Author: Salman M. Arnab, Yves Chevalier, Samuel Gagné, Adrian F. Howansky, Luc Laperrière, Xiangyi Wu, Wei Zhao 👨🔬
Affiliation: Stony Brook Medicine, Analogic Canada 🌍
Purpose: A direct-indirect dual-layer flat-panel-detector (DI-DLFPD) is under development for patient motion artifact-free contrast-enhanced digital mammography (CEDM). DI-DLFPD comprises a direct front layer (FL) detector for low-energy x-ray image acquisition and an indirect back layer (BL) detector for simultaneous high-energy acquisition. This study presents measurements from DI-DLFPD prototypes and compares them with theoretical calculations. The goal is to investigate the impact of BL detector thickness on imaging performance and optimize the DI-DLFPD design for CEDM.
Methods: Two DI-DLFPD prototypes were developed, both utilizing a 200 µm thick a-Se for the FL. The BL employed CsI with thicknesses of 200 µm and 400 µm, respectively. Measurements were conducted using a 49kV tungsten x-ray spectrum, shaped with a K-edge filter. Modulation transfer function (MTF) and noise power spectra (NPS) were evaluated using both experimental and theoretical methods. The detectability of contrast-enhanced lesions was calculated for both prototypes.
Results: As BL CsI thickness increased from 200 µm to 400 µm, the MTF decreased by ~ 21% at 1 cycles/mm (f1) due to increased optical blurring. Its effect on signal-to-noise ratio was outweighed by the reduction in NNPS across all frequencies: ~ 48% at zero frequency attributed to improved x-ray absorption, and ~56% at f1 due to the decrease in MTF. Both FL and BL are x-ray quantum noise limited across the clinical exposure range. The detectability of 8 mm contrast-enhanced lesions increased by 15% in the prototype with 400 µm CsI compared to that with 200 µm CsI.
Conclusion: Increasing BL CsI thickness to 400 µm improved lesion detectability for CEDM, while both detectors met the requirement of x-ray quantum noise limited performance. Excellent agreement was achieved between experimental and theoretical results, which validated our model for investigating the impact of DI-DLFPD design parameters on contrast-enhanced lesion detectability and other imaging tasks.