Exploring trade-offs between target coverage, healthy tissue sparing, and the placement of catheters in HDR brachytherapy for prostate cancer using a novel multi-objective model-based mixed-integer evolutionary algorithm

Abstract

Brachytherapy is a form of radiotherapy whereby a radiation source is guided near tumors, using devices such as catheter implants. In the present clinical work ow, catheters are rst placed inside or close to the tumor based on clinical expertise. Subsequently, soware is used to design a plan for the delivery of radiation. Treatment planning is essentially a multi-objective optimization problem, where con icting objectives represent radiation delivered to tumor cells and healthy cells. However, current clinical so ware collapses this information into a single-objective, constrained optimization problem. Moreover, catheter positioning is typically not included. As a consequence, it is hard to obtain insight into the true nature of the trade-o s between key planning objectives and the placement of catheters. Such insights are however crucial in understanding how be er treatment plans may be constructed. To obtain such insights, we interface with real-world clinical so ware and derive potential catheter positions for real-world patients. Selecting and con guring catheters requires mixed-integer optimization. For this reason, we extend the recently-proposed Genetic Algorithm for Model-Based mixed-Integer opTimization (GAMBIT) to tackle multi-objective optimization problems. Our results indicate that clinically acceptable plans of high quality may be achievable with less catheters than typically used in current clinical practice.

Better and faster catheter position optimization in HDR brachytherapy for prostate cancer using multi-objective real-valued GOMEA

“…This requires 5, 4, and 5 catheters for patients 1, 2, and 3, respectively. For comparison, the Pareto fronts presented in [11] required 7, 7, and 14 catheters, respectively. The treatment plans that were clinically used for these patients are all outside the Golden Corner.…”

“…The results are compared to the results obtained with GAMBIT in [11]. These results were reported to have been obtained with a duration of optimization limited to 48 hours per catheter number.…”

“…Clinical data from three treated patients is used for optimization. In order to allow a comparison with GAMBIT, the patient data used in [11] was requested and used as the patient cases. We follow the approach used for GAMBIT regarding the number of dosecalculation points.…”

“…Multiple works already exist on this, e.g., [2,6], using single-objective optimizers. Recently, a fully multi-objective approach, the Multi-Objective Genetic Algorithm for Model-Based mixed-Integer opTimization (MO-GAMBIT), has been applied to this optimization problem [11]. Three objectives were considered, namely number of catheters, coverage of targets, and sparing of OARs, and promising results were found.…”

Uncertainty Handling in Surrogate Assisted Optimisation of Games

“…Real-world optimisation problems often exhibit a so called observation uncertainty, because physical measurements usually contain errors [4]. This uncertainty can cause issues, especially if the application is sensitive, such as in case of medical treatments [111]. Research in the field of noisy optimisation and real-world applications targets this and related issues.…”