Clustering analysis as a new automatic method for distinguishing static and dynamic exposure to high energetic gamma rays using thermoluminescence detectors and a reader with a CCD camera

摘要

AbstractIn this study, we developed a new automatic method for distinguishing static and dynamic exposure to high energetic gamma rays using thermoluminescent LiF:Mg,Cu,P detectors with customized geometry and a reader with a CCD (Charge-Coupled Device) camera. The original TLDtool software was developed further for this study. We conducted measurements in 69 experiments that tested exposure to radiation with photons at an energy of 661?keV in three different conditions (static, static angular, and dynamic) in the dose range appropriate to standard radiation protection practices (1–25?mSv). After registering the CCD camera, images of the dose distribution on the detector's surface were processed and analyzed using the TLDtool desktop application. The reprocessed images were automatically classified to the dynamic or static ionizing radiation exposure groups by the k-means clustering algorithm implemented in the TLDtool. The images were successfully classified at a rate of 100% by the software for the two test configurations of the dosimeter. Thus, we developed an innovative approach that uses a data clustering algorithm for analyzing dosimetric quantities.Highlights?A two-dimensional thermoluminescence dos