Click here to


Are you sure ?

Yes, do it No, cancel

An Experimental Method to Measure Zero-Frequency DQE in the Presence of System Drift

X Ji*, M Feng , R Zhang , G Chen , K Li , University of Wisconsin-Madison, Madison, WI


(Sunday, 7/14/2019) 5:00 PM - 6:00 PM

Room: 303

Purpose: For the measurement of local or detector panel-specific DQE, it is more desirable to perform a large number of independent image acquisitions instead of shifting a region-of-interest in a single or a few images. However, data acquired through repeated scans are usually contaminated by system drift; one percent of system drift could lead to more than an order of magnitude underestimation of zero-frequency DQE. The purpose of this work was to develop an experimental method that corrects drift-induced error and enables direct and accurate measurement of DQE(k) at k=0.

Methods: First, we performed a cascaded systems analysis to pinpoint the quantitative relationship between system drift and errors in the zero-frequency NPS and DQE. Under the guidance of the theoretical analysis, a method was developed to isolate the drift-induced error term based on the measured autocovariance data themselves without requiring any knowledge about the drift function. Next, simulations with known ground truth for DQE(0) were performed to validate the proposed method. Finally, experimental studies were performed using a photon counting detector-based x-ray imaging system to further validate the proposed method.

Results: Drift leads to a positive error in the measured NPS(0). Magnitude of this error is proportional to the area of the image data used for NPS measurement. For the simulation data, percent error of NPS(0) was reduced from 3400% without correction to only 2.4% with correction. Percent error of DQE(0) was reduced from 97% to 2.5%. For the experimental data, the proposed method generated a DQE(k) curve that is not only continuous across k=0 but also consistent with our previous theoretical DQE(0) model.

Conclusion: An experimental method was developed to correct system drift-induced error in the measured DQE and to facilitate the characterization of panel-specific and detector system intrinsic DQE(0).

Funding Support, Disclosures, and Conflict of Interest: The work is supported by an NIH grant R01EB020521 and a DOD Breakthrough Award W81XWH-16-1-0031.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email