Metaheuristic Optimization Method for Neutron Capture Therapy
C Busch1*, S Bogetic1, C Ramsey1,2, (1) University of Tennessee, Knoxville, TN, (2) Thompson Cancer Survival Center, Knoxville, TN
Presentations
MO-I430-BReP-F3-2 (Monday, 7/11/2022) 4:30 PM - 5:30 PM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 3
Purpose: There has been a increase in interest in Boron Neutron Capture Therapy (BNCT) thanks to advancements in neutron source modeling combined with new delivery compounds. A novel advanced optimization tool is introduced to efficiently optimize the neutron spectra. The optimization software developed uses a general purpose meta-heuristic optimization algorithm coupled to a radiation transport solver to generate a Beam Shaping Assembly (BSA) design given a set of patient related constraints and dose objectives.
Methods: A framework is developed to couple the Gnowee metaheuristic optimization algorithm to a Monte Carlo Neutron transport code [MCNP 6.2]. The goal of Gnowee/MCNP is to optimize the materials, geometry and dimensions of the BSA. The cylindrical dimensions that could be changed include the cylindrical exit radius, internal and external reflector boundary, the overall external BSA radius, and the total length. The filter materials in the BSA could also be changed.
Results: Starting from an initial population of 200 random BSA design solutions, the Gnowee/MCNP framework optimization adjusted 38 design parameters over 175 generations to evaluate 3500 BSA designs. The most significant improvement found by Gnowee/MCNP is a dramatically shorter optimal design search period, and smaller dimension (70x70cm vs. 140x140cm) for higher neutron intensity in the target. The smaller dimension and higher intensity allow the patient to be placed much closer to the source and decrease the beam-on time for treatment delivery.
Conclusion: The pairing of Gnowee/MCNP allows for precise tailoring of the neutron beam for the energy and shape. The treatment also projects a sparing of tissue while maximizing dose to a desired location and minimizing the dose in healthy tissue. The advances allow for a general optimization that can be used for a wide range of patient and tumor applications to generate and evaluate a wide range of designs of new treatment systems.
Keywords
Neutron Capture Therapy, Optimization
Taxonomy
TH- External Beam- Particle/high LET therapy: Neutron therapy- BNCT
Contact Email