United States)
Robert W. Cox, NIMH/National Institutes of Health (United States)

Published in SPIE Proceedings Vol. 4684:
Medical Imaging 2002: Image Processing
Milan Sonka; J. Michael Fitzpatrick, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?