First evaluation of a novel method for a single run 3D Angiography

• Published in: Clinical neuroradiology (Munich) Vol. 28; no. S1; p. S62
• Main Authors: Lang, Stefan, Holter, Philip, Schmidt, Manuel, Lucking, Hannes, Kaethner, Christian, Kowarschik, Markus, Dorfler, Arnd
• Format: Journal Article
• Language: English
• Published: Springer 01.09.2018
• Subjects: Analysis; Aneurysms; Angiography; Methods
• ISSN: 1869-1439; 1869-1447
• DOI: 10.1007/S00062-018-0719-8

Purpose: 3D DSA consists of mask and fill runs for generating subtracted 3D images. To save patient dose (PD), a novel AI-based method for single run 3D Angiography (3DA) has been developed that generates DSA-like 3D images based on fill runs only. Our aim was an initial evaluation of 3DA. Methods: 3D DSA datasets of cerebral aneurysms were reconstructed using conventional and prototype software (Siemens Healthineers AG, Erlangen, Germany). Corresponding reconstructions have been analyzed by 2 neuroradiologists in consensus reading in terms of quantitative (aneurysm size AS in mm, parent vessel's diameter VD in mm) and qualitative parameters (image quality IQ, localization, aneurysmal configuration, neck width). Results: In total 10 datasets ([n.sub.male] = [6;n.sub.female] = [4;age.sub.mean] = 62.8 years) have been successfully reconstructed using both conventional and prototype software. In all cases corresponding reconstructions demonstrated complete congruency in terms of IQ ([IQ.sub.3D DSA/3DA] = 4), localization ([n.sub.ACA] = [2;n.sub.ACI] = [4;n.sub.BA] = [2;n.sub.MCA] = 2), aneurysmal configuration ([saccular.sub.3D DSA/3DA] = 10) and neck width ([narrow.sub.3D DSA/3DA] = 2; [medium.sub.3D] DSA/3DA = 2; [wide.sub.3D DSA/3DA] = 6). Analysis of the aneurysm size ([AS.sub.3D DSA] = 5.2 [+ or -] 3.2 mm;[AS.sub.3DA] = 5.2 [+ or -] 3.1 mm) and the parent vessel's diameter ([VD.sub.3D DSA] = 3.6 [+ or -] 0.8 mm;[VD.sub.3DA] = 3.7 [+ or -] 0.8 mm) did not show significant differences (p < 0.05) between both reconstructions. Conclusion: 3DA is a promising method and shows comparable results to 3D DSA in terms of quantitative and qualitative parameters and might help to reduce PD.