In vivo validation of a novel algorithm for automatic premature ventricular contractions recognition

Area of Science:

• Cardiology
• Medical Imaging
• Algorithm Development

Background:

• Template-matching algorithms are standard in premature ventricular contractions (PVCs) catheter ablation.
• A lack of systematic analysis exists for the accuracy and spatial resolution of these systems.
• This study validates a novel automated template-matching algorithm in vivo.

Purpose of the Study:

• To systematically validate the performance of a novel automated template-matching algorithm.
• To assess the accuracy and spatial resolution of the algorithm for PVC identification.
• To determine the clinical utility of the algorithm in catheter ablation.

Main Methods:

• Utilized a porcine model with simulated PVCs from various cardiac origins.
• Evaluated the algorithm's ability to discriminate between sinus rhythm and PVCs.
• Assessed spatial resolution by measuring QRS morphology discrimination at distances of 0-6 mm from a simulated PVC focus.

Main Results:

• Achieved 99.6% specificity and 85.3% sensitivity for recognizing simulated PVCs.
• Demonstrated significant discrimination differences between PVC origin and distances of 2, 4, and 6 mm (P < 0.001).
• Reported a pace-mapping resolution of 2 mm with a c-statistic of 0.93 for discrimination.

Conclusions:

• The automated template-matching algorithm exhibits high specificity and sensitivity.
• The algorithm provides good spatial discrimination with a resolution of approximately 2 mm.
• Clinical implementation of this algorithm could enhance catheter ablation for PVCs.