This project will implement an image-domain vendor-independent dose-reduction simulator that will become a valuable tool to reduce the risk of unnecessary radiation from computed tomography (CT) scans, while preserving diagnostic image quality. CT usage has increased substantially in the past two decades, raising concerns about the risks of ionizing radiation for patients, and leading the radiology community to strive towards the ALARA principle, using radiation exposures that are "As Low as Reasonably Achievable". Progress towards optimizing CT protocols to balance radiation exposure versus image quality has been slow due to the lack of efficient, objective methods to determine the effects of image noise on diagnostic performance. A very promising tool for such studies is a sinogram-domain CT dose-reduction simulator, which adds synthetic noise to sinogram data to create an image that corresponds to a lower-dose scan. To date, such simulators are not generally available, utilize nonstandard input file formats, and require knowledge of proprietary scanner properties. The goal of this application is to develop an image-domain CT dose-reduction simulator that can work with any CT vendor's scanner, using standard images that are available on clinical networks. To accomplish this, we propose a novel technique: 1) forward-project an image to form an intermediary sinogram;2) calculate sinogram noise magnitudes and use a random number generator to form synthetic noise;3) perform a filtered-backprojection of sinogram noise to create image noise;4) add this noise to the original image to simulate a low-dose scan. The critical gap in knowledge for implementing such a simulator is a method to estimate internal scanner parameters (bowtie profile, reconstruction kernel shape, tube-current modulation, flux scaling, and system noise) in the absence of proprietary information. Our research in modeling noise properties of CT scanners has recently led to novel and innovative concepts, based on measuring the propagation of noise during acquisition and image reconstruction that allow us to estimate these unknown proprietary parameters through a series of baseline images acquired on a given system. In this project we propose two specific aims:
Aim 1 - Implement methods to estimate CT scanner parameters;
Aim 2 - Develop and test vendor-independent image-domain CT dose-reduction simulation software. Successful completion of these aims will provide a tool that can allow objective evaluation of ALARA protocols, establish target reference noise limits for clinical practice, and train/educate radiologists to acclimate to low-dose images that do not impair observer performance. Our team is uniquely positioned to lead this effort. Optimizing radiation dose in CT protocols is a significant healthcare challenge recently recognized by the public, the FDA, Congress, and the NIH, in which dose-reduction simulators can play a major role if available in the clinical community. Our long-term objective is to optimize CT protocols based on scientific, objective methods for reducing radiation exposure without loss of diagnostic accuracy, and to provide educational tools to disseminate ALARA procedures into the radiological community. This vendor-independent image-domain CT dose-reduction simulator would represent a significant step towards optimizing CT protocols, allowing researchers and clinicians access too much needed tools.

Public Health Relevance

Computed tomography (CT) scans provide great diagnostic benefit but involve the risk of radiation exposure. The vendor-independent, dose-reduction CT simulator proposed in this project would enable the establishment of protocols that optimize this image quality-dose tradeoff, serve as a training tool for radiologists to acclimate to the use of low-dose images, and make possible the practice of CT imaging with radiation usage as low as reasonably achievable.

National Institute of Health (NIH)
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
Exploratory/Developmental Grants (R21)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-SBIB-J (80))
Program Officer
Pai, Vinay Manjunath
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Pittsburgh
Schools of Medicine
United States
Zip Code
Whiting, Bruce R; Evans, Joshua D; Dohatcu, Andreea C et al. (2014) Measurement of bow tie profiles in CT scanners using a real-time dosimeter. Med Phys 41:101915