J Nucl Med. 2008; 49 (Supplement 1):159P
InfoSNM Computer PresentationsInfoSNM III |
Software tools for standardized analysis of FDG whole body studies in multi-center trials
Ronald Boellaard1,
Otto Hoekstra1 and
Adriaan Lammertsma1
1 Nuclear Medicine & PET Research, VU University Medical Centre, Amsterdam, Netherlands
661
Learning Objectives: To understand the benefits of standardized automated analysis of FDG PET studies in multi-center trials.
Abstract (summary): Exchange of SUV in multi-center trials is hampered by differences in methodology used. Therefore, in the Netherlands, a protocol for standardization of FDG whole body scans was developed (NEDPAS). As part of this standardization, volume of interest (VOI) software tools were developed in order to standardize VOI definition and SUV calculation. The programs developed allow for dicom import of images from all main PET scanner manufacturers. Special care was taken to correctly read scaling and calibration factors from dicom file headers. After importing a whole body scan, it can be viewed in all 3 orientations with various zoom factors. The user can identify a tumor by ‘left clicking’ on the tumor. Next, the program will automatically generate 3D VOI as defined in NEDPAS, i.e. 41, 50 and 70% isocount contours with and without local background correction. Automatic VOI generation can be restricted to within a user specified volume to avoid inclusion of non-tumor voxels. SUV normalized for body weight, BSA and LBM (with and without correction for plasma glucose level) will be calculated automatically and written to an ascii file. Using the tool, SUV is obtained in a semi-automatic and reproducible way with minimal user intervention, thereby minimizing human errors and/or variability. All relevant data are written to a file, making it easy to check that correct data were used in the calculations, such as patient information, injected dose, acquisition and calibrationt times, VOI position and corrections applied. The software tool allows for reproducible and consistent VOI definitions and SUV calculations across all institutes participating in multi-center trials.
