Head-to-head comparison of adaptive statistical and model-based iterative reconstruction algorithms for submillisievert coronary CT angiography

• Published in: European heart journal cardiovascular imaging 2018-02, Vol.19 (2), p.193-198
• Main Authors: Benz, Dominik C, Fuchs, Tobias A, Gräni, Christoph, Studer Bruengger, Annina A, Clerc, Olivier F, Mikulicic, Fran, Messerli, Michael, Stehli, Julia, Possner, Mathias, Pazhenkottil, Aju P, Gaemperli, Oliver, Kaufmann, Philipp A, Buechel, Ronny R
• Format: Article
• Language: English
• Subjects: Analysis of Variance; Cohort Studies; Computed Tomography Angiography - methods; Coronary Angiography - methods; Coronary Artery Disease - diagnostic imaging; Coronary Artery Disease - physiopathology; Databases, Factual; Female; Humans; Image Processing, Computer-Assisted; Male; Prospective Studies; Sensitivity and Specificity; Signal-To-Noise Ratio
• ISSN: 2047-2404; 2047-2412
• DOI: 10.1093/ehjci/jex008

Abstract Aims Iterative reconstruction (IR) algorithms allow for a significant reduction in radiation dose of coronary computed tomography angiography (CCTA). We performed a head-to-head comparison of adaptive statistical IR (ASiR) and model-based IR (MBIR) algorithms to assess their impact on quantitative image parameters and diagnostic accuracy for submillisievert CCTA. Methods and results CCTA datasets of 91 patients were reconstructed using filtered back projection (FBP), increasing contributions of ASiR (20, 40, 60, 80, and 100%), and MBIR. Signal and noise were measured in the aortic root to calculate signal-to-noise ratio (SNR). In a subgroup of 36 patients, diagnostic accuracy of ASiR 40%, ASiR 100%, and MBIR for diagnosis of coronary artery disease (CAD) was compared with invasive coronary angiography. Median radiation dose was 0.21 mSv for CCTA. While increasing levels of ASiR gradually reduced image noise compared with FBP (up to − 48%, P