ABSTRACT

Background: Polymer gels are an emerging new class of dosimeters which are being applied to the challenges of modern radiotherapy modalities. Research on gel dosimetry involves several scientific domains, one of which is the imaging techniques with which dose data is extracted from the dosimeters. In the current work, we present our preliminary results of investigating capability of X-ray CT for extracting brachytherapy dose distributions from a normoxic gel dosimeter.

Materials and methods: A normoxic radiosensitive polymer gel was fabricated under normal atmospheric conditions and poured into three phantoms. Using Cs137 brachytherapy sources, the phantoms were irradiated with different dose distributions with a LDR Selectron remote after-loader. To improve SNR, 25 images were obtained of each slice for image averaging and an averaged background image of an un-irradiated gel phantom was then subtracted for artifact removal. To further improve the accuracy, a self-consistent normalized method was used for calibration of the dosimeters based on an assumption of a linear dose response between zero and maximum dose regions in the gel.

Results: Although results reveal very similar CT-number gradients to that of brachytherapy dose distributions, but the method does not fulfill brachytherapy dosimetry requirements. This might be due to the high prescribed doses in this study which in turn results in a large change in the CT numbers. This change in the CT numbers of the images can not be considered to have a linear relationship with dose which was the basic assumption of our calibration method, so the results are just qualitatively comparable.

Conclusion: In this study, the results of using X-ray CT for brachytherapy polymer gel dosimetry is promising but not still satisfying. Improving a proper calibration method for correlating CT numbers to dose will be significantly helpful for performing measurements with CT. The main limitation for CT is still a low signal to noise ratio especially in lower dose areas.