Answers are grounded in this paper and cite their sources.
Hyperthermia has been attracting great attention, research resources and clinical translation efforts as a cancer treatment. Metallic nanoparticles can enhance heat deposition in tumors when subjected to external energy sources like lasers. However, challenges remain in accurately estimating state variables, such as the temperature and heat sources, during treatments. This study presents a combined Physics-Informed Neural Network (PINN) and particle filter approach for state estimation in a model, representing a sample of a nanofluid heated by a near-infrared diode laser. The PINN is trained to solve the heat transfer model and serve as the state evolution model in the particle filter. Synthetic and actual temperature measurements from heating experiments involving a nanofluid of palladium-cerium oxide nanoparticles are used in the solution of the state estimation problem. Verification tests show that the particle filter can robustly estimate states with 850 particles in few seconds of computational time, due to the efficient PINN predictions. Overall, the combined PINN - particle filter approach demonstrates potential for solving state estimation problems in complex engineering systems, such as in cancer thermotherapy.
No takes yet. Share an insight, caveat, or question.
Pedruzzi et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: