Validation of AI Algorithm Interpretation for Pre Participation Ecg in Athletes

Q: What are the key findings of this study?

AI-ECG significantly correlated with male sex (r=0.724, p<0.0001) and age (r=0.199, p<0.0001) in athletes, but not with low EF (p=0.703), HCM (p=0.086), or AF (p=0.113).

Q: What is the clinical evidence from this study?

Study design: Cohort. Population: Pre-participation cardiovascular screening in athletes (n=845). Intervention: Artificial intelligence-ECG (AI-ECG) prediction model vs. Cardiologist-interpreted ECG (international criteria) and follow-up echocardiogram. Primary outcome: Accuracy of AI-ECG prediction for determining age, sex, probability of low ejection fraction, hypertrophic cardiomyopathy, and atrial fibrillation (r = 0.724 (male sex); r = 0.199 (age); r = -0.039 (low EF); r = 0.173 (HCM); r = 0.053 (AF), p=<0.0001).

Olubadewa A. Fatunde; Juliana Savic; Wallace Klein Schwengber; Winston T. Wang; Claire Yee; Zachi Itzhak Attia; Michael John Ackerman; Konstantinos C. Siontis; Luis R. Scott

doi:10.1016/s0735-1097(25)02985-7

March 29, 2025Journal of the American College of CardiologyOpen Access

Validation of AI Algorithm Interpretation for Pre Participation Ecg in Athletes

View Full Paper

Ask AI

Bookmark

Share

Key result

AI-ECG predicts age and sex in athletes but fails to identify structural heart disease or AF.

r = 0.724 (male sex); r = 0.199 (age); r = -0.039 (low EF); r = 0.173 (HCM); r = 0.053 (AF)
P<0.0001
n=845

Study Design

Type

Cohort (n=845)

Multicenter

No

PICO

P

Population

845 collegiate or professional athletes (24.3% female) undergoing pre-participation cardiovascular screening with 12-lead ECG at Mayo Clinic Arizona between January 2018 and September 2024.

E

Exposure / Comparator

Artificial intelligence-ECG (AI-ECG) prediction model vs Cardiologist-interpreted ECG (international criteria) and follow-up echocardiogram

O

Primary Outcome

Accuracy of AI-ECG prediction for determining age, sex, probability of low ejection fraction, hypertrophic cardiomyopathy, and atrial fibrillation — r = 0.724 (male sex); r = 0.199 (age); r = -0.039 (low EF); r = 0.173 (HCM); r = 0.053 (AF), p=<0.0001

Main Result

Effect estimate: r = 0.724 (male sex); r = 0.199 (age); r = -0.039 (low EF); r = 0.173 (HCM); r = 0.053 (AF)

p-value: p=<0.0001

Limitations

Underpowered to test AI model performance in predicting hypertrophic cardiomyopathy and atrial fibrillation due to low event rates
Lower model performance in women (84.5% vs 98.6% in men) and compared to the general population
Follow-up echocardiogram was only performed in 99 individuals (11.2%)
Lack of matched analysis to non-athletes

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

AI-ECG sex/age signals in athletes should not change screening; leaves open validation for low EF, HCM, or AF detection.

Abstract

No abstract yet. The journal hasn't published one we can show.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Fatunde et al. (2025) conducted a cohort in Pre-participation cardiovascular screening in athletes (n=845). Artificial intelligence-ECG (AI-ECG) prediction model vs. Cardiologist-interpreted ECG (international criteria) and follow-up echocardiogram was evaluated on Accuracy of AI-ECG prediction for determining age, sex, probability of low ejection fraction, hypertrophic cardiomyopathy, and atrial fibrillation (r = 0.724 (male sex); r = 0.199 (age); r = -0.039 (low EF); r = 0.173 (HCM); r = 0.053 (AF), p=<0.0001). AI-ECG significantly correlated with male sex (r=0.724, p<0.0001) and age (r=0.199, p<0.0001) in athletes, but not with low EF (p=0.703), HCM (p=0.086), or AF (p=0.113).

synapsesocial.com/papers/6acf69dd3316c606f341fa12 https://doi.org/10.1016/s0735-1097(25)02985-7

View Full Paper

Ask AI

Bookmark

Share

Validation of AI Algorithm Interpretation for Pre Participation Ecg in Athletes

Study Design

PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Validation of AI Algorithm Interpretation for Pre Participation Ecg in Athletes

Study Design

PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading