PO-06-205 Deep learning based assessment of T:R ratios during prolonged screening in S-ICD patients experienced inappropriate shocks secondary to T-wave over-sensing

Toon, Lin-Thiri, ElRefai, Mohamed, Abouelasaad, Mohamed, Wiles, Benedict, Ward, Samuel, Dunn, Anthony, Zemkoho, Alain, Roberts, Paul and Paisey, John (2025) PO-06-205 Deep learning based assessment of T:R ratios during prolonged screening in S-ICD patients experienced inappropriate shocks secondary to T-wave over-sensing. Heart Rhythm, 22 (4), S723-S724. (doi:10.1016/j.hrthm.2025.03.1745).

Abstract

Background: the conventional S-ICD screening involves a static 10-second ECG screening at various body positions with an eligibility cut-off T:R ratio of 1:3. Despite this screening, S-ICD patients experience higher rates of inappropriate shocks (IAS) compared to transvenous ICD patients, with the primary cause of IAS being T-wave over-sensing (TWO). In this study, we explored cases of patients who experienced IAS due to TWO using a 24-hour Holter recording and a deep-learning AI tool.

Objective: to explore the changes of T:R ratios over 24-hour period in atients who experienced IAS due to TWO using a 24-hour Holter recording and a deep-learning AI tool.

Methods: patients who had IAS secondary to TWO were identified and fitted with a 24-hour Holter monitor to record their S-ICD vectors. The AI tool was then applied to assess the T:R ratio across the entire duration of the recordings.

Results: ten patients (mean age 59 ± 20 years, 60% male) were recruited. Of these, three patients had ischemic cardiomyopathy, three had non-ischemic cardiomyopathy, and the remaining patients had hypertrophic cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy (ARVC), congenital heart disease, or valvular heart disease. Although each of these patients had passed conventional screening before implant, none maintained a T:R ratio threshold of 1:3 across all vectors over 24 hours.

Conclusion: T:R ratios appear to fluctuate based on factors such as physical activity and body position. We believe these patients experienced IAS due to these fluctuations in T:R ratios, despite initially passing the standard screening. We propose the adoption of prolonged screening to optimise selection, implant position and vector selection for S-ICD candidates, with the goal of reducing inappropriate shocks and minimising related psychosocial impacts.

This record has no associated files available for download.

Contributors

Author: Alain Zemkoho

Download statistics