Research PET / CT Lung Attenuation Correction During Imaging And Segmentation Methods

Positron emission tomography (PET) is an important part of medical imaging technology. With the development of image fusion technology, PET/CT has become a kind of brand-new imaging-anatomical function imaging. PET/CT for early diagnosis of cancer and other diseases, can find the hidden small lesions which have not incurred in tumor anatomical structure change. PET can reflect living biochemical metabolism, and has the characteristics of high specificity and sensitivity, so the degree of accuracy of the PET data and image is very important for disease diagnosis and basic life process of cognition and understanding. However, due to PET technology itself characteristics and reconstruction image limit, which result that PET image resolution and contrast is not high and serious image artifacts, and cause the image uneven and distortion, which can seriously affect the precision and accuracy of PET/CT imaging.One of the most serious impact of PET image quantitative accuracy is tissue attenuation effects in the imaging process. Tissue attenuation will give PET data acquisition great influence, causing artifact and local organizations of radioactive distribution inequality, making the data signal to noise ratio (SNR) and the precision of the reconstruction image to reduce. In order to obtain the real radioactive distribution information of tested content object, in the image reconstruction must organize attenuation correction. The traditional PET image attenuation correction method used transmission scanning, but due to the transmission scanning data statistical shortage, affecting attenuation correction accuracy. In PET/CT, CT image also belong to the transmission scanning, the use of a CT image correction provides a new solution for the attenuation correction of the PET data.The main content of this article is designed based on CT data of PET/CT attenuation correction simulation experiments, the experiments using real human data from the voxellized phantom NCAT (NURBs (Non Uniform rational B-Splines))-Based Cardiac Torso) as object detection, through the GATE (Geant4Simulation Application for Tomographic Emission) open source software package structure PET/CT scan simulation platform to produce original PET projection data, in the process of PET image reconstruction adding PET attenuation correction factor transformed by CT attenuation correction factor, and to complete PET image attenuation correction.After PET/CT imaging, can do PET and CT image segmentation. The segmentation of PET/CT lung images can be used for quantitative analysis and PET imaging data body effect correction. According to the complicated medical image, including the characteristics of the large amount of information, this paper puts forward PET/CT lung images segmentation and extraction method based on wavelet multi-scale edge detection and information entropy iterative threshold theory. The segmentation and extraction PET/CT lung image contour using this method is clear, with edge positioning accurate. The extracted PET/CT lung image not only can be used for quantitative analysis, but also can be directly used for the correction of respiratory motion artifacts in PET imaging data.