Purpose: To evaluate the clinical performance of a dose reduction system utilizing AI guidance and a shuttered attenuator system to reduce radiation doses, to both patients and operators, during fluoroscopically guided interventional procedures.
Methods: Two interventional fluoroscopy systems were retrofit with a shuttered attenuator system that intermittently attenuates the x-ray field outside of the primary focus area of the interventionalist. An integrated AI system adjusts the primary focus area based on interpretation of the tools and actions of the interventionalist to maintain high image quality in a dynamic, primary region of interest (ROI). Evaluation of the clinical performance of the system included characterization of entrance exposure, exposure from scattered radiation, high contrast resolution, low contrast resolution, and tracking performance of the AI system.
Results: Dosimetry measurements demonstrate that the dose reductions of both DAP and exposure measurements scale in proportion to the relative areas of the attenuated and primary focus areas. However, the dose reduction algorithm is not immediately effective upon initiating an exposure. Fluoroscopists who utilize multiple short exposures are unlikely to realize the dose savings afforded by the system. High contrast spatial resolution in the peripheral region is reduced slightly compared to that in the primary ROI, but maintain spatial resolution of 2.36 lp per mm. CNR and contrast detail object detectability are more significantly affected by the increased noise present in the area outside of the primary ROI. The temporal response of the AI system is on the order of 1 s and can result in image discontinuities between the regions inside and outside of the primary ROI due to different refresh rates.
Conclusion: The AI system generally performs as expected although subtle features of the system performance may require adjustments to operator technique to fully realize the dose reduction benefits of the system.