Artificial Intelligence Benchmarking of Wearable-Triggered Arrhythmia Detection Under Unmeasured Atrial Fibrillation Episode Boundary Ambiguity: A Healthcare False-Alarm Robustness Evaluation

Wearable-triggered arrhythmia detectors are benchmarked against reference annotations whose atrial fibrillation episode boundaries are themselves uncertain, yet the effect of that boundary ambiguity on reported false-alarm performance has not been characterised. The framework is then embedded in a before-after-control-impact design with a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64) and evaluated across 2000 replications per design cell for cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points.54 clusters are required to detect a 5 percentage point change at 80% power, and that the minimum detectable effect at 80% power falls steeply as clusters increase from 8 to 32. The contribution is a boundary-aware benchmarking protocol that separates detector error from annotation error, together with the design parameters needed to evaluate it credibly.