Investigating the effects of different kernels used for CT image reconstruction on dose distributions in treatment planning of kidney cancer radiotherapy

Introduction: The quality of CT images used for treatment planning of cancer patients is an important issue in accurate outlining of the tumor volume and organs at risk. Different kernels in CT scanner systems are available for improving the image quality. Applying these kernels  on   CT  images  will  change   the   CT  numbers  and  electron   density   of   tissues, consequently. Therefore, the aim of this study was to assess effects of different CT kernels on Hounsfield unit variations, the related computed electron densities and the calculated dose distributions in radiation treatment planning system. Materials and Methods: The 16 slice Siemens CT scanner used in this work. The B30, B35, B41 and B50 kernels applied on abdomen CT images for a kidney cancer patient. The ISOgray treatment planning system was used for radiation treatment planning and calculating dose of 6MV photon beam energy. The dose volume histogram (DVH) of left kidney as target volume and spinal cord as organ at risk were calculated. Results: The B50 kernel, had the greatest effect on calculated CT numbers for considered reference points relative to standard kernel B30, among the applied kernels for image reconstruction (16-19 HU variations). The average of calculated percent dose in target volume for 3 reference points obtained %99.19, %100.36T %99.89 and %100.87 for standard kernel B30 and other B35, B41 and B50 kernels, respectively. Mean dose in DVH was 2.03 for B30 and 2.05 for other kernels. Conclusion: The Hounsfield units of the selected reference points, in the target volume of reconstructed CT images by various kernels had relatively high variations respect to B30 kernel. Despite these variations, electron density and consequently the average of calculated percent dose in the target volume did not show considerable changes. Hence, it can be concluded that the use of image reconstruction kernels to improve the quality of CT images will help to determine the edges and outlines of tumors and organs at risk more precisely. While, applying these kernels does not significantly affect the amount of calculated doses in the treatment planning system