Abstract
Although the novel respiratory-gated PET imaging is a commercially available
technique, no details of its accuracy and its optimal parameter have been reported. Therefore,
the accuracy and optimal acquisition window size were evaluated using a dynamic motion
phantom which can simulate breathing movements of the human body. A PET scan was
performed using a moving NEMA/IEC body phantom with six hot spheres（ diameter: 10 ─
37 mm） by changing the acquisition window size as an index of the respiration phase. In the
reconstructed PET images obtained from different window sizes, the image quality was
improved when the window size was narrow. Especially, the improvement of image quality
for the 10 mm sphere was significantly higher than that of the other hot spheres. However,
the statistical noise contained in the acquisition PET data was increased when the window
size was narrow. The optimal acquisition window size was 25％ of one breathing cycle.
