Zellweger MJ, Tsirkin A, Vasilchenko V, Failer M, Dressel A, Kleber ME, Ruff P, März W
A memetic pattern-based algorithm (MPA) was assessed for non-invasive coronary artery disease (CAD) diagnosis in 987 LURIC patients, 68% of whom had documented CAD. The optimised MPA achieved AUC values of 0.88 and 0.87 in training and test cohorts, exceeding the Framingham Risk Score (AUC 0.69), with positive predictive values for CAD identification of 93% and 94%, suggesting AI-driven pattern recognition as a viable alternative to conventional diagnostic procedures.
The EPMA Journal, 16. August 2018 | PMID 30174760