Attenuation correction vs non-attenuation correction in detecting lesions of normal and overweight patients in 18F-FDG coincidence imaging

2120 Objectives : Attenuation is a major problem in nuclear imaging and has a particularly significant impact on PET imaging. The loss of counts due to attenuation can exceed 50% to 95% which increases image noise, artifacts, and distortion in a large patient. The purpose of this study is to compare the non-attenuation corrected (NAC) tomographic coincidence imaging with arithmetic attenuation-corrected (AC) tomographic coincidence imaging for lesion detection and correlate these findings with body mass index (BMI). Methods : 10 cases (6 male, 4 female) of normal weight (BMI Results : In the normal weight cases, NAC images were rated "better resolution" in 67%. The total number of lesions counted between both physicians among all images given was 107, and there were no significant discrepancies. In the overweight cases, NAC images were rated "better resolution" in only 54%. The total number of lesions counted was 72, with an inter-reader variability of 5 lesions. However, there was no significant discrepancy in lesion detection between AC and NAC techniques among individual readers. Conclusions : Overall, NAC images were perceived to have better resolution. Our results demonstrate that there is no significant difference in lesion detection between AC and NAC whole body linear tomographic coincidence imaging in normal or overweight cancer patients.