Phantom attenuation research of 2 detectors SPECT with coincident

Objective The photons from the body would be attenuated in their intensity by the soft tissue before they reach the detectors. Some of them lost more intensity than others when coming from the deeper body. Attenuation of the photons would distort the image, affect the image quality, and may lead to misdiagnosis. In this research the X-ray and γ-ray were used as transmission resources for accurate assess-ment of attenuation in each part of the body. The transmission data was used to correct the emission data from the body. Iteration algorithm was used for the reconstruction of attenuation corrected tomography ima-ges. Evaluate the efficacy of attenuation correction in duel head hybrid SPECT/coincidence PET systems (SPECT/PET) and SPECT using both the Jaszczak and the myocardial phantoms. Methods For PET phantom study, the Jaszczak phantom was filled with the aqueous solution of 18F-fluorodeoxyglucose (FDG) and was scanned with two hybrid SPECT/coincidence PET systems respectively (Vertex Plus, Netherlands Philips ADAC Company and Hawkeye, USA GE Company). The PET images were acquired and reconstruc-ted. Results The lesion, which the diameter was less than 12 mm of uncorrected image, could not be identified. But it, after correction, diameter 9 mm, would be recognized clearly, less than 7 mm would not be distinguished. Noncorrected image could only identify lesions which the diameter were 9, 12, 14, 18, 22 and 38 mm, while in the corrected image, not only it was much better, but also the le-sion of 6 mm was distinguished. In Hawkeye, the same situation, the corrected image of was improved obvi-ously; the disparity of counts between surrounding and deep center less, and the counts in deep center were compensated. There were only 3 round cold lesions could be recognized, the diameter: 14, 18 and 20 mm; all of them which diameter were less than 12 mm could not be distinguished. One in the center was al-most as background, and could not be identified. The radioactive accounts of surrounding and center were quite different. The corrected image 5 lesions could be found, 9, 12, 14, 18 and 20 mm, but7 mm could not be found. The noncorrected image, only 12, 14, 18, 22 and 38 mm hot lesions could be found, and by the surrounding, there was an arc shape lesion. But it could distinguish 9, 12, 14, 18, 22 and 38 mm lesions on the corrected image, and the image was much better than noncorrected one. Conclusions The corrected image of Vertex Plus was improved obviously with the disparity of counts between surrounding and deep center less, and the counts in deep center was compensated. The result of at-tenuation with phantom showed that 137Cs was better than CT in both image quality and spatial resolution. Key words: Tomography, emission-computed, single-photon;  Coincidence;  Quality control