Abstract Deterministic direct-search methods have been successfully used to address real-world challenging optimization problems, including the beam angle optimization (BAO) problem in radiation therapy treatment planning. BAO is a highly non-convex optimization problem typically treated as the optimization of an expensive multi-modal black-box function which results in a computationally time consuming procedure. For the recently available modalities of radiation therapy with protons (instead of photons) further efficiency in terms of computational time is required despite the success of the different strategies developed to accelerate BAO approaches. Introducing randomization into otherwise deterministic direct-search approaches has been shown to lead to excellent computational performance, particularly when considering a reduced number (as low as two) of random poll directions at each iteration. In this study several randomized direct-search strategies are tested considering different sets of polling directions. Results obtained using a prostate and a head-and-neck cancer cases confirmed the high-quality results obtained by deterministic direct-search methods. Randomized strategies using a reduced number of polling directions showed difficulties for the higher dimensional search space (head-and-neck) and, despite the excellent mean results for the prostate cancer case, outliers were observed, a result that is often ignored in the literature. While, for general global optimization problems, mean results (or obtaining the global optimum once) might be enough for assessing the performance of the randomized method, in real-world problems one should not disregard the worst-case scenario and beware of the possibility of poor results since, many times, it is only possible to run the optimization problem once. This is even more important in healthcare applications where the mean patient does not exist and the best treatment possible must be assured for every patient.

Wastewater Treatment Plants (WWTPs) are complex structures, with nonlinear characteristics, that have strict quality criteria, and carry out the treatment using various resources, with special emphasis on electricity. The optimization of WWTPs is a necessity to achieve sustainability. This work proposes the use of metaheuristic algorithms to optimize the functioning of a WWTP. Genetic Algorithm, Particle Swarm Optimization, and Simulated Annealing are used to minimize the aeration energy consumption prediction by an artificial neural network. The optimization framework aims dynamically adjusts the reference value for the proportional-integral (PI) controller. The proposed approach was able to dynamically adjust the oxygen pumping requirement and thus reduce the electricity consumption compared to the default PI controller adopted by the Benchmark Simulation Model no. 2 (BSM2).

Radiation therapy is one of the main treatments against cancer. Intensity Modulated Radiation Therapy (IMRT) is one type of radiation therapy that allows a high degree of conformity between the radiation intensities and the areas to treat. The planning of a radiation treatment for a given patient is crucial for obtaining the desired goals (being able to destroy cancer cells while preserving the healthy ones). In clinical practice, treatment planning is most of the times based on a lengthy trial-and-error procedure during which the planner interacts with a treatment planning system trying to find a treatment that complies with the medical prescription. One of the first decisions the planner has to make is on the angles to be used to deliver radiation. In clinical practice, most of the times, the number of angles to be used is defined a priori based on the experience of the planner with similar cases. Often, the solution that is used is the equidistant solution, where all angles are equally apart. In this paper we propose the use of Differential Evolution (DE) for determining in an automated way the set of angles that should be used in a given IMRT treatment. Solutions obtained after the DE optimization are then compared with the equidistant solution. Preliminary results considering ten already treated patients at the Portuguese Institute of Oncology of Coimbra IPOCFG are presented.

Radiotherapy is one of the main treatments used against cancer. Radiotherapy uses radiation to destroy cancerous cells trying, at the same time, to minimize the damages in healthy tissues. The planning of a radiotherapy treatment is patient dependent, resulting in a lengthy trial and error procedure until a treatment complying as most as possible with the medical prescription is found. Intensity Modulated Radiation Therapy (IMRT) is one technique of radiation treatment that allows the achievement of a high degree of conformity between the area to be treated and the dose absorbed by healthy tissues. Nevertheless, it is still not possible to eliminate completely the potential treatments' side-effects. In this retrospective study we use the clinical data from patients with head-and-neck cancer treated at the Portuguese Institute of Oncology of Coimbra and explore the possibility of classifying new and untreated patients according to the probability of xerostomia 12 months after the beginning of IMRT treatments by using a logistic regression approach. The results obtained show that the classifier presents a high discriminative ability in predicting the binary response “at risk for xerostomia at 12 months”.