Theoretical Examination of Charge Transport Effects in Particle-In-Binder Mercuric Iodide (PIB-HgI2) on X-ray Detector Performance

Abstract

Mammographic breast screening is commonly performed using active matrix flat-panel imagers (AMFPIs). However, AMFPIs suffer from low signal-to-noise ratio at the low x-ray exposures encountered in digital breast tomosynthesis (DBT), resulting in decreased detective quantum efficiency (DQE). This loss in DQE can be mitigated through replacement of the amorphous selenium (a-Se) and cesium iodide (CsI:Tl) converters commonly employed in AMFPIs with particle-in-binder mercuric iodide (PIB HgI2) – a material which provides 3 to 10 times higher x-ray induced signal. Unfortunately, the high levels of image lag (over 10%) exhibited by PIB HgI2 precludes its use in clinical systems. Generally, lag refers to imaging signal in a given frame resulting from the trapping and subsequent release of charge carriers generated by incident X-rays in preceding frames. A strategy for significantly reducing lag is to diminish the contribution of hole induced signal (which is believed to be the major contributor of lag) through incorporation of a Frisch grid structure into the PIB HgI2 converter.

This dissertation focuses on the use of physics-based modeling to identify Frisch grid designs that minimize hole induced signal and lag, while maintaining high total induced signal. The physics in the modeling includes the drift of electrons and holes through the converter, trapping and release of holes in the converter material, and charge accumulation on the insulating pillars that support and encapsulate the Frisch grid as well as the effect of that accumulated charge on the electric field. Signal properties, including total signal, hole suppression, image lag, line spread function and modulation transfer function, have been examined as a function of the voltage applied to the electrodes, grid pitch (the center-to-center distance between neighboring grid wires) and RGRID (the ratio of grid wire widths to grid pitch), x-ray spectrum and frame time. In addition, DQE performance was calculated using cascaded system analysis.

Modeling results show that, through judicious selection of grid design, hole induced signal can be suppressed by more than 90%, leading to a reduction in lag by over an order of magnitude down to levels less than 1% – without compromising total induced signal or spatial resolution. For the most favorable grid designs, DQE performance was found to be largely maintained as exposure decreases.