Xinhui Duan PhD, Presenter: Nothing to Disclose

Cynthia H. McCollough PhD, Abstract Co-Author: Research Grant, Siemens AG

Lack of access to projection data from patient CT scans is a major limitation for development and validation of new reconstruction algorithms for dose reduction. To meet this critical need, we are developing a freely-available library of reference patient data sets, which will include image and projection data. To accomplish, we sought to develop and validate a standardized, vendor-neutral format for CT projection data, which will be used in our reference patient library.

The framework for the projection data format was developed by Battelle Memorial Institute as a proposed extension to current DICOM standards. In this framework, five groups of newly-defined Information Object Definitions were included as DICOM tags, containing essential information about data acquisition necessary for image reconstruction. For this work, only a subset of these tags was needed. Projection data were stored as binary data in a pre-defined order and format. To validate the implementation, CT projection data were acquired from two CT scanners (Siemens Definition Flash and GE Discovery 750HD), where the ACR phantom was scanned in axial and helical modes. After decoding (by us for Siemens, by the manufacturer for GE), the projection data were converted to the DICOM format. An off-line CT reconstruction was performed using only the information stored in the converted data file.

The converted projection data files are compatible with current DICOM standards, so the projection data can be directly viewed using existing DICOM software. The reconstructed images were similar in spatial resolution compared with the manufacturer’s reconstruction, even though different reconstruction approaches were used. This confirms that the geometric parameters used in the reconstruction were accurate and the data were correctly stored and retrieved.

 A standardized format for CT projection data was implemented for use in a reference patient data library for algorithm development and evaluation.

 The successful implementation of an open format of projection data will allow algorithm development and realistic evaluation of dose reduction in CT.

Duan, X, McCollough, C, Implementation of an Open Data Format for CT Projection Data.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14006609.html