Can artificial intelligence lower the global sudden cardiac death rate? A narrative review

Area of Science:

• Cardiology
• Artificial Intelligence
• Medical Informatics

Background:

• Sudden cardiac arrest (SCA) is a significant global cause of mortality.
• Current survival rates for SCA remain low, underscoring the need for improved prediction and prevention strategies.
• The World Health Organization (WHO) defines sudden cardiac death (SCD) based on timeframes relative to symptom onset or observation of the individual.

Purpose of the Study:

• To review the current applications of Artificial Intelligence (AI) in predicting and preventing sudden cardiac arrest (SCA).
• To explore the potential of AI, machine learning (ML), and deep learning (DL) in identifying individuals at risk for SCA.

Main Methods:

• A literature search was conducted using PubMed and Web of Science databases.
• Studies were included if they focused on sudden cardiac death prediction and prevention, and utilized artificial intelligence, machine learning, or deep learning techniques.

Main Results:

• AI, ML, and DL demonstrate significant potential for risk stratification in SCA.
• These technologies can aid in identifying individuals at higher risk of experiencing SCA.
• Current AI models require further development and validation for widespread clinical use.

Conclusions:

• AI, ML, and DL offer promising avenues for improving SCA survival rates through enhanced risk stratification.
• Further research is essential, focusing on explainable AI, larger and diverse datasets, external validation, multimodal approaches, and ethical considerations.
• Addressing bias and ensuring robust testing are critical steps to fully realize the potential of AI in SCA prevention.