BibTex format
@article{Palermi:2026:10.1016/j.ijcard.2026.134677,
author = {Palermi, S and Zeidaabadi, B and Vecchiato, M and Anselmino, M and Adorisio, R and Biffi, A and Borrelli, F and Brugin, E and Cantarutti, N and Cavarretta, E and Cominacini, M and Corsi, M and DAscenzi, F and De, Feo V and Di, Gioia G and Dorelli, G and Foccardi, G and Gallina, S and Giangrandi, S and Graziano, F and Aggour, H and El-Medany, A and Pastika, L and Barker, J and Patlatzoglou, K and Khattak, GR and Lodi, E and Livio, A and Maestrini, V and Manfredi, GL and Mansour, D and Modena, M and Neunhaeuserer, D and Nigro, A and Palermi, A and Pellegrino, A and Pelliccia, A and Quattrini, FM and Ricci, F and Scarzella, F and Squeo, MR and Tonelli, R and Zanardo, E and Zorzi, A and Peters, NS and Kramer, DB and Waks, JW and De, Ferrari GM and Ng, FS and Sau, A and Saglietto, A},
doi = {10.1016/j.ijcard.2026.134677},
journal = {Int J Cardiol},
title = {Development and external validation of AI-ECG models in athlete pre-participation screening: Performance, limitations, and clinical implications.},
url = {http://dx.doi.org/10.1016/j.ijcard.2026.134677},
year = {2026}
}
RIS format (EndNote, RefMan)
TY - JOUR
AB - BACKGROUND: Pre-participation screening (PPS) in competitive athletes aims to identify cardiovascular diseases associated with sudden cardiac death (SCD). Although the 12lead electrocardiogram (ECG) represents the cornerstone of PPS, structural abnormalities may demonstrate limited or incomplete electrical expression, particularly in asymptomatic athletes with physiological remodeling. Artificial intelligence (AI)-enabled ECG models have shown promising performance in hospital-based populations, but their transportability to low-prevalence athlete screening environments remains uncertain. OBJECTIVES: To develop and externally validate a deep learning (DL)-based AI-ECG ensemble model for detecting imaging-confirmed structural heart disease in competitive athletes undergoing PPS. METHODS: A convolutional neural network (CNN) ensemble was trained using hospital-derived ECG images from Beth Israel Deaconess Medical Center (BIDMC, Boston, USA) and externally validated in the Italian Team for Athlete CARDiac evaluation and AI-based Risk prediction (ITACARD-AI) registry. Separate CNNs were developed for valvular heart disease (VHD) and cardiomyopathies (CM) and combined using XGBoost meta-learning. Model performance was assessed using area under the receiver operating characteristic curve (AUROC), subgroup analyses, and threshold-based evaluation. RESULTS: The ITACARD-AI cohort included 1115 competitive athletes (mean age 26±13years; 70% male), including 48 athletes (4.3%) with VHD and 30 (2.7%) with CM. External validation demonstrated substantial performance degradation compared with hospital-based internal validation. AUROC values decreased to 0.70 (95% CI 0.64-0.75) for VHD and 0.69 (95% CI 0.60-0.78) for CM, indicating only modest discrimination in the screening population. Threshold analyses showed high negative predictive values (~99%) but persistently low positive predictive values (≤8%), reflecting limited disease enrichment and strong prevalence dep
AU - Palermi,S
AU - Zeidaabadi,B
AU - Vecchiato,M
AU - Anselmino,M
AU - Adorisio,R
AU - Biffi,A
AU - Borrelli,F
AU - Brugin,E
AU - Cantarutti,N
AU - Cavarretta,E
AU - Cominacini,M
AU - Corsi,M
AU - DAscenzi,F
AU - De,Feo V
AU - Di,Gioia G
AU - Dorelli,G
AU - Foccardi,G
AU - Gallina,S
AU - Giangrandi,S
AU - Graziano,F
AU - Aggour,H
AU - El-Medany,A
AU - Pastika,L
AU - Barker,J
AU - Patlatzoglou,K
AU - Khattak,GR
AU - Lodi,E
AU - Livio,A
AU - Maestrini,V
AU - Manfredi,GL
AU - Mansour,D
AU - Modena,M
AU - Neunhaeuserer,D
AU - Nigro,A
AU - Palermi,A
AU - Pellegrino,A
AU - Pelliccia,A
AU - Quattrini,FM
AU - Ricci,F
AU - Scarzella,F
AU - Squeo,MR
AU - Tonelli,R
AU - Zanardo,E
AU - Zorzi,A
AU - Peters,NS
AU - Kramer,DB
AU - Waks,JW
AU - De,Ferrari GM
AU - Ng,FS
AU - Sau,A
AU - Saglietto,A
DO - 10.1016/j.ijcard.2026.134677
PY - 2026///
TI - Development and external validation of AI-ECG models in athlete pre-participation screening: Performance, limitations, and clinical implications.
T2 - Int J Cardiol
UR - http://dx.doi.org/10.1016/j.ijcard.2026.134677
UR - https://www.ncbi.nlm.nih.gov/pubmed/42468692
ER -