Effect of charge trapping and release on the performance of polycrystalline mercuric iodide x-ray converters incorporating Frisch grid structures for digital breast tomosynthesis

Shen, Liuxing; El‐Mohri, Youcef; Zhao, Qingwei; Antonuk, Larry E.

doi:10.1117/12.2610762

Cited by 1 publication

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such behavior is the result of even more effective suppression of the effect of 4). Finally, it is worth noting that for a given grid pitch and EFR, grid designs with R GRID values higher than optimal R GRID exhibit MTFs that deviate from the sinc function due to signal loss affecting the shape of the corresponding LSFs (Shen et al 2022). DQE results for a PIB HgI 2 detector with a Frisch grid determined using cascaded systems analysis are shown in figure 11.…”

Section: Resultsmentioning

confidence: 99%

“…Early investigations focused on improving the PIB HgI 2 material have reported lag levels of 4% to 5% (Cho et al 2008, Kim et al 2011 but have not led to practical devices. Our group has been investigating the effect of introducing a Frisch grid into the converter to suppress the contribution of holes to the signal through theoretical modeling-with the finding that hole signal and lag can be largely suppressed without significantly affecting either the total imaging signal or MTF (Shen et al 2020, Shen et al 2021a, 2021b, Shen et al 2022. Frisch grids have long been employed in gas detectors (Frisch 1944) as well as in solid-state, gamma-ray spectroscopic detectors in order to achieve single polarity charge sensing (He 2001).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Minimization of image lag in polycrystalline mercuric iodide converters through incorporation of Frisch grid structures for digital breast tomosynthesis

Shen¹,

Antonuk²,

El‐Mohri³

et al. 2023

Phys. Med. Biol.

View full text Add to dashboard Cite

Objective. Polycrystalline mercuric iodide photoconductive converters fabricated using particle-in-binder techniques (PIB HgI2) provide significantly more detected charge per x-ray interaction than from a-Se and CsI:Tl converters commonly used with active matrix flat-panel imagers (AMFPIs). This enhanced sensitivity makes PIB HgI2 an interesting candidate for applications involving low x-ray exposures—since the relatively high levels of additive electronic noise exhibited by AMFPIs incorporating a-Se and CsI:Tl reduce detective quantum efficiency (DQE) performance under such conditions. A theoretical study is reported on an approach for addressing a major challenge impeding practical use of PIB HgI2 converters—the high lag exhibited by the material (over 10%) which would lead to undesirable image artifacts in applications involving acquisition of consecutive images such as digital breast tomosynthesis. Approach. Charge transport modeling accounting for the trapping and release of holes (thought to be the primary contributor to lag) was used to examine signal properties, including lag, of pillar-supported Frisch grids embedded in the photoconductor for 100 μm pitch AMFPI pixels. Performance was examined as a function of electrode voltage, grid pitch (center-to-center distance between neighboring grid wires) and the ratio of grid wire width to grid pitch. Main results. Optimum grid designs maximizing suppression of signal generated by hole transport, without significantly affecting the total signal due to electron and hole transport, were identified and MTF was determined. For the most favorable designs, additional modeling was used to determine DQE. The results indicate that, through judicious choice of grid design and operational conditions, first frame lag can be significantly reduced to below 1%—less than the low levels exhibited by a-Se. DQE performance is shown to be largely maintained as exposure decreases—which should help to maintain good image quality. Significance. Substantial reduction of lag in PIB HgI2 converters via incorporation of Frisch grids has been demonstrated through modeling.

show abstract

Section: Resultsmentioning

confidence: 99%