RT Journal Article
SR Electronic
T1 Positron range correction with S-deconvolution
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 1482
OP 1482
VO 50
IS supplement 2
A1 Richard Laforest
A1 Xiaodong Liu
YR 2009
UL http://jnm.snmjournals.org/content/50/supplement_2/1482.abstract
AB 1482 Objectives In high resolution small animal PET imaging, positron range is an important factor affecting quantitative accuracy. This work explores the performance of a sinogram restoration algorithm aiming at recovering the loss of resolution from the camera geometry and physics of positron decay. The technique employs the inverse S-transform to perform deconvolution on the sinograms by a spatially variant resolution kernel that accounts for the detector response, photon acolinearity and the positron range. Methods Spatial resolution kernels were determined for long range positron emitting nuclides from experimentally measured point spread function in water and converted to projection space. The S-transform is a non-iterative one step approach based on a piecewise Taylor expansion of the projection data. The technique was applied to various phantoms filled with Br-76 and to small animal data. The technique operates by first de-blurring the sinograms (3D sinograms converted to 2D with FORE algorithm) and then followed by the image reconstruction using the standard 2D-FBP algorithm. Performance of this algorithm and its noise-resolution properties is also compared to an iterative penalized maximum likelihood approach (MAP with range correction). Results The sinogram deconvolution technique is efficient at reducing the loss of resolution due to the long positron range. In particular, better spatial resolutions and improved recovery coefficients were observed than from reconstruction from conventional FBP image reconstruction. Conclusions The S-Transform algorithm is not as efficient as MAP with range correction at recovering the loss of resolution although substantial resolution recovery is achieved. The main advantage of this technique is speed. Research Support This work was made possible from the financial support from the RadioNuclide Resource for Cancer Applications (RRCA), NIH grant R24 CA86307.