Electrocardiography-based artificial intelligence predicts the upcoming future of heart failure with mildly reduced ejection fraction

Dae-Young Kim ¹ Sang-Won Lee ² Dong-Ho Lee ³ Sang-Chul Lee ³ Ji-Hun Jang ¹ Sung-Hee Shin ¹ Dae Hyeok Kim ¹ Wonik Choi ² Yong-Soo Baek ^1*

• ¹ Division of Cardiology, Department of Internal Medicine, Inha University College of Medicine and Inha University Hospital, Incheon, Republic of Korea
• ² Inha University, Incheon, Republic of Korea
• ³ DeepCardio Co., Ltd., Incheon, Republic of Korea

Background: Heart failure with mildly reduced ejection fraction (HFmrEF) has emerged as the predominant subtype of heart failure (HF). This study aimed to develop artificial intelligence (AI)-electrocardiography (ECG) to identify and predict the prognosis of patients with HFmrEF. Methods: We collected 104,336 12-lead ECG datasets from April 2009 to December 2021 in a tertiary centre. The AI-ECG encompasses a novel model that combines an automatic labelling preprocessing method with a transformer architecture incorporating a triplet loss for HFmrEF analysis. Results: The receiver operating characteristic analyses revealed that the area under the curve of AI-ECG for identifying all types of HF was acceptable (0.873, 95% confidence interval [CI]: 0.864–0.893), while that for identifying patients with HFmrEF was relatively lower (0.824, 95% CI: 0.794–0.863) than that for those with HF with reduced ejection fraction (EF) (0.875, 95% CI: 0.844–0.912) and those with normal EF (0.870, 95% CI: 0.842–0.894). The analysis of ECG features showed significant increases in QRS duration (p=0.001), QT interval (p=0.045), and corrected QT interval (p=0.041) with increasing ‘Severity by Euclidean distance’. Following the predictability analysis with another group of 953 patients for improvements of follow-up EF in HFmrEF, the patients were grouped into three clusters based on the AI-Euclidean distance; Cluster 1 had the most severe cases and poorer outcomes than Clusters 2 (p<0.001) and 3 (p<0.001). Conclusions: AI-ECG presents an innovative approach for the prognostic stratification of cardiac contractility in patients with HFmrEF. In patients with HFmrEF, disease progression can be predicted using AI-ECG.