Multiobjective Optimization of Microwave Phased Array Excitation for Targeted Tissue Heating With Reduced Channel Power in Hyperthermia Treatment Planning

The multiobjective genetic algorithm (MOGA) is proposed for determining phased array (PA) excitation to selectively heat the tumor with minimal healthy tissue hotspot. Power absorption indicators, such as mean specific absorption rate (SAR) in tumor target ( ${{SAR}}_{{targ}})$ , hotspot to tumor quotient (HTQ), average power absorption ratio (aPA ,ratio), and 50% iso-SAR tumor coverage ( ${{TC}}_{{50}})$ , were combined in pairs as an objective function in MOGA. Coupled electromagnetic (EM) and thermal simulations of the heterogeneous breast with locally advanced breast cancer (LABC) were used for evaluating MOGA and single-objective GA (SOGA)-based treatment plans; 25 LABC patient models with varying breast (144.66-1339.3 cc) and tumor (10.53-88.88 cc) volumes and tumor location were employed. The treatment plans were assessed using aforementioned SAR indicators, steady-state thermal metrics in tumors, healthy tissue hotspots in the hyperthermia range (40 degrees C-44 degrees C), and channel power consumption of the PA. Among the treatment plans, MOGA with $[1/SAR_{{targ}}, HTQ]_{obj}$ yielded 104.63% rise in ${{SAR}}_{{targ}}$ across patients with thermal metrics comparable to SOGA with ${{HTQ}}_{{obj}}$ [state of the art (SoA)]. MOGA with $[1/SAR_{{targ}}, HTQ]_{{obj}}$ required only 50.5% channel power compared to the SoA, as the tumor was efficiently heated without wasting power in the healthy tissues. This is a significant contribution to developing an affordable PA system.