What is the clinical evidence from this study?

Study design: Observational. Population: Paroxysmal atrial fibrillation (n=318321). Intervention: Deep neural network (AI) model for 12-lead ECG. Primary outcome: Prediction of atrial fibrillation onset within one month of normal sinus rhythm (AUROC 0.905).

What does this research mean for the field?

A deep learning model using standard 12-lead ECGs can predict paroxysmal atrial fibrillation onset within one month of normal sinus rhythm with high accuracy (AUROC 0.905), relying significantly on features near the T wave and ST segment. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

February 21, 2025Open Access

Explainable paroxysmal atrial fibrillation diagnosis using an artificial intelligence-enabled electrocardiogram

YJYeongbong JinSeoul National University BKBonggyun KoChonnam National University WCWoo-Jin ChangSeoul National University

Key Result

A deep learning model using standard 12-lead ECGs predicted paroxysmal atrial fibrillation onset within one month of normal sinus rhythm with an AUROC of 0.905.

Study Design

Type

Observational (n=318,321)

Multicenter

Structured PICO

Does a deep learning algorithm applied to normal sinus rhythm 12-lead ECGs predict the onset of paroxysmal atrial fibrillation within one month?

Population

318,321 patients with 552,372 normal sinus rhythm (NSR) 12-lead ECGs recorded between 2013 and 2020.

Intervention

Deep neural network (1d-CNN) trained on standard 12-lead ECGs to predict paroxysmal atrial fibrillation (PAF) onset within one month of NSR.

Outcome

Prediction of atrial fibrillation onset within one month of normal sinus rhythm (early diagnosis of PAF), measured by area under the receiver operating characteristic curve (AUROC).

An AI-enabled 12-lead ECG model can accurately predict the onset of paroxysmal atrial fibrillation within one month from normal sinus rhythm, with the ST segment and T wave identified as key predictive features.

Main Result

Effect estimate: AUROC 0.905

Limitations

Single-center design, which requires validation across diverse healthcare systems
Data imbalance may influence specificity and AUROC
Absence of personalized information extraction hinders individualized analysis
ECG summarization based on true positive and negative cases may reduce detail on specific ECG intervals

Abstract

BACKGROUND/AIMS: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). METHODS: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. RESULTS: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. CONCLUSION: Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Ask AI

Helpful

Bookmark

View Full Paper