Computed tomography (CT) scans are performed both with and without intravenous (IV) and/or enteric contrast, drugs used to improve the discrimination of vascular structures, improve tissue characterization, search for bowel perforation or fistulae, and many related applications. IV contrast is nephrotoxic, especially in patients with impaired renal function. Consequently, many patients cannot or should not have IV contrast, so non-contrast CT (NC-CT) scans are often performed, despite their intrinsic limitations, which include poor contrast-to-noise properties. There is a critical and immediate need to improve the diagnostic performance of such unenhanced scans. The long-term goal of this project is to overcome the limitations of NC-CT using statistically principled image reconstruction optimized for specific applications. The intrinsic contrast-to-noise will be increased by reducing noise while preserving resolution using projection-domain smoothing and restoration with explicit models of measurement statistics. Successful completion of this work may provide superior CT imaging performance for patients with chronic kidney disease, contrast allergy, obesity, and for screening applications where radiation dose is limited. There are 4 specific aims: (1) An instrument-specific mathematical model will be synthesized using the statistical and physical properties of a particular clinical CT scanner. (2) Penalized-likehood sinogram restoration methods will be implemented on NC-CT exams. (3) The NC-CT image reconstruction strategy will be implemented on dedicated hardware to achieve clinically useful reconstruction times. (4) The penalized-likelihood sinogram restoration strategy will be optimized to deliver tailored algorithms for specific NC-CT applications and tested on a database of clinical cases. On completion, this project will provide a validated means to reconstruct non-contrast CT scans with significantly improved signal-to-noise and contrast-to-noise, thereby improving the diagnostic performance for emergency examinations. The system will be flexible and clinically feasible for multicenter testing in selected applications. The reconstruction methods will be optimized for the most promising applications, and preliminary measurements of diagnostic performance will be available. Project Narrative The long term goal of this project is to develop and apply statistically principled image reconstruction approaches for non-contrast computed tomography (CT). The intrinsic contrast-to-noise will be increased by reducing noise while preserving resolution using projection-domain smoothing and restoration with explicit models of measurement statistics.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Research Project (R01)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-SBIB-P (02))
Program Officer
Baker, Houston
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Chicago
Schools of Medicine
United States
Zip Code
Bindschadler, Michael; Modgil, Dimple; Branch, Kelley R et al. (2014) Simulation Evaluation of Quantitative Myocardial Perfusion Assessment from Cardiac CT. Proc SPIE Int Soc Opt Eng 9033:903303
Fu, Geng; Meng, Ling-Jian; Eng, Peter et al. (2013) Experimental demonstration of novel imaging geometries for x-ray fluorescence computed tomography. Med Phys 40:061903
Petschke, Adam; La Riviere, Patrick J (2013) Comparison of photoacoustic image reconstruction algorithms using the channelized Hotelling observer. J Biomed Opt 18:26009
Vargas, Phillip A; La Riviere, Patrick J (2011) Comparison of sinogram- and image-domain penalized-likelihood image reconstruction estimators. Med Phys 38:4811-23
La Riviere, Patrick; Vargas, Phillip; Xia, Dan et al. (2010) Region of interest reconstruction in x-ray fluorescence computed tomography for negligible attenuation. IEEE Trans Nucl Sci 57:234-241
Forthmann, Peter; Koehler, Thomas; Defrise, Michel et al. (2010) Comparing implementations of penalized weighted least-squares sinogram restoration. Med Phys 37:5929-38