Artificial intelligence in ventricular arrhythmias and sudden cardiac death: A guide for clinicians

Figures

• 

  Fig. 1. AI fusion for VT ablation planning. Workflow illustrating the integration of imaging and AI for guiding ablation strategies in VT. The process begins with cardiac imaging to delineate structural abnormalities, followed by scar segmentation to identify core and border zones. Extracted features are processed using AI-based analysis, incorporating radiomics and computational modelling. These outputs are integrated into VT ablation planning, including creation of three-dimensional meshes and digital twin simulations, with subsequent generation of activation maps and identification of potential reentry circuits. The final stage supports clinical decision-making, enabling targeted ablation of arrhythmogenic substrate. AI: artificial intelligence; VT: ventricular tachycardia; 3D: three-dimensional. LGE: late gadolinium enhancement. MRI: magnetic resonance imaging.

• 

  Fig. 2. Core domains of AI in VT and SCD. This schematic highlights four principal applications of AI relevant to risk stratification and management. Top left: ECG-based risk prediction, where AI algorithms applied to the 12-lead electrocardiogram can identify high-risk patients beyond traditional criteria. Top right: AI combined with cardiac imaging facilitates the quantification of scars and mapping of fibrosis, thereby enhancing substrate assessment. Bottom left: integration of ICD and wearable device data allows AI models to forecast arrhythmic events and refine event prediction. Bottom right: liability and ethical considerations, addressing medicolegal concerns and the trustworthiness of AI-generated alerts in clinical practice. AI artificial intelligence. ICD: intracardiac defibrillator.

• 

  Fig. 3. Cardiac magnetic resonance imaging with late gadolinium enhancement in hypertrophic cardiomyopathy. Arrows indicate areas of patchy mid-wall and subepicardial fibrosis in hypertrophied myocardial segments. Such non-ischaemic enhancement patterns represent an arrhythmogenic substrate and are associated with increased risk of ventricular arrhythmias and sudden cardiac death.

• 

  Fig. 4. Twelve-lead electrocardiogram illustrating classical features of hypertrophic cardiomyopathy. The tracing demonstrates increased QRS voltages, deep narrow (“dagger-like”) Q waves in the lateral leads (I, aVL, V5-V6), and widespread repolarisation abnormalities, including biphasic T waves.