Construction of 18F-FDG and 11C-PiB PET Healthy Control Statistical Databases
Objective: To create databases of both 18F-FDG and 11C-PiB PET images of Healthy Control patients to which acquired images may be compared to using statistical analysis software.
Methods: The image databases consist of 25 healthy controls. The patients have a mean age of 72.3 yrs, have been determined to be PiB negative (defined by a neocortical SUVR less than 1.6), are MRI normal and performed normally when neuropschologically assessed.
The same patients were used in the construction of both FDG and PiB databases.
The PiB images used were summed 30 min frame, 40 min post-injection, 370 MBq dose, and the FDG images were a 15 min frame, 30 min post-injection, 296 MBq. All images were acquired on a Philips Allegro PET system.
The images used were all normalised to a SUV scale.
Due to the high level of PiB uptake in the scalp, it was necessary to segment the PiB images prior to coregistration. The segmented grey matter was coregistered into a standard space and then the transformation matrix was applied to the unsegmented image. Coregistration of the FDG images to the standard space was performed directly.
Once the images were spatially normalised, they were summed to create the image representing the database of these 25 healthy control patients.
Results: The use of these databases in three dimensional spatial normalisation and statistical analysis has enhanced the resources available for both our physicians in a clinical and research setting.