Search

The MICOS Compton spectrometer is a tool for the routine spectrometric control of X‐ray generators, mainly in the medical field (<150 kV). The use of a pen‐type NaI (Tl) detector with thin Be window allows to extend its range down to 6 keV, for a better evaluation of the low‐E side of the spectrum, which is determined by filtration. A specific problem is encountered in the deconvolution of the spectrum at very low energies, dominated by the K escape of iodine. The proposed solution combines an approximate step‐by‐step application of the GRAVEL algorithm, whereas the MAXED unfolding software is used for reaching the final solution. A new computer code was developed that combines spectrum acquisition and deconvolution and calculates the main characteristics of the source spectrum: end point energy, mean energy, energy at 50% of the left slope, and air kerma at 1 m from the source. [ABSTRACT FROM AUTHOR]

We examine the dosimetric control of 90Sr–90Y source trains used for the prevention of restenosis by endovascular irradiation. Beta sources have many advantages in this respect, but their dosimetric control is not easy, because of the very steep dose gradients.Radiochromic films are exposed to the β radiation in a phantom, at the reference distance of 2 mm. Exposure to 60Co gamma-rays and 4.5 MeV electron linac beam are used for dose calibration. No significant difference is found between 60Co and electron calibrations. A bad reproducibility (up to ±8%) is observed in dose measurements with the source train, attributed to fluctuations of the position of the individual sources in the catheter through which the sources are conveyed. This problem is solved by simultaneously exposing two films, on the two sides of a 4-mm thick phantom, with the catheter at the centre. After film digitisation and conversion to dose, the geometrical mean of the corresponding doses at 2 mm in the two images is calculated. A much better reproducibility is obtained (±2%). A software has been written for the analysis and averaging of the images.The results are consistent with Monte-Carlo calculations for a source of the same activity. They disagree with the initial dose calibration of the source train, although traceable to NIST. [Copyright &y& Elsevier]