Model-Based Image Reconstruction for X-Ray CT in Lung Imaging
To develop a computer program that will improve CT image quality and decrease the amount of x-ray radiation that future patients may be exposed to when they have a CT examination.
|Study Design:||Endpoint Classification: Efficacy Study
Intervention Model: Single Group Assignment
Masking: Single Blind (Investigator)
Primary Purpose: Basic Science
|Official Title:||Model-Based Image Reconstruction for X-Ray CT in Lung Imaging|
- Developement of a computer program to improve CT image quality [ Time Frame: 6 years ] [ Designated as safety issue: Yes ]
Sinograms will be retrieved by an archive system and processed by MBIR (model-based image reconstruction) methods that we are developing that improve image quality (reduce noise, improve spatial resolution, reduce artifacts). We will evaluate the image quality both quantitatively and qualitatively.
We hope to develop and benchmark methods for algorithm acceleration to enable routine clinical use of MBIR (model-based image reconstruction) methods.
|Study Start Date:||August 2010|
|Estimated Study Completion Date:||December 2016|
|Estimated Primary Completion Date:||August 2015 (Final data collection date for primary outcome measure)|
Experimental: CT Imaging and Reconstruction
To develop a MBIR (model-based image reconstruction) method that will improve X-ray CT lung imaging by improving image quality and reducing dose.
Procedure: CT Imaging and Reconstruction
Patients will be consented for their permission to save and use the sinogram from their CT chest/lung scan. The sinogram will be de-identified and sent to an archive system for storage.
Later it will be exported to a computer for processing using MBIR (model based image reconstruction). The newly processed images will then be read by blinded readers. The quality of the images will be reviewed to determine if they are as readable and accurate as CT images created with the software that is currently being used.
We will be asking patients for their permission to save and use the sinogram from their CT scan. The sinogram will be de-identified and sent to an archive system for storage. It will be exported to a computer for processing using MBIR (model based iterative reconstruction). MBIR (model based iterative reconstruction) is a new method being developed to process CT sinograms. The newly reconstructed images will be reviewed by experts to determine if they are as readable and accurate as CT images created with the software that is currently being used. Sinogram data and the reconstructed images will be shared with collaborating researchers at General Electric Global Research.
|Contact: Jeffrey Fessler, PhD||734-763-1434|
|Contact: James Pool, Jr., CCRCfirstname.lastname@example.org|
|United States, Michigan|
|University of Michigan Hospital||Recruiting|
|Ann Arbor, Michigan, United States, 48109|
|Contact: Jeffrey Fessler, PhD 734-763-1434|
|Contact: James Pool Jr., CCRC 734-615-7391 email@example.com|
|Principal Investigator: Jeffrey Fessler, PhD|
|Principal Investigator:||Jeffrey Fessler, PhD||University of Michigan Hospital|
|Principal Investigator:||Jeffrey Fessler, PhD||University of Michigan|