The use of mathematical models to study complex systems such as physical and biological phenomena allows understanding their behavior, specifically regarding variables and parameters that are difficult to obtain. Additionally, studying optimization techniques has made it possible to approximate the characteristics of these systems by correlating numerical simulations and experimentation. Radiofrequency hyperthermia therapy for cancer treatment is currently under consideration for future medical applications. However, some of its properties are complex to measure, which could prevent their control. This is the case of electrical conductivity, which depends on the induction frequency and the tissue characteristics. In this paper, radiofrequency hyperthermia therapy was simulated via the finite element method. Then, an estimation of the electrical conductivity involved in the treatment was performed using the particle swarm optimization method. The execution time and the difference between the estimated parameter and the exact value were evaluated and compared with those obtained using the Levenberg-Marquardt method. The results indicate a significant agreement between the estimated and exact values in three different cases. The Levenberg-Marquardt method has a difference of 0,1942% and a performance time of 22 minutes, whereas the particle swarm optimization method has a difference of 0,0967% and a performance time of 327 minutes. The latter performs better in terms of parameter value estimation, whereas the former has better computational times. These techniques may help medical doctors to prescribe treatment protocols and may open the possibility of devising control strategies for hyperthermia therapy as a cancer treatment.