13 March 2009 MTF compensation for digital radiography system with indirect conversion flat panel detector
Author Affiliations +
Proceedings Volume 7258, Medical Imaging 2009: Physics of Medical Imaging; 725852 (2009) https://doi.org/10.1117/12.812471
Event: SPIE Medical Imaging, 2009, Lake Buena Vista (Orlando Area), Florida, United States
Abstract
The MTF (modulation transfer function) of digital radiography systems can be enhanced in the spatial frequency domain due to their high signal to noise ratio. A Wiener filter, which requires prior estimation of the noise and signal power spectrum of the images, was used to compensate MTF of the detector and thereby optimally restore the images details. We studied the noise characteristics of two flat panel detectors with structured columnar scintillator (CsI) and granular scintillator (Gd2O2S). A noise model formulating noise transfer process was applied to estimate the noise components for the filter. Signal model was based on dose of the application. We revisited the noise and signal model that was used in previous work by Souchay et al. for mammography application [1], considering the difference in detector characteristics and the applications (extremity x-ray) that we are specifically investigating. Starting with real clinical images, we used an observer study method to measure the visually optimal parameter for the Wiener filter. A set of clinical images was used to evaluate the radiologists' preferences to compensated images against the reference images. Statistical results from three experienced radiologists ranking results show that the compensated images are preferred over the reference images.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Wei Zhang, Henri Souchay, Chao Yang, "MTF compensation for digital radiography system with indirect conversion flat panel detector", Proc. SPIE 7258, Medical Imaging 2009: Physics of Medical Imaging, 725852 (13 March 2009); doi: 10.1117/12.812471; https://doi.org/10.1117/12.812471
PROCEEDINGS
8 PAGES


SHARE
RELATED CONTENT

Low-light NV-CMOS image sensors for day/night imaging
Proceedings of SPIE (May 31 2013)
Minimum resolvable power contrast model
Proceedings of SPIE (January 12 2018)
Noise Sources In Digital Fluorography
Proceedings of SPIE (November 04 1981)
Multidimensional Morphological Edge Detection
Proceedings of SPIE (October 13 1987)

Back to Top