Share Email Print
cover

Proceedings Paper

Noise performance studies of model-based iterative reconstruction (MBIR) as a function of kV, mA and exposure level: Impact on radiation dose reduction and image quality
Author(s): Daniel Gomez-Cardona; Ke Li; Meghan G. Lubner; Perry J. Pickhardt; Guang-Hong Chen
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The significance of understanding the noise properties of clinical CT systems is twofold: First, as the diagnostic performance (particularly for the detection of low contrast lesions) is strongly limited by noise, a thorough study of the dependence of image noise on scanning and reconstruction parameters would enable the desired image quality to be achieved with the least amount of radiation dose; Second, a clear understanding of the noise properties of CT systems would allow the limitations in existing CT systems to be identified and improved. The recent introduction of the model-based iterative reconstruction (MBIR) method has introduced strong nonlinearity to clinical CT systems and violated the classical relationship between CT noise properties and CT system parameters, therefore it is necessary to perform a comprehensive study on the noise properties of MBIR. The purpose of this study was to systematically study the dependence of the noise magnitude and noise texture of MBIR on x-ray tube potential (kV), tube current (mA), and radiation dose level. It has been found that the noise variance σ2 of MBIR has relaxed dependence on kV and mA, which can be described as power-law relationships as σ2 ∝kV−1 and σ2 ∝ mA−0.4, respectively. The shape of the noise power spectrum (NPS) demonstrated a strong dependence on kV and mA, but it remained constant as long as the radiation dose level was the same. These semi-empirical relationships can be potentially used to guide the optimal selection of kV and mA when prescribing CT scans with the maximal dose reduction.

Paper Details

Date Published: 18 March 2015
PDF: 7 pages
Proc. SPIE 9412, Medical Imaging 2015: Physics of Medical Imaging, 941238 (18 March 2015); doi: 10.1117/12.2082334
Show Author Affiliations
Daniel Gomez-Cardona, Univ. of Wisconsin-Madison (United States)
Ke Li, Univ. of Wisconsin-Madison (United States)
Meghan G. Lubner, Univ. of Wisconsin-Madison (United States)
Perry J. Pickhardt, Univ. of Wisconsin-Madison (United States)
Guang-Hong Chen, Univ. of Wisconsin-Madison (United States)


Published in SPIE Proceedings Vol. 9412:
Medical Imaging 2015: Physics of Medical Imaging
Christoph Hoeschen; Despina Kontos, Editor(s)

© SPIE. Terms of Use
Back to Top