Digital angiography: Ghost hunters

Abstract

Low dose Angiography (LDA) systems rely on significant recursive filtration (aka temporal averaging) to maintain image quality. The use of this type of processing is specific to the imaging application (cardiac, neuro, vascular…) and can be varied by the system vendor. In situations of rapid motion, such processing can have a detrimental effect on images obtained, producing temporal ghosting. LDA enables the use of lower kV and mAs resulting in lower staff and patient doses. It also allows manufacturers to overcome output limitations when using smaller focal spots. While the benefits are clear very little work has been carried out in relation to the drawbacks of this approach. This work aims to characterise recursive filtration in LDA systems. Common test objects such as the Leeds Test Objects do not allow for the dynamic assessment of Image Quality. In this study, a modified NEMA Cardiology Phantom and a homemade linear motion phantom were used to investigate the impact of Motion Recursive Filtering on image quality in LDA. Systems assessed were: Philips Alura FD 20, Alura Clarity FD 10 and Siemens Artis Zee systems. All systems using LDA demonstrated different levels of Temporal Ghosting. The Intensity and spacing of the ghost images was dependent on system type, protocol and pulse rate. The impact of temporal ghosting impinges on the effective optimisation of clinical systems. The impact of temporal ghosts should be assessed during the commissioning and optimisation stages for all digital angiography systems.