Digital Reference Object
Updated: October 20, 2013

DRO Description

The PET/CT Digital Reference Object (DRO) is a set of synthetically generated DICOM image files of known voxel values for positron emission tomography (PET) and x-ray computed tomography (CT). The PET/CT DRO is intended to test the computation of standardized uptake values (SUVs) by PET/CT display stations. It is also intended to test region of interest (ROI) calculations and alignment between the PET and CT images. This is motivated by vendor-specific variations in PET DICOM formats used for SUVs. The PET/CT DRO was developed at the University of Washington Imaging Research Laboratory and is supported by the Quantitative Imaging Biomarker Alliance (QIBA).

The primary goals and objectives of the PET/CT Digital Reference Object are to support the QIBA FDG-PET 'Technical Validation' efforts for profile development. This will be done by (1) evaluation and validation of SUV calculations in PET images, (2) evaluation and validation of ROI calculations and (3) providing a common reference standard that can be adopted and modified by PET/CT scanner and display station manufacturers.

Network Flow
This image shows the flow of PET/CT DICOM image files in a heterogeneous manufacturer and vendor network. The insertion of digital reference objects (DROs) for testing validity of SUV and ROI analysis is also indicated.

The PET and CT components of the Images of the DRO are each a set of DICOM format files, one file per image slice. Each set of files are typically grouped as a stack to form an image volume. Representative sections through the CT and PET image volumes are shown below. The synthetic test object is based on, but is not identical to, the NEMA NU-2 PET test phantom [J Nucl Med, vol. 43 no. 10 1398-1409, 2002].

Further details are given in the DRO User's Guide.

Representative 2-D sections through the CT and PET DRO image volumes.

Users of the DRO are asked to download the package, import the PET and CT objects into their viewing software, perform region of interest (ROI) analyses, and submit the results back to this website.