Analysis of noisy 2D angiographic images for improved blood flow rate quantification in dialysis access

Measurement of access blood flow rate pre- and post-angioplasty is a marker of interventional success and confirmation of dialysis access function restoration during endovascular procedures. In this paper, we discuss the effects of noise on indicator-dilution curves at various acquisition and flow conditions and investigate the methods to minimize such errors through the use of noise reducing techniques. It was found that for peak to peak algorithm using gamma variate curve fit, the overall mean accuracy of all simulated conditions was 23% above measured flow, while for cross-correlation algorithm using mean filter method the overall mean accuracy was 18% above measured flow. The mean quantification accuracy based on all measured flow conditions was considered best at 3 F/s-which correlate to conditions for minimum radiation risk. The results of this study will be useful to understand the impact of noise on accuracy and for selecting optimum methods and acquisition parameters to improve computational accuracy while aiming for minimum radiation dose.