David J Winkel1,2, V Reddappagari Suryanarayana3, A Mohamed Ali3, Johannes Görich4, Sebastian Johannes Buß4, Axel Mendoza2, Chris Schwemmer5, Puneet Sharma2, U Joseph Schoepf6, Saikiran Rapaka2. 1. Department of Radiology, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland. 2. Siemens Healthineers, 755 College Rd E, 08540 Princeton, NJ, USA. 3. Siemens Healthcare Private Limited, Unit No. 9A, 9th Floor, North Tower, Mumbai 400079, India. 4. Das Radiologische Zentrum - Radiology Center, Sinsheim-Eberbach-Walldorf-Heidelberg, Germany. 5. Siemens Healthineers, Siemensstrasse 1, 91301 Forchheim, Germany. 6. Division of Cardiovascular Imaging, Department of Radiology and Radiological Science, Medical University of South Carolina, 25 Courtenay Drive, 29425 Charleston, SC, USA.
Abstract
AIMS: To present and validate a fully automated, deep learning (DL)-based branch-wise coronary artery calcium (CAC) scoring algorithm on a multi-centre dataset. METHODS AND RESULTS: We retrospectively included 1171 patients referred for a CAC computed tomography examination. Total CAC scores for each case were manually evaluated by a human reader. Next, each dataset was fully automatically evaluated by the DL-based software solution with output of the total CAC score and sub-scores per coronary artery (CA) branch [right coronary artery (RCA), left main (LM), left anterior descending (LAD), and circumflex (CX)]. Three readers independently manually scored the CAC for all CA branches for 300 cases from a single centre and formed the consensus using a majority vote rule, serving as the reference standard. Established CAC cut-offs for the total Agatston score were used for risk group assignments. The performance of the algorithm was evaluated using metrics for risk class assignment based on total Agatston score, and unweighted Cohen's Kappa for branch label assignment. The DL-based software solution yielded a class accuracy of 93% (1085/1171) with a sensitivity, specificity, and accuracy of detecting non-zero coronary calcium being 97%, 93%, and 95%. The overall accuracy of the algorithm for branch label classification was 94% (LM: 89%, LAD: 91%, CX: 93%, RCA: 100%) with a Cohen's kappa of k = 0.91. CONCLUSION: Our results demonstrate that fully automated total and vessel-specific CAC scoring is feasible using a DL-based algorithm. There was a high agreement with the manually assessed total CAC from a multi-centre dataset and the vessel-specific scoring demonstrated consistent and reproducible results. Published on behalf of the European Society of Cardiology. All rights reserved.
AIMS: To present and validate a fully automated, deep learning (DL)-based branch-wise coronary artery calcium (CAC) scoring algorithm on a multi-centre dataset. METHODS AND RESULTS: We retrospectively included 1171 patients referred for a CAC computed tomography examination. Total CAC scores for each case were manually evaluated by a human reader. Next, each dataset was fully automatically evaluated by the DL-based software solution with output of the total CAC score and sub-scores per coronary artery (CA) branch [right coronary artery (RCA), left main (LM), left anterior descending (LAD), and circumflex (CX)]. Three readers independently manually scored the CAC for all CA branches for 300 cases from a single centre and formed the consensus using a majority vote rule, serving as the reference standard. Established CAC cut-offs for the total Agatston score were used for risk group assignments. The performance of the algorithm was evaluated using metrics for risk class assignment based on total Agatston score, and unweighted Cohen's Kappa for branch label assignment. The DL-based software solution yielded a class accuracy of 93% (1085/1171) with a sensitivity, specificity, and accuracy of detecting non-zero coronary calcium being 97%, 93%, and 95%. The overall accuracy of the algorithm for branch label classification was 94% (LM: 89%, LAD: 91%, CX: 93%, RCA: 100%) with a Cohen's kappa of k = 0.91. CONCLUSION: Our results demonstrate that fully automated total and vessel-specific CAC scoring is feasible using a DL-based algorithm. There was a high agreement with the manually assessed total CAC from a multi-centre dataset and the vessel-specific scoring demonstrated consistent and reproducible results. Published on behalf of the European Society of Cardiology. All rights reserved.
Authors: Thomas Sartoretti; Antonio G Gennari; Elisabeth Sartoretti; Stephan Skawran; Alexander Maurer; Ronny R Buechel; Michael Messerli Journal: J Nucl Cardiol Date: 2022-03-17 Impact factor: 5.952
Authors: Pedro Covas; Eison De Guzman; Ian Barrows; Andrew J Bradley; Brian G Choi; Joseph M Krepp; Jannet F Lewis; Richard Katz; Cynthia M Tracy; Robert K Zeman; James P Earls; Andrew D Choi Journal: Front Cardiovasc Med Date: 2022-03-21
Authors: Claudia Morf; Thomas Sartoretti; Antonio G Gennari; Alexander Maurer; Stephan Skawran; Andreas A Giannopoulos; Elisabeth Sartoretti; Moritz Schwyzer; Alessandra Curioni-Fontecedro; Catherine Gebhard; Ronny R Buechel; Philipp A Kaufmann; Martin W Huellner; Michael Messerli Journal: Diagnostics (Basel) Date: 2022-08-03
Authors: Giuseppe Muscogiuri; Valentina Volpato; Riccardo Cau; Mattia Chiesa; Luca Saba; Marco Guglielmo; Alberto Senatieri; Gregorio Chierchia; Gianluca Pontone; Serena Dell'Aversana; U Joseph Schoepf; Mason G Andrews; Paolo Basile; Andrea Igoren Guaricci; Paolo Marra; Denisa Muraru; Luigi P Badano; Sandro Sironi Journal: Heliyon Date: 2022-10-05