Model Predictive Control Applied to a Reduced-order Model of a Magnetic Nanoparticle Hyperthermia Treatment
Open Access
- Author:
- Aguilar, Dario
- Graduate Program:
- Electrical Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- June 15, 2022
- Committee Members:
- Rafic Bachnak, Professor in Charge/Director of Graduate Studies
Kiana Karami, Thesis Advisor/Co-Advisor
Ma'Moun Abdel Abu-Ayyad, Committee Member
Anilchandra Attaluri, Thesis Advisor/Co-Advisor
Mohammad-Reza Tofighi, Committee Member - Keywords:
- Magnetic Nanoparticles
Hyperthermia
Eddy currents
Model Predictive Control - Abstract:
- Magnetic Nanoparticles Hyperthermia (MNPH) is a heating technique used in medicine to assist chemotherapy and radiotherapy of cancerous tumor treatments. The nanoparticles are manufactured with Iron (Fe), a magnetic material; when an Alternating Magnetic Field (AMF) is applied, the nanoparticles experience Specific Power Loss (SPL), meaning that part of the applied power is transformed into heat. As a result, the technique is used for localized heating. On the other hand, when an AMF is applied to a conductive material, the tissue in this case, closed-loop currents known as Eddy currents, are induced, and some power is also lost as non-specific heating. Temperatures greater than 43 (◦C) (hyperthermia) applied for a certain period of time has been demonstrated to be effective against malignant tumors. However, one challenge with this technique is controlling the temperature in the tumor area while maintaining a low temperature of less than 39 (◦C) in the surrounding healthy tissue, which receives heat mainly due to Eddy currents. A mathematical reduced-order model of the MNPH treatment is proposed to design a Model Predictive Control (MPC) capable of maintaining a temperature in the tumor at about 43 (◦C) while preserving temperatures in the healthy tissue at less than 39 (◦C). Finally, a Proportional Integral (PI) control is also designed for comparison purposes. The MPC allows the tumor to reach a steady temperature of 43 (◦C) 85 (s) earlier than the PI control does when Eddy current heating is not considered in the tumor temperature control. However, when Eddy current heating effect is considered, the settling time increases significantly, 405 (s) for the MPC and 520 (s) for the PI control; additionally, the tumor temperature response becomes oscillatory for both controllers but within the hyperthermia region. Finally, the tumor temperature starts oscillating in the hyperthermia region at 100 (s) when the MPC is used, this is 15 (s) sooner than it does when the PI control is used. The MPC uses less aggressive changes in the AMF amplitude, 10.4 (kA/m/s) less than the PI control, to drive the tumor to hyperthermia.