Assumption-bounded reduced electro-thermal modeling of lesion formation in radiofrequency cardiac ablation: power-time allocation, thermal latency, perfusion, and temperature limitation.

Radiofrequency ablation comparisons are confounded by energy, latency, perfusion, assumptions, and controller action. An assumption-bounded two-dimensional electro-thermal model compared 30 W/30 s, 50 W/10 s, fixed 90 W/4 s, and temperature-limited 90 W/4 s protocols. A 30 s post-pulse endpoint passed 12/12 cases. Equal-energy peak temperatures differed. In the 4 mm-buffer grid, the limiter reduced power in 29/60 cells. Peak-temperature and protocol-associated perfusion-effect orderings persisted across return-location tests, whereas the presence of the Arrhenius damage contour Ω ≥ 1 and controller intervention were return-sensitive. This supports screening, not absolute or patient-level prediction.