A Monte Carlo study of verification imaging in high dose rate brachytherapy

We have been evaluating the practicality of monitoring the position of an 192Ir source during high dose rate (HDR) brachytherapy treatments using x-ray fluoroscopy. The EGS4 Monte Carlo code has been used to simulate the interactions of 192Ir photons with the patient and the CsI phosphor of an x-ray image intensifier to predict what signals will be generated by these 192Ir photons. The calculations show that it is the 192Ir photons scattered within the patient that are mainly responsible for generating the spurious signals in the x-ray image intensifier that degrade image quality. The scattered 192Ir photons are distributed in the energy range (15-200 keV), which is markedly lower than the average energy of the primaries (360 keV), and therefore interact more efficiently with the CsI phosphor of the x-ray image intensifier. Experimental measurements support these observations, demonstrating that spurious signals produced by the 192Ir source become appreciably larger when the 192Ir source is located within a scattering object rather than air. For a 10 cm airgap, the signal-to-noise ratio (SNR) can decrease by factors ranging between 3 and 10 (no antiscatter grid), depending on the position of a 7 Ci 192Ir source inside a 30 cm thick water phantom. In typical clinical situations, a focused grid (Pb, 12:1, 40 lines/cm) can increase the SNR by about a factor of 2. Furthermore, the SNR rapidly increases with increasing airgap, such that a 20 cm airgap can be as effective as a 12:1 air interspaced grid in eliminating the spurious signals. Our results suggest that use of a high-current x-ray fluoroscopy technique, a large airgap, and a well-designed anti-scatter grid can make the fluoroscopic monitoring of source position in HDR brachytherapy feasible. This, in turn, can improve the quality assurance of such treatments.