Deep Learning-Driven Automated Assessment of Left Ventricular Diastolic Function in Echocardiography

Area of Science:

• Cardiology
• Artificial Intelligence in Medicine
• Medical Imaging Analysis

Background:

• Left ventricular diastolic function (LVDF) assessment is crucial for diagnosing heart conditions.
• Traditional echocardiography analysis can be time-consuming and subject to inter-observer variability.
• Automated analysis tools are emerging to improve efficiency and accuracy.

Purpose of the Study:

• To evaluate the feasibility and accuracy of a fully automated deep learning model for LVDF assessment.
• To compare AI-driven measurements with expert manual interpretations using 2D transthoracic echocardiography (TTE).
• To assess the diagnostic performance of the AI model in classifying diastolic dysfunction grades.

Main Methods:

• A prospective observational study included 302 patients undergoing 2D TTE for suspected diastolic dysfunction.
• AI-based software (Ligence Heart) automatically analyzed diastolic parameters.
• AI measurements were compared against expert manual measurements.

Main Results:

• The AI model demonstrated a high success rate for spectral and tissue Doppler parameters.
• Strong correlations were observed for E velocity (r=0.93), A velocity (r=0.88), E/A ratio (r=0.94), and left atrial volume index (LAVi) (r=0.92).
• High accuracy was achieved in classifying normal and severe diastolic dysfunction grades.

Conclusions:

• Automated AI evaluation of LVDF using 2D TTE is feasible and accurate.
• The AI model provides guideline-based measurements with strong correlation to expert interpretation.
• This AI approach shows potential to enhance diagnostic consistency and efficiency in clinical practice.