Automated Classification of Severity in Cardiac Dyssynchrony Merging Clinical Data and Mechanical Descriptors

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Cardiac resynchronization therapy (CRT) benefits patients with left ventricle malfunction and conduction disorders.
• A significant portion (20%-30%) of CRT patients do not respond to treatment.
• Mechanical dyssynchrony is a potential indicator for CRT response.

Purpose of the Study:

• To develop and validate an automated classification model for assessing the severity of ventricular contraction dyssynchrony.
• To identify key clinical and imaging factors predictive of CRT response.

Main Methods:

• Utilized clinical data including left ventricular ejection fraction (LVEF), QRS, and P-R intervals.
• Applied factor analysis of dynamic structures to equilibrium radionuclide angiography images to extract mechanical contraction behavior.
• Developed automated classifiers to distinguish between absent, mild, and moderate-severe dyssynchrony.

Main Results:

• Achieved 90%, 50%, and 80% hit rates for interventricular dyssynchrony (absent, mild, moderate-severe).
• Observed 100%, 50%, and 90% hit rates for intraventricular dyssynchrony (absent, mild, moderate-severe).
• Demonstrated high accuracy in classifying dyssynchrony severity in heart failure patients.

Conclusions:

• The automated model shows promise for clinical application in patient follow-up for CRT.
• Accurate dyssynchrony assessment can potentially improve patient selection and outcomes for CRT.
• This method offers a valuable tool for managing patients with ventricular dysfunction and conduction disorders.