Can Artificial Intelligence Identify the Ideal Ablation Area for the Patients With Atypical Atrial Flutter? A Proof...

Area of Science:

• Cardiology
• Medical Technology
• Artificial Intelligence in Medicine

Background:

• Catheter ablation for atypical atrial flutter (AFL) presents challenges due to complex circuits.
• Success rates in AFL ablation vary based on circuit complexity and operator experience.

Purpose of the Study:

• To develop artificial intelligence (AI) software for identifying the critical isthmus of atypical AFL using 3D mapping data.
• To create a tool that assists in pinpointing optimal ablation targets for complex arrhythmias.

Main Methods:

• Extracted 3D mapping data from 19 atypical AFL episodes.
• Designed AI software analyzing local data (location, timing, voltage) to detect critical isthmus characteristics (wavefront convergence, conduction slowing).
• Validated software by comparing estimated critical isthmus location with actual AFL termination areas.

Main Results:

• The AI software identified the critical isthmus in all cases within 24.9 seconds.
• Achieved a high concordance rate of 79% between the software's Top 1 prediction and the actual AFL termination area.
• Potential arrhythmia termination was observed in two cases where ablation was performed at the estimated critical isthmus.

Conclusions:

• The developed AI software accurately identifies the optimal treatment area for atypical AFL.
• This tool can lead to faster and more precise catheter ablation procedures.
• The software may help less experienced operators achieve outcomes comparable to those of experienced clinicians.