Generalizability of electrocardiographic artificial intelligence
摘要
Electrocardiographic artificial intelligence (ECG-AI) uses AI-based algorithms to analyze ECGs. Recent literature has revealed the untapped potential of ECG beyond traditional diagnostics. ECG-AI is now used not only to detect arrhythmias but also to assess risk and identify both cardiovascular and non-cardiovascular conditions. This perspective article summarizes evidence from published literature to support the conclusion that ECG-AI models are highly generalizable and have the potential to revolutionize healthcare.