WE‐C‐L100J‐06: Linear Systems Analysis for a New Solid State X‐Ray Image Intensifier (SSXII) Based On Electron‐Multiplying Charge‐Coupled Devices (EMCCDs)

Purpose: To objectively evaluate and optimize performance of a new SSXII using linear‐systems analysis with varying constituent elements. Method and Materials: An imaging module of the SSXII consists of a fluorescent phosphor, a minifying fiber‐optic taper (FOT), and a fiber‐optic plate (FOP) coupled directly to the sensor of an EMCCD camera. An array of such modules is used to achieve the desired field‐of‐view (FOV). Linear‐systems analysis was applied to a module to determine system performance for various combinations of components whose properties were estimated using known physical constants and manufacturer specifications. A 350μm thick CsI(Tl) structured phosphor deposited on an additional FOP was considered in this analysis; however the phosphor may be grown directly on the FOT for improved optical transfer efficiency. New back‐thinned sensors, offering high optical quantum efficiencies (>90%), are also considered. Various FOT minifications were studied to evaluate the tradeoffs between a larger field‐of‐view (FOV) per module and potential DQE degradation. Results: Initial calculations indicate elimination of a FOP in the imaging chain could improve the integral DQE by 30–80%, depending on the FOT minification. Use of a back‐thinned sensor could offer additional improvements of 10–25% over a front‐illuminated EMCCD. Increasing the FOT minification factor from 3:1 to 6:1 would tend to decrease the integral DQE by ∼50%, which could be more than compensated for by making the above two design improvements. Calculated MTFs and DQEs will be presented for various minifications, exposures, and gains. A prototype SSXII (16μm pixels) demonstrated 20 lp/mm bar‐pattern resolution radiographically and fluoroscopic image sequences with exposures of <3 μR/frame. Conclusion: The linear‐systems analysis indicates current designs for the SSXII will provide high‐resolution, low‐noise, real‐time imaging. With component optimization, an acceptable DQE can be maintained with taper minifications as large as 6:1. (Partial support: the UB Foundation and NIH grants R01‐EB002873, R01‐NS43924.).