Comparative Evaluation of Radiation Dose and Image Quality In Two Cardiac Fluoroscopy Systems with Distinct Image Processing Algorithms
Abstract
Purpose
To quantitatively compare radiation dose output and image quality performance between two cardiac fluoroscopy systems operated under matched exposure conditions, and to assess the impact of proprietary image processing on low-contrast detectability.
Methods
A Leeds TOR 18-FG fluoroscopy phantom was sandwiched between 18cm of polymethylmethacrylate (PMMA) and imaged on two clinical cardiac catheterization systems (Fluoroscopy System 1 and Fluoroscopy System 2). Each system was operated in its respective low- and high-dose clinical fluoroscopy modes. Comparable dose settings were identified by matching air kerma rates measured at the FDA reference point. For each condition, last-image-hold frames were evaluated subjectively by counting visible low-contrast detectability targets and line-pair groups. Objective measurements were performed by calculating contrast and noise from standardized regions of interest for each contrast disc, enabling construction of low-contrast detectability performance.
Results
For matched low-dose modes, air kerma rates differed by 11.2% (System 1 Low Dose: 5.44 mGy/min; System 2 Low Dose: 6.05 mGy/min). For matched high-dose modes, air kerma rates differed by 7.0% (System 1 High Dose: 19.38 mGy/min; System 2 High Dose: 18.05 mGy/min). At these commensurate dose levels, both systems demonstrated comparable visibility of line-pair groups and high-contrast targets. However, low-contrast detectability analysis revealed atypical preservation of low-contrast object visibility on Fluoroscopy System 1 relative to Fluoroscopy System 2, with contrast-detail performance degrading less rapidly at lower contrast levels. This behavior is consistent with enhancement from proprietary image processing.
Conclusion
When operated at matched air kerma rates, the two fluoroscopy systems demonstrated comparable radiation dose output and spatial resolution performance. Differences in low-contrast detectability were primarily attributable to differences in image processing rather than exposure conditions. These results highlight the importance of considering algorithm-driven image appearance, in addition to dose metrics, when comparing fluoroscopy system performance.