Room: AAPM ePoster Library
Purpose: assess image quality and potential dose reduction in using Statistical Pixel Angiography (SPA) over Digital Subtraction Angiography (DSA) via grid elimination. The SPA algorithm expands upon the principles of kinetic imaging, which calculates pixel variance across a series of radiographic images.
Methods: phantom experiment was conducted with a Philips fluoroscope to simulate contrast flowing through an artery. Varying thicknesses of acrylic were used to simulate three patient diameter sizes: small (18 cm), average (24 cm), and large (35 cm). The acrylic was stacked over a 1.8 mm diameter catheter, which served as a surrogate for an abdominal artery. Omnipaque contrast was injected into the catheter, and fluorographic images were acquired with and without the grid. The DSA images were summed to generate single sumDSA images for reference comparisons. SPA images were then generated from the un-subtracted images. Contrast-to-Noise Ratios (CNRs) were calculated by analyzing regions of interest across the catheter in both sets of images, and ratios of SPA-to-DSA CNR (RCs) were reported.
Results: SPA algorithm resulted in significantly improved CNRs over the DSA images. The average RC(SPA/DSA) for the images acquired with and without the grid was 2.03 and 2.42, respectively. The SPA images without the grid were able to achieve up to 85% of the CNR of the DSA images with the grid for a small patient. This resulted in a dose reduction of 75% without any optimization of the automatic exposure control rate (AERC) system.
Conclusion: images can generate over twice the CNR of DSA images. The SPA algorithm in conjunction with grid elimination can produce comparable image quality, with a significant reduction in radiation output. The improvements in SPA image quality over DSA can be leveraged to eliminate the use of grids in interventional radiology to reduce patient doses.