Search

Neural Architecture Search (NAS) has recently become a topic of great interest. However, there is a potentially impactful issue within NAS that remains largely unrecognized: noise. Due to stochastic factors in neural network initialization, training, and the chosen train/validation dataset split, the performance evaluation of a neural network architecture, which is often based on a single learning run, is also stochastic. This may have a particularly large impact if a dataset is small. We therefore propose to reduce this noise by evaluating architectures based on average performance over multiple network training runs using different random seeds and cross-validation. We perform experiments for a combinatorial optimization formulation of NAS in which we vary noise reduction levels. We use the same computational budget for each noise level in terms of network training runs, i.e., we allow less architecture evaluations when averaging over more training runs. Multiple search algorithms are considered, including evolutionary algorithms which generally perform well for NAS. We use two publicly available datasets from the medical image segmentation domain where datasets are often limited and variability among samples is often high. Our results show that reducing noise in architecture evaluations enables finding better architectures by all considered search algorithms.

Explainable artificial intelligence (XAI) is an important and rapidly expanding research topic. The goal of XAI is to gain trust in a machine learning (ML) model through clear insights into how the model arrives at its predictions. Genetic programming (GP) is often cited as being uniquely well-suited to contribute to XAI because of its capacity to learn (small) symbolic models that have the potential to be interpreted. Nevertheless, like many ML algorithms, GP typically results in a single best model. However, in practice, the best model in terms of training error may well not be the most suitable one as judged by a domain expert for various reasons, including overfitting, multiple different models existing that have similar accuracy and unwanted errors on particular data points due to typical accuracy measures like mean squared error. Hence, to increase chances that domain experts deem a resulting model plausible, it becomes important to be able to explicitly search for multiple, diverse, high-quality models that trade-off different meanings of accuracy. In this paper, we achieve exactly this with a novel multi-modal multi-tree multi-objective GP approach that extends a modern model-based GP algorithm known as GP-GOMEA that is already effective at searching for small expressions.

Model-Based Evolutionary Algorithms (MBEAs) can be highly scalable by virtue of linkage (or variable interaction) learning. This requires, however, that the linkage model can capture the exploitable structure of a problem. Usually, a single type of linkage structure is attempted to be captured using models such as a linkage tree. However, in practice, problems may exhibit multiple linkage structures. This is for instance the case in multi-objective optimization when the objectives have different linkage structures. This cannot be modelled sufficiently well when using linkage models that aim at capturing a single type of linkage structure, deteriorating the advantages brought by MBEAs. Therefore, here, we introduce linkage kernels, whereby a linkage structure is learned for each solution over its local neighborhood. We implement linkage kernels into the MBEA known as GOMEA that was previously found to be highly scalable when solving various problems. We further introduce a novel benchmark function called Best-of-Traps (BoT) that has an adjustable degree of different linkage structures. On both BoT and a worst-case scenario-based variant of the well-known MaxCut problem, we experimentally find a vast performance improvement of linkage-kernel GOMEA over GOMEA with a single linkage tree as well as the MBEA known as DSMGA-II.

Dimensionality reduction (DR) is an important technique for data exploration and knowledge discovery. However, most of the main DR methods are either linear (e.g., PCA), do not provide an explicit mapping between the original data and its lower-dimensional representation (e.g., MDS, t-SNE, isomap), or produce mappings that cannot be easily interpreted (e.g., kernel PCA, neural-based autoencoder). Recently genetic programming (GP) has been used to evolve interpretable DR mappings in the form of symbolic expressions. There exists a number of ways in which GP can be used to this end and no study exists that performs a comparison. In this paper, we fill this gap by comparing existing GP methods as well as devising new ones. We evaluate our methods on several benchmark datasets based on predictive accuracy and on how well the original features can be reconstructed using the lower-dimensional representation only. Finally we qualitatively assess the resulting expressions and their complexity. We find that various GP methods can be competitive with state-of-the-art DR algorithms and that they have the potential to produce interpretable DR mappings.

The Multi-Objective Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (MO-RV-GOMEA) has been proven effective and eficient in solving real-world problems. A prime example is optimizing treatment plans for prostate cancer brachytherapy, an internal form of radiation treatment, for which equally important clinical aims from a base protocol are grouped into two objectives and bi-objectively optimized. This use of MO-RV-GOMEA was recently successfully introduced into clinical practice. Brachytherapy can also play an important role in treating cervical cancer. However, using the same approach to optimize treatment plans often does not immediately lead to clinically desirable results. Concordantly, medical experts indicate that they use additional aims beyond the cervix base protocol. Moreover, these aims have different priorities and can be patient-specifically adjusted. For this reason, we propose a novel adaptive objective configuration method to use with MO-RV-GOMEA so that we can accommodate additional aims of this nature. Based on results using only the base protocol, in consultation with medical experts, we configured key additional aims. We show how, for 10 patient cases, the new approach achieves the intended result, properly taking into account the additional aims. Consequently, plans resulting from the new approach are preferred by medical specialists in 8/10 cases.

Genetic Programming (GP) can make an important contribution to explainable artificial intelligence because it can create symbolic expressions as machine learning models. Nevertheless, to be explainable, the expressions must not become too large. This may, however, limit their potential to be accurate. The re-use of subexpressions has the unique potential to mitigate this issue. The Genetic Programming Gene-pool Optimal Mixing Evolutionary Algorithm (GP-GOMEA) is a recent model-based GP approach that has been found particularly capable of evolving small expressions. However, its tree representation offers no explicit mechanisms to re-use subexpressions. By contrast, the graph representation in Cartesian GP (CGP) is natively capable of re-use. For this reason, we introduce CGP-GOMEA, a variant of GP-GOMEA that uses graphs instead of trees. We experimentally compare various configurations of CGP-GOMEA with GP-GOMEA and find that CGP-GOMEA performs on par with GP-GOMEA on three common datasets. Moreover, CGP-GOMEA is found to produce models that re-use subexpressions more often than GP-GOMEA uses duplicate subexpressions. This indicates that CGP-GOMEA has unique added potential, allowing to find even smaller expressions than GP-GOMEA with similar accuracy., Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Algorithmics

Objectives: Children with cancer often experience sleep problems, which are associated with many negative physical and psychological health outcomes, as well as with a lower quality of life. Therefore, interventions are strongly required to improve sleep in this population. We evaluated interactive education with respect to sleep hygiene with a social robot at a pediatric oncology outpatient clinic regarding the feasibility, experiences, and preliminary effectiveness. Methods: Researchers approached children (8 to 12 years old) who were receiving anticancer treatment and who were visiting the outpatient clinic with their parents during the two-week study period. The researchers completed observation forms regarding feasibility, and parents completed the Children’s Sleep Hygiene Scale before and two weeks after the educational regimen. The experiences of children and parents were evaluated in semi-structured interviews. We analyzed open answers by labeling each answer with a topic reflecting the content and collapsed these topics into categories. We used descriptive statistics to describe the feasibility and experiences, and a dependent-samples t-test to evaluate the preliminary effectiveness. Results: Twenty-eight families participated (58% response rate) and all interactions with the robot were completed. The children and parents reported that they learned something new (75% and 50%, respectively), that they wanted to learn from the robot more often (83% and 75%, respectively), and that they applied the sleeping tips from the robot afterwards at home (54%). Regarding the preliminary effectiveness, children showed a statistically significant improvement in their sleep hygiene (p = 0.047, d = 0.39). Conclusions: Providing an educational regimen on sleep hygiene in a novel, interactive way by using a social robot at the outpatient clinic seemed feasible, and the children and parents mostly exhibited positive reactions. We found preliminary evidence that the sleep h, Interactive Intelligence

SPIE BiOS (6-12 March, 2021), Item does not contain fulltext, Functional near infrared spectroscopy (fNIRS) is used for brain hemodynamic assessment. Cortical hemodynamics are reliably estimated when the recorded signal has a sufficient quality. This is acquired when fNIRS optodes have proper scalp coupling. A lack of proper scalp coupling causes false positives and false negatives. Therefore, developing an objective algorithm for determining fNIRS signal quality is of great importance. In this study, we developed a machine learning-based algorithm for quantitatively rating fNIRS signal quality. Our promising results confirm the efficacy of the algorithm in determining fNIRS signal quality and hence decreasing misinterpretations.

Contains fulltext : 233623.pdf (Publisher’s version ) (Open Access)

Perioperative applications of near-infrared spectroscopy (NIRS) to monitor regional tissue oxygenation and perfusion in cardiac and noncardiac surgery are of increasing interest in neonatal care. Complex neonatal surgery can impair adequate oxygen delivery and tissue oxygen consumption and increase the risk of neurodevelopmental delay. Coupled with conventional techniques, NIRS monitoring may enable targeted hemodynamic management of the circulation in both cardiac and noncardiac surgical procedures. In this narrative review, we discuss the application of perioperative NIRS in specific neonatal interventions, including surgical intervention for congenital heart defects, definitive closure of the patent ductus arteriosus, neurological and gastrointestinal disorders, and use of extracorporeal membrane oxygenation. We identified areas for future research within disease-specific indications and offer a roadmap to aid in developing evidence-based targeted diagnostic and management strategies in neonates. Impact: There is growing recognition that perioperative NIRS monitoring, used in conjunction with conventional monitoring, may provide critical hemodynamic information that either complements clinical impressions or delivers novel physiologic insight into the neonatal circulatory and perfusion pathways.Copyright © 2021, The Author(s), under exclusive licence to the International Pediatric Research Foundation, Inc.

The Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (RV-GOMEA) has previously been successfully used to achieve highly scalable optimization of various real-world problems in a gray-box optimization setting. Deformable Image Registration (DIR) is a multi-objective problem, aimed at finding the most likely non-rigid deformation of a given source image so that it matches a given target image. We specifically consider the case where the deformation model allows for finite-element-type modeling of tissue properties. This optimization problem is non-smooth, necessitating techniques like EAs to get good results. Though the objectives of DIR are non-separable, non-neighboring regions of the deformation grid are conditionally independent. We show that GOMEA allows to exploit such knowledge through the large-scale parallel application of variation steps, where each is only accepted when leading to an improvement, on a Graphics Processing Unit (GPU). On various 2-dimensional DIR problems, we find that this way, similar results can be achieved as when sequential processing is performed, while allowing for substantial speed-ups (up to a factor of 111) for the highest-dimensional problems (i.e., the highest deformation-grid resolution). This work opens the door to the extension of this type of DIR to larger (3-dimensional) deformation grids, and its application to other real-world problems.

The aim of multimodal optimization (MMO) is to obtain all global optima of an optimization problem. In this chapter, we introduce a general framework for two-phase MMO evolutionary algorithms (EAs), in which different high-fitness regions (niches) are located in the first phase via clustering, and each of the located niches is separately optimized with a core search algorithm in the second phase. One such two-phase MMO EA is the Hill-Valley Evolutionary Algorithm (HillVall-EA). In HillVallEA, the remarkably simple hill-valley clustering method is used. The idea behind hill-valley clustering is that two solutions belong to the same niche (valley) when there is no hill in between them, which can be easily tested by performing additional function evaluations. We compare hill-valley clustering to two other recently introduced fitness-informed clustering methods: nearest-better clustering and hierarchical Gaussian mixture learning. We show how these clustering methods, as well as different core search algorithms, influence the resulting optimization performance of the two-phase MMO framework on the commonly used CEC 2013 niching benchmark suite. Our results show that HillVallEA, equipped with the core search algorithm Adapted Maximum-Likelihood Gaussian Model Univariate (AMu) as core search algorithm, outperforms all other MMO EAs, both within the limited benchmark budget, and in the long run. HillVallEA-AMu was the winner of the GECCO niching competition in 2018 and 2019, and is currently, to the best of our knowledge, the best performing algorithm on this benchmark suite.

The Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) is a model-based EA framework that has been shown to perform well in several domains, including Genetic Programming (GP). Differently from traditional EAs where variation acts blindly, GOMEA learns a model of interdependencies within the genotype, that is, the linkage, to estimate what patterns to propagate. In this article, we study the role of Linkage Learning (LL) performed by GOMEA in Symbolic Regression (SR). We show that the non-uniformity in the distribution of the genotype in GP populations negatively biases LL, and propose a method to correct for this. We also propose approaches to improve LL when ephemeral random constants are used. Furthermore, we adapt a scheme of interleaving runs to alleviate the burden of tuning the population size, a crucial parameter for LL, to SR. We run experiments on 10 real-world datasets, enforcing a strict limitation on solution size, to enable interpretability. We find that the new LL method outperforms the standard one, and that GOMEA outperforms both traditional and semantic GP. We also find that the small solutions evolved by GOMEA are competitive with tuned decision trees, making GOMEA a promising new approach to SR.

Purpose. Recently, we introduced a bi-objective optimization approach based on dose-volume indices to automatically create clinically goodHDRprostate brachytherapy plans. To calculate dose-volume indices, a reconstruction algorithm is used to determine the 3D organ shape from 2D contours, inevitably containing settings that influence the result.Weaugment the optimization approach to quickly find plans that are robust to differences in 3D reconstruction. Methods. Studied reconstruction settings were: Interpolation between delineated organ contours, overlap between contours, and organ shape at the top and bottom contour. Two options for each setting yields 8 possible 3D organ reconstructions per patient, over which the robust model defines minimax optimization. For the original model, settings were based on our treatment planning system. Both models were tested on data of 26 patients and compared by re-evaluating selected optimized plans both in the original model (1 organ reconstruction, the difference determines the cost), and in the robust model (8 organ reconstructions, the difference determines the benefit). Results. Robust optimization increased the run time from 3 to 6 min. The median cost for robust optimization as observed in the original model was-0.25% in the dose-volume indices with a range of [-0.01%,-1.03%]. The median benefit of robust optimization as observed in the robust model was 0.93% with a range of [0.19%, 4.1

We propose a novel surrogate-assisted Evolutionary Algorithm for solving expensive combinatorial optimization problems. We integrate a surrogate model, which is used for fitness value estimation, into a state-of-the-art P3-like variant of the Gene-Pool Optimal Mixing Algorithm (GOMEA) and adapt the resulting algorithm for solving non-binary combinatorial problems. We test the proposed algorithm on an ensemble learning problem. Ensembling several models is a common Machine Learning technique to achieve better performance. We consider ensembles of several models trained on disjoint subsets of a dataset. Finding the best dataset partitioning is naturally a combinatorial non-binary optimization problem. Fitness function evaluations can be extremely expensive if complex models, such as Deep Neural Networks, are used as learners in an ensemble. Therefore, the number of fitness function evaluations is typically limited, necessitating expensive optimization techniques. In our experiments we use five classification datasets from the OpenML-CC18 benchmark and Support-vector Machines as learners in an ensemble. The proposed algorithm demonstrates better performance than alternative approaches, including Bayesian optimization algorithms. It manages to find better solutions using just several thousand fitness function evaluations for an ensemble learning problem with up to 500 variables., Algorithmics

It is well known nowadays that radioactivity can destroy the living cells it interacts with. it is therefore unsurprising that radioactive sources, such as iodine-125, were historically developed for treatment purposes within radiation oncology with the goal of damaging malignant cells. however, since then, new techniques have been invented that make creative use of the same radioactivity properties of these sources for medi- cal applications. here, we review two distinct kinds of therapeutic uses of radioactive sources with applications to prostate, cervical, and breast cancer: brachytherapy and radioactive seed localization. in brachytherapy (BT), the radioactive sources are used for internal radiation treatment. current approaches make use of real-time image guidance, for instance by means of magnetic resonance imaging, ultrasound, computed tomog- raphy, and sometimes positron emission tomography, depending on clinical availability and cancer type. Such image-guided BT for prostate and cervical cancer presents a promising alternative and/or addition to external beam radiation treatments or surgical resections. radioactive sources can also be used for radio-guided tumor localization during surgery, for which the example of iodine-125 seed use in breast cancer is given. radioactive seed localization (rSl) is increasingly popular as an alternative tumor localization technique during breast cancer surgery. Advantages of applying RSL include added flexibility in the clinical scheduling logistics, an increase in tumor localization accuracy, and higher patient satisfaction; safety measures do however have to be employed. We exemply the implementation of rSl in a clinic through our experi- ences at the netherlands cancer institute.

code for "On genetic programming representations and fitness functions for interpretable dimensionality reduction"

SAGOMEA is an efficient surrogate-assisted Evolutionary Algorithm for combinatorial (discrete) optimization problems with expensive fitness functions.

The advent of Machine Learning (ML) is proving extremely beneficial in many healthcare applications. In pediatric oncology, retrospective studies that investigate the relationship between treatment and late adverse effects still rely on simple heuristics. To capture the effects of radiation treatment, treatment plans are typically simulated on virtual surrogates of patient anatomy called phantoms. Currently, phantoms are built to represent categories of patients based on reasonable yet simple criteria. This often results in phantoms that are too generic to accurately represent individual anatomies. We present a novel approach that combines imaging data and ML to build individualized phantoms automatically. We design a pipeline that, given features of patients treated in the pre-3D planning era when only 2D radiographs were available, as well as a database of 3D Computed Tomography (CT) imaging with organ segmentations, uses ML to predict how to assemble a patient-specific phantom. Using 60 abdominal CTs of pediatric patients between 2 to 6 years of age, we find that our approach delivers significantly more representative phantoms compared to using current phantom building criteria, in terms of shape and location of two considered organs (liver and spleen), and shape of the abdomen. Furthermore, as interpretability is often central to trust ML models in medical contexts, among other ML algorithms we consider the Gene-pool Optimal Mixing Evolutionary Algorithm for Genetic Programming (GP-GOMEA), that learns readable mathematical expression models. We find that the readability of its output does not compromise prediction performance as GP-GOMEA delivered the best performing models.

Feature construction can substantially improve the accuracy of Machine Learning (ML) algorithms. Genetic Programming (GP) has been proven to be effective at this task by evolving non-linear combinations of input features. GP additionally has the potential to improve ML explainability since explicit expressions are evolved. Yet, in most GP works the complexity of evolved features is not explicitly bound or minimized though this is arguably key for explainability. In this article, we assess to what extent GP still performs favorably at feature construction when constructing features that are (1) Of small-enough number, to enable visualization of the behavior of the ML model; (2) Of small-enough size, to enable interpretability of the features themselves; (3) Of sufficient informative power, to retain or even improve the performance of the ML algorithm. We consider a simple feature construction scheme using three different GP algorithms, as well as random search, to evolve features for five ML algorithms, including support vector machines and random forest. Our results on 21 datasets pertaining to classification and regression problems show that constructing only two compact features can be sufficient to rival the use of the entire original feature set. We further find that a modern GP algorithm, GP-GOMEA, performs best overall. These results, combined with examples that we provide of readable constructed features and of 2D visualizations of ML behavior, lead us to positively conclude that GP-based feature construction still works well when explicitly searching for compact features, making it extremely helpful to explain ML models.

Purpose: Bi-objective simultaneous optimization of catheter positions and dwell times for high-dose-rate (HDR) prostate brachytherapy, based directly on dose-volume indices, has shown promising results. However, optimization with the state-of-the-art evolutionary algorithm MO-RV-GOMEA so far required several hours of runtime, and resulting catheter positions were not always clinically feasible. The aim of this study is to extend the optimization model and apply GPU parallelization to achieve clinically acceptable computation times. The resulting optimization procedure is compared with a previously introduced method based solely on geometric criteria, the adapted Centroidal Voronoi Tessellations (CVT) algorithm. Methods: Bi-objective simultaneous optimiza

There has recently been great progress in automatic segmentation of medical images with deep learning algorithms. In most works observer variation is acknowledged to be a problem as it makes training data heterogeneous but so far no attempts have been made to explicitly capture this variation. Here, we propose an approach capable of mimicking different styles of segmentation, which potentially can improve quality and clinical acceptance of automatic segmentation methods. In this work, instead of training one neural network on all available data, we train several neural networks on subgroups of data belonging to different segmentation variations separately. Because a priori it may be unclear what styles of segmentation exist in the data and because different styles do not necessarily map one-on-one to different observers, the subgroups should be automatically determined. We achieve this by searching for the best data partition with a genetic algorithm. Therefore, each network can learn a specific style of segmentation from grouped training data. We provide proof of principle results for open-sourced prostate segmentation MRI data with simulated observer variations. Our approach provides an improvement of up to 23% (depending on simulated variations) in terms of Dice and surface Dice coefficients compared to one network trained on all data.

We address the real-world problem of automating the design of high-quality prostate cancer treatment plans in case of high-dose-rate brachytherapy, a form of internal radiotherapy. For this, recently a bi-objective real-valued problem formulation was introduced. With a GPU parallelization of the Multi-Objective Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (MO-RV-GOMEA), good treatment plans were found in clinically acceptable running times. However, optimizing a treatment plan and delivering it to the patient in practice is a two-stage decision process and involves a number of uncertainties. Firstly, there is uncertainty in the identified organ boundaries due to the limited resolution of the medical images. Secondly, the treatment involves placing catheters inside the patient, which always end up (slightly) different from what was optimized. An important factor is therefore the robustness of the final treatment plan to these uncertainties. In this work, we show how we can extend the evolutionary optimization approach to find robust plans using multiple scenarios without linearly increasing the amount of required computation effort, as well as how to deal with these uncertainties efficiently when taking into account the sequential decision-making moments. The performance is tested on three real-world patient cases. We find that MO-RV-GOMEA is equally well capable of solving the more complex robust problem formulation, resulting in a more realistic reflection of the treatment plan quali

It is known that to achieve efficient scalability of an Evolutionary Algorithm (EA), dependencies (also known as linkage) must be properly taken into account during variation. In a Gray-Box Optimization (GBO) setting, exploiting prior knowledge regarding these dependencies can greatly benefit optimization. We specifically consider the setting where partial evaluations are possible, meaning that the partial modification of a solution can be efficiently evaluated. Such problems are potentially very difficult, for example, non-separable, multimodal, and multiobjective. The Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) can effectively exploit partial evaluations, leading to a substantial improvement in performance and scalability. GOMEA was recently shown to be extendable to real-valued optimization through a combination with the real-valued estimation of distribution algorithm AMaLGaM. In this article, we definitively introduce the Real-Valued GOMEA (RV-GOMEA), and introduce a new variant, constructed by combining GOMEA with what is arguably the best-known real-valued EA, the Covariance Matrix Adaptation Evolution Strategies (CMA-ES). Both variants of GOMEA are compared to L-BFGS and the Limited Memory CMA-ES (LM-CMA-ES). We show that both variants of RV-GOMEA achieve excellent performance and scalability in a GBO setting, which can be orders of magnitude better than that of EAs unable to efficiently exploit the GBO setting.

Anatomical landmark correspondences in medical images can provide additional guidance information for the alignment of two images, which, in turn, is crucial for many medical applications. However, manual landmark annotation is labor-intensive. Therefore, we propose an end-to-end deep learning approach to automatically detect landmark correspondences in pairs of two-dimensional (2D) images. Our approach consists of a Siamese neural network, which is trained to identify salient locations in images as landmarks and predict matching probabilities for landmark pairs from two different images. We trained our approach on 2D transverse slices from 168 lower abdominal Computed Tomography (CT) scans. We tested the approach on 22,206 pairs of 2D slices with varying levels of intensity, affine, and elastic transformations. The proposed approach finds an average of 639, 466, and 370 landmark matches per image pair for intensity, affine, and elastic transformations, respectively, with spatial matching errors of at most 1 mm. Further, more than 99% of the landmark pairs are within a spatial m

Purpose: Current phantoms used for the dose reconstruction of long-term childhood cancer survivors lack individualization. We design a method to predict highly individualized abdominal three-dimensional (3-D) phantoms automatically. Approach: We train machine learning (ML) models to map (2-D) patient features to 3-D organ-at-risk (OAR) metrics upon a database of 60 pediatric abdominal computed tomographies with liver and spleen segmentations. Next, we use the models in an automatic pipeline that outputs a personalized phantom given the patient's features, by assembling 3-D imaging from the database. A step to improve phantom realism (i.e., avoid OAR overlap) is included. We compare five ML algorithms, in terms of predicting OAR left-right (LR), anterior-posterior (AP), inferior-superior (IS) positions, and surface Dice-Sørensen coefficient (sDSC). Furthermore, two existing human-designed phantom construction criteria and two additional control methods are investigated for comparison. Results: Different ML algorithms result in similar test mean absolute errors: ∼8 mm for liver LR, IS, and spleen AP, IS; ∼5 mm for liver AP and spleen LR; ∼80 % for abdomen sDSC; and ∼60 % to 65% for liver and spleen sDSC. One ML algorithm (GP-GOMEA) significantly performs the best for 6/9 metrics. The control methods and the human-designed criteria in particular perform generally worse, sometimes substantially (+5-mm error for spleen IS,-10 % sDSC for liver). The automatic step to improve realism generally results in limited metric accuracy loss, but fails in one case (out of 60). Conclusion: Our ML-based pipeline leads to phantoms that are significantly and substantially more individualized than currently used human-designed criteria.

Performing large-scale three-dimensional radiation dose reconstruction for patients requires a large amount of manual work. We present an image processing-based pipeline to automatically reconstruct radiation dose. The pipeline was designed for childhood cancer survivors that received abdominal radiotherapy with anterior-to-posterior and posterior-to-anterior field set-up. First, anatomical landmarks are automatically identified on two-dimensional radiographs. Second, these landmarks are used to derive parameters to emulate the geometry of the plan on a surrogate computed tomography. Finally, the plan is emulated and used as input for dose calculation. For qualita

To study radiotherapy-related adverse effects, detailed dose information (3D distribution) is needed for accurate dose-effect modeling. For childhood cancer survivors who underwent radiotherapy in the pre-CT era, only 2D radiographs were acquired, thus 3D dose distributions must be reconstructed from limited information. State-of-the-art methods achieve this by using 3D surrogate anatomies. These can however lack personalization and lead to coarse reconstructions. We present and validate a surrogate-free dose reconstruction method based on Machine Learning (ML). Abdominal planning CTs (n = 142) of recently-treated childhood cancer patients were gathered, their organs at risk were segmented, and 300 artificial Wilms' tumor plans were sampled automatically. Each artificial plan was automatically emulated on the 142 CTs, resulting in 42,600 3D dose distributions from which dose-volume metrics were derived. Anatomical features were extracted from digitally reconstructed radiographs simulated from the CTs to resemble historical radiographs. Further, patient and radiotherapy plan features typically available from historical treatment records were collected. An evolutionary ML algorithm was then used to link features to dose-volume metrics. Besides 5-fold cross validation, a further evaluation was done on an independent dataset of five CTs each associated with two clinical plans. Cross-validation resulted in mean absolute errors ≤ 0.6 Gy for organs completely inside or outside the field. For organs positioned at the edge of the field, mean absolute errors ≤ 1.7 Gy for Dmean, ≤ 2.9 Gy for 2cc,}, and ≤ 13% for V5 Gy10 Gy, were obtained, without systematic bias. Similar results were found for the independent dataset. To conclude, we proposed a novel organ dose reconstruction method that uses ML models to predict dose-volume metric values given patient and plan features. Our approach is not only accurate, but also efficient, as the setup of a surrogate is no longer needed.

Often, real-world problems are of the gray-box type. It has been shown that the Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (RV-GOMEA) is in principle capable of exploiting such a Gray-Box Optimization (GBO) setting using linkage models that capture dependencies between problem variables, resulting in excellent performance and scalability on both benchmark and real-world problems that allow for partial evaluations. However, linkage models proposed for RV-GOMEA so far cannot explicitly capture overlapping dependencies. Consequently, performance degrades if such dependencies exist. In this paper, we therefore introduce various ways of using conditional linkage models in RV-GOMEA. Their use is compared to that of non-conditional models, and to VkD-CMA, whose performance is among the state of the art, on various benchmark problems with overlapping dependencies. We find that RV-GOMEA with conditional linkage models achieves the best scalability on most problems, with conditional models leading to similar or better performance than non-conditional models. We conclude that the introduction of conditional linkage models to RV-GOMEA is an important contribution, as it expands the set of problems for which optimization in a GBO setting results in substantially improved performance and scalability. In future work, conditional linkage models may prove to benefit the optimization of real-world problems.

We present an automatic bi-objective parameter-tuning approach for inverse planning methods for high-dose-rate prostate brachytherapy, which aims to overcome the difficult and time-consuming manual parameter tuning that is currently required to obtain patient-specific high-quality treatment plans. We modelled treatment planning as a bi-objective optimization problem, in which dose-volume-based planning criteria related to target coverage are explicitly separated from organ-sparing criteria. When this model is optimized, a large set of high-quality plans with different trade-offs can be obtained. This set can be visualized as an insightful patient-specific trade-off curve. In our parameter-tuning approach, the parameters of inverse planning methods are automatically tuned, aimed to maximize the two objectives of the bi-objective planning model. By generating trade

Evolutionary algorithms (EAs) are the preferred method for solving black-box multi-objective optimization problems, but when gradients of the objective functions are available, it is not straightforward to exploit these efficiently. By contrast, gradient-based optimization is well-established for single-objective optimization. A single-objective reformulation of the multi-objective problem could therefore offer a solution. Of particular interest to this end is the recently introduced uncrowded hypervolume (UHV) indicator, which is Pareto compliant and also takes into account dominated solutions. In this work, we show that the gradient of the UHV can often be computed, which allows for a direct application of gradient ascent algorithms. We compare this new approach with two EAs for UHV optimization as well as with one gradient-based algorithm for optimizing the well-established hypervolume. On several bi-objective benchmarks, we find that gradient-based algorithms outperform the tested EAs by obtaining a better hypervolume with fewer evaluations whenever exact gradients of the multiple objective functions are available and in case of small evaluation budgets. For larger budgets, however, EAs perform similarly or better. We further find that, when finite differences are used to approximate the gradients of the multiple objectives, our new gradient-based algorithm is still competitive with EAs in most considered benchmarks. Implementations are available at https://github.com/scmaree/uncrowded-hypervolume.

The aim of bi-objective optimization is to obtain an approximation set of (near) Pareto optimal solutions. A decision maker then navigates this set to select a final desired solution, often using a visualization of the approximation front. The front provides a navigational ordering of solutions to traverse, but this ordering does not necessarily map to a smooth trajectory through decision space. This forces the decision maker to inspect the decision variables of each solution individually, potentially making navigation of the approximation set unintuitive. In this work, we aim to improve approximation set navigability by enforcing a form of smoothness or continuity between solutions in terms of their decision variables. Imposing smoothness as a restriction upon common domination-based multi-objective evolutionary algorithms is not straightforward. Therefore, we use the recently introduced uncrowded hypervolume (UHV) to reformulate the multi-objective optimization problem as a single-objective problem in which parameterized approximation sets are directly optimized. We study here the case of parameterizing approximation sets as smooth Bézier curves in decision space. We approach the resulting single-objective problem with the gene-pool optimal mixing evolutionary algorithm (GOMEA), and we call the resulting algorithm BezEA. We analyze the behavior of BezEA and compare it to optimization of the UHV with GOMEA as well as the domination-based multi-objective GOMEA. We show that high-quality approximation sets can be obtained with BezEA, sometimes even outperforming the domination- and UHV-based algorithms, while smoothness of the navigation trajectory through decision space is guaranteed.

This repository contains implementations (C++) of different uncrowded-hypervolume based methods for (gradient-free) multi-objective optimization. Using the (uncrowded) hypervolume, multi-objective optimization problems can be formulated as (high-dimensional) single-objective optimization problems, in the hypervolume of a solution sets is optimized. For details, see the corresponding publications/preprints.