Development of a quasi-ideal photon counting x-ray detector with spectral coincidence counters

Johns Hopkins University, Baltimore MD

Investigators

Abstract

Project Summary The proposed project aims at developing the next generation of photon counting detectors (PCDs) with multi- energy inter-pixel coincidence counter (MEICC) and digital pulse energy correction (DPEC) for x-ray computed tomography (CT). We believe that MEICC–DPEC design will bring PCDs much closer to ideal devices and brings many clinical dreams surrounding PCD CT into reality. In a prior R21 EB029739 grant, we developed MEICC concept using Monte Carlo simulations and found that MEICC improved the dose efficiency by a factor of 3.9, which is 87% (=3.9/4.5 with respect to dose efficiency) toward the ideal spectral PCD. Simulations showed that DPEC made the PCD performance almost insensitive to the intensity of x-rays. Building upon this foundation, under this BPI project, we propose to translate our theoretical results into the real world, fabricate physical application-specific integrated circuits (ASICs) with MEICC–DPEC functionality, develop algorithms to process MEICC–DPEC data, and evaluate the performance of PCD using a tabletop x-ray CT system. Our goas are to achieve Cramér–Rao lower bound values (i.e., the minimum noise variance with unbiased estimators) that are >75% lower than the current PCD with 2 energy bins for spectral imaging tasks, which corresponds to dose efficiency improvements by a factor of >4. The specific aims are: (SA1) Optimize MEICC– DPEC design parameters and develop algorithms. (SA2) Develop MEICC–DPEC PCD. (SA3) Assess MEICC– DPEC PCD using a tabletop CT system. We will compare results against (a) the same PCDs but with 2 energy bins without MEICC–DPEC, (b) a clinical PCD CT system, and (c) two clinical dual-energy CT systems. Experimental results in SA3 will be compared against computer simulation results in SA1.