Portal MV imaging with thin-film high-energy current X-ray detectors: A Monte Carlo study.

PURPOSE: We investigate the potential of the recently introduced high-energy current (HEC) thin-film detector as an alternative design to existing electronic portal imaging devices (EPID). The HEC radiation detectors employ multiple nano-/micrometer layers made of disparate atomic number (Z) conductors separated by solid or gaseous (e.g., air) dielectrics. The HEC detector may be designed as an external structure or an auxiliary device integrated into the existing EPID.

METHOD: Using Geant4 Monte Carlo simulations, we compare the image contrast of a HEC detector to that of a commercial EPID device (AS500) for a 2.5 MV X-ray beam. The detector response was modeled using a series of monoenergetic incident photons with detector signal scored as the energy deposition in the air gaps (HEC) and in the phosphor layer (EPID). Las Vegas (LV) phantom was employed to test the spatial resolution and contrast of the single- and multielement HEC detector. The HEC detector pixel size was the same as for AS500 (0.78 mm × 0.78 mm). In addition, image contrast of a water/bone phantom using both the multielement HEC and EPID detectors was simulated and compared.

RESULTS: The HEC detector has higher relative response to low-energy photons compared to EPID. The multielement HEC has 32.3 times greater response at 100 keV than at 500 keV, while the EPID without copper plate shows a factor of 6.8 between the same energies. LV phantom images indicate that the image contrast is approximately the same for single- and multielement HEC detectors, but the latter has lower noise. Both single- and multielement HEC could resolve a 2 mm diameter hole with an image magnification factor of 1.2. In the present design, the HEC detector has much less material (9.66 mg/cm2 gold) compared to EPID (133 mg/cm2 Gd2 O2 S) to interact with incident photons. For imaging bony structures, the HEC detector needs about nine times greater photon flux as the EPID to acquire data at same uncertainty level. Despite this, the HEC sensor requires less than 1 cGy dose to obtain images with statistical uncertainty better than 2.5%. In the case when the field diameter is 10 cm, the multielement HEC image contrast is 14.3% higher than that of EPID with copper removed. When the field diameter is decreased to 5 cm, HEC contrast is 27.1% better than the EPID contrast.

CONCLUSION: It is demonstrated that the image contrast of HEC detectors are comparable and in some cases better than that of the standard EPID design. This opens a potential for complementary HEC and EPID designs that may better utilize the kV portion of the spectrum.