Takuya Mitsumoto, Presenter: Nothing to Disclose

Yoshiaki Taguchi, Abstract Co-Author: Nothing to Disclose

Fumio Sunaoka, Abstract Co-Author: Nothing to Disclose

Shingo Kawaguchi, Abstract Co-Author: Nothing to Disclose

Takashi Sato, Abstract Co-Author: Nothing to Disclose

Ryogo Minamimoto MD, PhD, Abstract Co-Author: Nothing to Disclose

We compared the maximum and peak standard uptake values (SUV max, SUV peak) measured from PET image reconstructed with 3D-OSEM (HD) and 3D-OSEM-PSF (HD-S), using NEMA body phantom. The purpose of this study was to determine effect of PSF correction base on SUV max and SUV peak.

The PET/CT device was performed using a Discovery PET/CT 600. The NEMA/IEC body phantom and the scatter phantom were used according to NEMA NU-2 2007. The hot spots were filled with 18F-FDG of 20 kBq/ml and the background radioactivity was set at 5 kBq/ml so that the ratio of radioactivity concentration of the hot spots and background would be 4:1. The radioactivity of the scatter phantom was set at 100 MBq. The reconstruction images extracted both HD and HD-S, iteration number of reconstruction parameter was changed from 1 to 19 at the constant subset number of 16. We measured SUV max and the SUV peak. The SUV peak assumed 1.2cm diameter region of interest that placed on the center of sphere.  

The value of SUV max and SUV peak were significantly improved using HD-S as compared with HD. HD-S had higher SUV values related to the resolution recovery and the edge artifacts, that enhance the image values near the sharp edges in the image. The SUV max from HD-S overestimated the value (up to 13% and 15% measured with the 17 mm and 22 mm sphere). In contrast, in the SUV peak, the effect of the overcorrection was not observed.

HD requires enormous computational complexity for signal recovery, but HD-S can realize the improvement of parameter condition, and better clinical usage. However, SUV max overestimated the SUV value with PSF correction.

When PSF correction is applied in a diagnosis, we recommend SUV peak in order that a problem of the overcorrection occurs in SUV max.

Mitsumoto, T, Taguchi, Y, Sunaoka, F, Kawaguchi, S, Sato, T, Minamimoto, R, Validation in SUV Max and SUV Peak Values with PSF-based PET Image Reconstruction: Influence of Reconstruction Iterations.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12033777.html