A Weak Signal Extracting Method for Defect Diagnosis in Spherical Radiation Source based on Active Imaging

Gamma-ray source is widely used in medical imaging. The inner structure of radiation source can be diagnosed using active imaging technique. The small defect appears to be an extremely weak signal in the diagnostic image of spherical
⁶⁰
Co-$\gamma$-ray source. This article reveals a weak signal extracting method called IFM (Iteration Filling Method) to distinguish the small defect based on mean filter and median filter. A program is developed based on Geant 4 toolkit to simulate the imaging process. A $\Phi 4 \mathrm{~cm}$ spherical 60
^Co
embedded by defect of $\Phi 1 \mathrm{~cm}, \Phi 8 \mathrm{~mm}, \Phi 6 \mathrm{~mm}$ and $\Phi 4 \mathrm{~mm}$ is chosen as the test object. The $\gamma$-ray intensity of the defect is 98 %, 95 %, 90 %, 85 %, 80 % and 0 % of the normal sections. After image processing, the defect shape can be identified clearly for intensity lower than 85 % and can be distinguished for intensity lower than 95%. What’s more, images of 1 %, 3 % and 10 % statistical fluctuation are tested. The shape and size of defect can be easily judged from image of 1 % statistical fluctuation. The defect edge turns to be irregular and distortion appears for image of 3% statistical fluctuation. Obvious distortion and artifact appear for image of 10 % statistical fluctuation. It indicates that the weak signal can be effectively extracted from the diagnostic image of strong background using IFM. Increasing image contrast and reducing statistical fluctuation is helpful to increase the extracting efficiency.