Streamlining Quality Control for Radiographic and Fluoroscopic Systems 📝

Author: Caroline Cheney, Patricia G Collins, Allen R. Goode, Preston Le, Andrew M. Polemi, Angela Snyder 👨‍🔬

Affiliation: UVA Health, Atirix Medical Systems 🌍

Abstract:

Purpose: Optimize ongoing quality control (QC) for radiographic (DR) and fluoroscopic (FL) systems by identifying metrics most sensitive to hardware disruptions among regional noise, pooled noise, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) measurements.
Methods: 337 images of custom contrast-detail phantoms were taken by various DR and FL systems under normal operation. Images were used to measure signal, contrast, regional noise, pooled noise, SNR and CNR among 36 phantom objects. Mean signal and regional noise were measured in ImageJ. Further calculations for pooled noise, SNR, and CNR were made in Excel. Regional noise was measured as standard deviation within a phantom object or background region of interest (ROI). Pooled noise combined noise from signal and background regions with various weighting options. Metric stability was assessed by coefficient-of-variation (CV). Correlation was assessed with linear regression. Differences were assessed by paired t-tests.
Results: Across all DR and FL systems analyzed, empirical differences among regional and pooled noises were negligible, and pooled noise measurements had an equal or lower CV. SNR and CNR measurements for each linearly correlated for each DR or FL system.
Conclusion: Collinearity between SNR and CNR suggest these metrics are governed by the same latent variables, hinting at the possible practicality of tracking just one of the metrics for ongoing QC. The linear correlation, in tandem with higher stability of SNR over CNR during normal operation [1], leads to the next natural postulate that SNR and CNR also track with each other respective of hardware disruptions, and that SNR might be a more sensitive indicator of such. The use of pooled noise in the denominator of SNR and CNR calculations retains higher stability.
[1] Goode et al. 2019. Signal and contrast to noise ratio evaluation of fluoroscopic loops for interventional fluoroscope quality control. J. Appl. Clin. Med. Phys. 20(10):172-180.

Back to List