Predictive Value of Artificial Intelligence-Enabled Electrocardiography in Patients With Takotsubo Cardiomyopathy

Area of Science:

• Cardiology
• Artificial Intelligence
• Medical Diagnostics

Background:

• Takotsubo cardiomyopathy (TC) patients face high risks of major adverse cardiovascular events.
• A validated tool for risk stratification in TC is currently lacking.
• This study investigates the prognostic utility of AI-ECG algorithms in TC.

Purpose of the Study:

• To evaluate the predictive value of AI-ECG algorithms for adverse outcomes in TC patients.
• To determine if AI-ECG findings can improve risk stratification beyond conventional factors.
• To explore the potential of AI in identifying subtle ECG patterns associated with TC prognosis.

Main Methods:

• Analysis of consecutive patients from the Mayo Clinic Takotsubo syndrome registry.
• Application of validated AI-ECG algorithms to estimate ECG-age, low ejection fraction probability, and atrial fibrillation probability.
• Construction of multivariable models, including Cox proportional hazards analysis, to assess AI-ECG association with major adverse cardiac events (MACE).

Main Results:

• 305 TC patients were analyzed with a median follow-up of 4.8 years.
• High-risk AI-ECG findings were associated with increased MACE.
• The presence of 2 or 3 high-risk AI-ECG findings significantly predicted MACE (HR, 4.419; P=0.001) after adjusting for conventional risk factors.

Conclusions:

• AI-ECG algorithms can detect subtle ECG patterns linked to poorer outcomes in TC.
• This AI-ECG approach shows promise for stratifying high-risk TC patients.
• Integrating AI-ECG into clinical practice may enhance TC patient management and outcomes.