Artificial intelligence for the assessment of diastolic function

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Left ventricular diastolic function assessment is a complex echocardiographic task.
• Artificial intelligence (AI) offers automation in data acquisition, analysis, and interpretation.
• This review focuses on AI advancements in diastolic function evaluation.

Purpose of the Study:

• To summarize recent progress in applying AI to echocardiographic assessment of diastolic function.
• To highlight AI's potential to overcome limitations of current guideline-based methods.

Main Methods:

• Review of recent studies on AI applications in diastolic function assessment.
• Analysis of trends including deep learning, reduced input variables, and clinical outcome validation.

Main Results:

• AI models demonstrate potential to match or surpass expert performance in diastolic function evaluation.
• AI improves reproducibility and workflow efficiency.
• Recent trends show a shift towards deep learning and continuous probabilistic scoring.

Conclusions:

• Guideline-based diastolic function assessment has limitations like variability and complexity.
• AI, including machine learning and deep learning, can integrate diverse data for advanced pattern identification.
• Future research will address AI explainability, generalizability, and clinical integration.