Optimization of attenuation and scatter corrections in sentinel lymph node scintigraphy using SPECT/CT systems.

OBJECTIVE Although SPECT/CT systems have been used for sentinel lymph node (SLN) imaging, few studies have focused on optimization of attenuation correction (AC) and scatter correction (SC). While SLNs could be detected in conventional planar images, they sometimes do not appear in SPECT/CT images. The purpose of this study was to investigate the optimal AC and SC and to improve the detectability of SLNs in examinations using SPECT/CT systems. MATERIALS AND METHODS The study group consisted of 56 female patients with breast cancer. In SPECT/CT imaging, four kinds of images were created with and without AC and SC; namely, AC-SC-, AC+SC-, AC-SC+ and AC+SC+. Five nuclear medicine physicians interpreted the planar and SPECT/CT images with five grades of confidence levels (1-5). The detection rate was calculated as the number of patients whose average confidence levels of interpretation were more than 4, divided by the total number of patients. RESULTS The confidence level of interpretation and the detection rate provided by the planar images were 4.76 ± 0.49 and 94.6 %, respectively. In SPECT/CT imaging, the AC+SC- provided the best detection rate (confidence level of interpretation, 4.81 ± 0.38; detection rate, 98.2 %), followed by the AC-SC- (4.70 ± 0.55, 89.3 %), and the AC-SC+ (4.39 ± 1.2, 78.6 %). The lowest values were obtained for the AC+SC+ (4.36 ± 1.22, 78.6 %). Regarding the confidence levels of interpretation, significant differences were observed between AC+SC- and AC-SC-, AC+SC- and AC+SC+, AC+SC- and AC-SC+, and between planar images and AC+SC+ (P = 0.0021, 0.0009, 0.0013, and 0.0056, respectively). CONCLUSIONS When SPECT/CT was used, AC improved the detection of SLNs. SC caused disappearance of a faint SLN in some cases and should not be performed.