Click here to

Session: Imaging General ePoster Viewing [Return to Session]

Automatic Detector QC and Image Quality Assessment in Digital Radiography

R DiTusa*, J So, Y Liu, B Peng, H Hsu, P Chaudhary, T Lin, S Jambawalikar, Columbia University Medical Center, New York, NY


PO-GePV-I-98 (Sunday, 7/10/2022)   [Eastern Time (GMT-4)]

ePoster Forums

Purpose: The aim of this study is to automate quality control (QC) measurements of digital radiography (DR) plates to decrease inter- and intra-operator variability.

Methods: Eight phantom images acquired from a radiography unit (Luminos Agile Max, Siemens) were analyzed using an automatic software. The automatic assessment includes signal-to-noise ratio (SNR), noise (standard deviation), and modulation transfer function (MTF). Signal ROIs were automatically sized and placed on the image with the ability to correct for the position of the ROIs over each contrast object. Once the ROIs have been placed, a noise ROI is appended next to the signal ROI. From these ROIs, SNR is calculated. ROIs were placed on cyclic bars ranging from 1.6 to 3.7 line-pairs per mm to calculate 20%, 35%, and 50% MTF. The line pairs visualized in the QC surveys were compared to the calculated MTF.

Results: SNR decreased below five for the first two QC reports analyzed—according to the Rose Criterion, an SNR greater than five is required for an object to be visualized. Before the third QC report was completed, a re-calibration of the detector was performed, resulting in SNR measurements of all seven contrast objects above five. The calculated 20%, 35%, and 50% MTF were about constant, with the calculated 35% MTF comparable to the physicist’s visual assessment during the QC surveys.

Conclusion: Automated QC software could be used to minimize user variability in QC measurements, and consequently allows for an increased confidence in QC survey data.


Digital Imaging, Image Analysis, Quality Control


IM- X-Ray: Image processing

Contact Email