Wavelet-based resolution recovery using anatomical prior provides quantitative recovery for human population phantom PET [<sup>11</sup>C]raclopride data

Abstract

The purpose of this study was the evaluation of a wavelet-based resolution recovery (RR) method, named structural and functional synergy for RR (SFS-RR), for a variety of simulated human brain [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C]raclopride PET images. Simulated datasets of 15 human phantoms were processed by SFS-RR using an anatomical prior. This anatomical information was in the form of a hybrid segmented-atlas, which combines an MRI for anatomical labelling and a PET image for functional labelling of each anatomical structure. First, the relationship between the FWHM of the original the PET image and its resolution recovered version was investigated. Then quantitative evaluation was performed by comparing caudate, putamen and cerebellum regions of the true image; simulated PET image; and RR image. The spatial resolution of the original PET image effected on how accurately SFS-RR recovers the image counts in striatum regions. The resolution in striatum, and cerebellum regions was successfully recovered for all the 15 human phantoms. The proposed methodology proved effective in the resolution recovery of small structures of brain [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup> C]raclopride PET images. The improvement was consistent across the anatomical variability of a simulated population, provided accurate anatomical segmentations are available.