In animal experiments involving local electromagnetic field (EMF) exposure, the accuracy of biological temperature rise strongly depends on the cooling effects due to blood flow, which are commonly modeled by the blood perfusion rate (B) in the Pennes bioheat transfer equation. However, applying literature-reported values of B under high-intensity exposure conditions leads to discrepancies with measurement data, while exhaustive searches for accurate parameters is computationally prohibitive. Therefore, this study proposes a novel method to efficiently perform inverse estimation of thermal parameters that satisfy both physical and physiological plausibility using a Weibull model and response surface methodology.