Abstract
Background: The method of Segmentation with Scatter and Photopeak window data in Attenuation Correction (SSPAC) recognizes the outlines of the body around chest and lungs using scatter window data after which an attenuation correction (μ) map can be constructed. We have developed a new extraction method, adding a masking process to SSPAC, because the extraction of outlines was otherwise incomplete.
Methods and Results: The masking process extracted right and left lung fields from chest images.The quality of the masking process was confirmed by the results from a low count phantom.In a case study, automatic extraction by SSPAC had a 44% success rate for low count (stress condition) myocardial single-photon emission computed tomography (SPECT) images, but reached a success rate of 99% with the addition of the new masking process. Outline truncation and low counts can cause unsuccessful SSPAC.
Conclusions: Our method for masking will contribute to a widespread use of SSPAC by improving the success rate in contouring.
