A comparison of reduced-order modelling techniques for application in hyperthermia control and estimation

Area of Science:

• Computational modeling
• Biomedical engineering
• Heat transfer

Background:

• Reduced-order modeling (ROM) is crucial for real-time control and state estimation in complex systems.
• In hyperthermia, ROM simplifies large thermal models for practical applications.
• Comparing ROM techniques is essential for optimizing clinical tools.

Purpose of the Study:

• To compare modal decomposition and balanced realization for hyperthermia thermal models.
• To evaluate ROM methods based on accuracy, robustness, and computational cost.
• To identify optimal ROM strategies for hyperthermia treatment.

Main Methods:

• Simulated hyperthermia heat transfer problems were used for comparison.
• Modal decomposition (MD) reduction was applied.
• Balanced realization (BR) based reduction was applied.

Main Results:

• MD models showed less error than BR models of similar order in low/moderate advection scenarios.
• MD is more robust to sensor/actuator placement changes than BR.
• MD transformation is computationally less demanding than BR.
• MD faced numerical instabilities in high advection cases.

Conclusions:

• Modal decomposition is recommended for hyperthermia models where advection is not dominant.
• Further research is needed to improve BR for real-time clinical applications.
• ROM techniques are vital for advancing hyperthermia control and estimation.