Embedding machine learning based toxicity models within radiotherapy treatment plan optimization.

Objective.This study addresses radiation-induced toxicity (RIT) challenges in radiotherapy (RT) by developing a personalized treatment planning framework. It leverages patient-specific data and dosimetric information to create an optimization model that limits adverse side effects using constraints learned from historical data.Approach.The study uses the optimization with constraint learning (OCL) framework, incorporating patient-specific factors into the optimization process. It consists of three steps: optimizing the baseline treatment plan using population-wide dosimetric constraints; training a machine learning (ML) model to estimate the patient's RIT for the baseline plan; and adapting the treatment plan to minimize RIT using ML-learned patient-specific constraints. Various predictive models, including classification trees, ensembles of trees, and neural networks, are applied to predict the probability of grade 2+ radiation pneumonitis (RP2+) for non-small cell lung (NSCLC) cancer patients three months post-RT. The methodology is assessed with four high RP2+ risk NSCLC patients, with the goal of optimizing the dose distribution to constrain the RP2+ outcome below a pre-specified threshold. Conventional and OCL-enhanced plans are compared based on dosimetric parameters and predicted RP2+ risk. Sensitivity analysis on risk thresholds and data uncertainty is performed using a toy NSCLC case.Main results.Experiments show the methodology's capacity to directly incorporate all predictive models into RT treatment planning. In the four patients studied, mean lung dose and V20 were reduced by an average of 1.78 Gy and 3.66%, resulting in an average RP2+ risk reduction from 95% to 42%. Notably, this reduction maintains tumor coverage, although in two cases, sparing the lung slightly increased spinal cord max-dose (0.23 and 0.79 Gy).Significance.By integrating patient-specific information into learned constraints, the study significantly reduces adverse side effects like RP2+ without compromising target coverage. This unified framework bridges the gap between predicting toxicities and optimizing treatment plans in personalized RT decision-making.

Extensive red blood cell matching considering patient alloimmunization risk.

BACKGROUND AND OBJECTIVES: Red blood cell (RBC) transfusions pose a risk of alloantibody development in patients. For patients with increased alloimmunization risk, extended preventive matching is advised, encompassing not only the ABO-D blood groups but also the most clinically relevant minor antigens: C, c, E, e, K, Fya, Fyb, Jka, Jkb, S and s. This study incorporates patient-specific data and the clinical consequences of mismatching into the allocation process. MATERIALS AND METHODS: We have redefined the MINimize Relative Alloimmunization Risks (MINRAR) model to include patient group preferences in selecting RBC units from a finite supply. A linear optimization approach was employed, considering both antigen immunogenicity and the clinical impact of mismatches for specific patient groups. We also explore the advantages of informing the blood bank about scheduled transfusions, allowing for a more strategic blood distribution. The model is evaluated using historical data from two Dutch hospitals, measuring shortages and minor antigen mismatches. RESULTS: The updated model, emphasizing patient group-specific considerations, achieves a similar number of mismatches as the original, yet shifts mismatches among patient groups and antigens, reducing expected alloimmunization consequences. Simultaneous matching for multiple hospitals at the distribution centre level, considering scheduled demands, led to a 30% decrease in mismatches and a 92% reduction in shortages. CONCLUSION: The reduction of expected alloimmunization consequences by incorporating patient group preferences demonstrates our strategy's effectiveness for patient health. Substantial reductions in mismatches and shortages with multi-hospital collaboration highlights the importance of sharing information in the blood supply chain.

Online adaptive planning methods for intensity-modulated radiotherapy.

Online adaptive radiation therapy aims at adapting a patient's treatment plan to their current anatomy to account for inter-fraction variations before daily treatment delivery. As this process needs to be accomplished while the patient is immobilized on the treatment couch, it requires time-efficient adaptive planning methods to generate a quality daily treatment plan rapidly. The conventional planning methods do not meet the time requirement of online adaptive radiation therapy because they often involve excessive human intervention, significantly prolonging the planning phase. This article reviews the planning strategies employed by current commercial online adaptive radiation therapy systems, research on online adaptive planning, and artificial intelligence's potential application to online adaptive planning.

Determining optimal locations for blood distribution centers.

BACKGROUND: Blood banks have to be thoughtful about supply chain decisions to effectively satisfy the blood product demand of hospitals. These decisions include the number and locations of distribution centers (DC), as this has a strong impact on both transportation cost and the ability to deliver emergency orders in time. STUDY DESIGN AND METHODS: We propose a mixed-integer linear programming approach to find optimal DC locations for supplying individual hospitals. The model maximizes the number of hospitals reachable from a DC within a given time-limit, and minimizes transportation cost. The minimal amount of data required is a set of hospital locations. The model can be further attuned to the user's needs by adding various model extensions. The model's use is demonstrated by two case studies, considering the blood banks of the Netherlands and Finland. RESULTS: For both case studies re-locating the DCs would result in a reduction of transportation cost of about 10% without affecting the reliability of delivery. In addition, to save facility exploitation costs, the number of DCs may be reduced in both countries while maintaining the reliability of delivery. The model was also shown to be robust against changes in hospital ordering behavior. DISCUSSION: We demonstrated the general usability and added value of the model by successfully optimizing the blood supply chains of the Netherlands and Finland, which differ substantially. Nonetheless, in both countries potential savings in both transportation and facility exploitation cost could be shown. The model code is open source and freely accessible online.

Robust dose-painting-by-numbers vs. nonselective dose escalation for non-small cell lung cancer patients.

PURPOSE: Theoretical studies have shown that dose-painting-by-numbers (DPBN) could lead to large gains in tumor control probability (TCP) compared to conventional dose distributions. However, these gains may vary considerably among patients due to (a) variations in the overall radiosensitivity of the tumor, (b) variations in the 3D distribution of intra-tumor radiosensitivity within the tumor in combination with patient anatomy, (c) uncertainties of the 3D radiosensitivity maps, (d) geometrical uncertainties, and (e) temporal changes in radiosensitivity. The goal of this study was to investigate how much of the theoretical gains of DPBN remain when accounting for these factors. DPBN was compared to both a homogeneous reference dose distribution and to nonselective dose escalation (NSDE), that uses the same dose constraints as DPBN, but does not require 3D radiosensitivity maps. METHODS: A fully automated DPBN treatment planning strategy was developed and implemented in our in-house developed treatment planning system (TPS) that is robust to uncertainties in radiosensitivity and patient positioning. The method optimized the expected TCP based on 3D maps of intra-tumor radiosensitivity, while accounting for normal tissue constraints, uncertainties in radiosensitivity, and setup uncertainties. Based on FDG-PETCT scans of 12 non-small cell lung cancer (NSCLC) patients, data of 324 virtual patients were created synthetically with large variations in the aforementioned parameters. DPBN was compared to both a uniform dose distribution of 60 Gy, and NSDE. In total, 360 DPBN and 24 NSDE treatment plans were optimized. RESULTS: The average gain in TCP over all patients and radiosensitivity maps of DPBN was 0.54 ± 0.20 (range 0-0.97) compared to the 60 Gy uniform reference dose distribution, but only 0.03 ± 0.03 (range 0-0.22) compared to NSDE. The gains varied per patient depending on the radiosensitivity of the entire tumor and the 3D radiosensitivity maps. Uncertainty in radiosensitivity led to a considerable loss in TCP gain, which could be recovered almost completely by accounting for the uncertainty directly in the optimization. CONCLUSIONS: Our results suggest that the gains of DPBN can be considerable compared to a 60 Gy uniform reference dose distribution, but small compared to NSDE for most patients. Using the robust DPBN treatment planning system developed in this work, the optimal DPBN treatment plan could be derived for any patient for whom 3D intra-tumor radiosensitivity maps are known, and can be used to select patients that might benefit from DPBN. NSDE could be an effective strategy to increase TCP without requiring biological information of the tumor.

When are static and adjustable robust optimization problems with constraint-wise uncertainty equivalent?

Adjustable robust optimization (ARO) generally produces better worst-case solutions than static robust optimization (RO). However, ARO is computationally more difficult than RO. In this paper, we provide conditions under which the worst-case objective values of ARO and RO problems are equal. We prove that when the uncertainty is constraint-wise, the problem is convex with respect to the adjustable variables and concave with respect to the uncertain parameters, the adjustable variables lie in a convex and compact set and the uncertainty set is convex and compact, then robust solutions are also optimal for the corresponding ARO problem. Furthermore, we prove that if some of the uncertain parameters are constraint-wise and the rest are not, then under a similar set of assumptions there is an optimal decision rule for the ARO problem that does not depend on the constraint-wise uncertain parameters. Also, we show for a class of problems that using affine decision rules that depend on all of the uncertain parameters yields the same optimal objective value as when the rules depend solely on the non-constraint-wise uncertain parameters. Finally, we illustrate the usefulness of these results by applying them to convex quadratic and conic quadratic problems.

A framework for inverse planning of beam-on times for 3D small animal radiotherapy using interactive multi-objective optimisation.

Advances in precision small animal radiotherapy hardware enable the delivery of increasingly complicated dose distributions on the millimeter scale. Manual creation and evaluation of treatment plans becomes difficult or even infeasible with an increasing number of degrees of freedom for dose delivery and available image data. The goal of this work is to develop an optimisation model that determines beam-on times for a given beam configuration, and to assess the feasibility and benefits of an automated treatment planning system for small animal radiotherapy. The developed model determines a Pareto optimal solution using operator-defined weights for a multiple-objective treatment planning problem. An interactive approach allows the planner to navigate towards, and to select the Pareto optimal treatment plan that yields the most preferred trade-off of the conflicting objectives. This model was evaluated using four small animal cases based on cone-beam computed tomography images. Resulting treatment plan quality was compared to the quality of manually optimised treatment plans using dose-volume histograms and metrics. Results show that the developed framework is well capable of optimising beam-on times for 3D dose distributions and offers several advantages over manual treatment plan optimisation. For all cases but the simple flank tumour case, a similar amount of time was needed for manual and automated beam-on time optimisation. In this time frame, manual optimisation generates a single treatment plan, while the inverse planning system yields a set of Pareto optimal solutions which provides quantitative insight on the sensitivity of conflicting objectives. Treatment planning automation decreases the dependence on operator experience and allows for the use of class solutions for similar treatment scenarios. This can shorten the time required for treatment planning and therefore increase animal throughput. In addition, this can improve treatment standardisation and comparability of research data within studies and among different institutes.

Dwell time modulation restrictions do not necessarily improve treatment plan quality for prostate HDR brachytherapy.

Inverse planning algorithms for dwell time optimisation in interstitial high-dose-rate (HDR) brachytherapy may produce solutions with large dwell time variations within catheters, which may result in undesirable selective high-dose subvolumes. Extending the dwell time optimisation model with a dwell time modulation restriction (DTMR) that limits dwell time differences between neighboring dwell positions has been suggested to eliminate this problem. DTMRs may additionally reduce the sensitivity for uncertainties in dwell positions that inevitably result from catheter reconstruction errors and afterloader source positioning inaccuracies. This study quantifies the reduction of high-dose subvolumes and the robustness against these uncertainties by applying a DTMR to template-based prostate HDR brachytherapy implants. Three different DTMRs were consecutively applied to a linear dose-based penalty model (LD) and a dose-volume based model (LDV), both obtained from literature. The models were solved with DTMR levels ranging from no restriction to uniform dwell times within catheters in discrete steps. Uncertainties were simulated on clinical cases using in-house developed software, and dose-volume metrics were calculated in each simulation. For the assessment of high-dose subvolumes, the dose homogeneity index (DHI) and the contiguous dose volume histogram were analysed. Robustness was measured by the improvement of the lowest D90% of the planning target volume (PTV) observed in the simulations. For (LD), a DTMR yields an increase in DHI of approximately 30% and reduces the size of the largest high-dose volume by 2-5 cc. However, this comes at a cost of a reduction in D90% of the PTV of 10%, which often implies that it drops below the desired minimum of 100%. For (LDV), none of the DTMRs were able to improve high-dose volume measures. DTMRs were not capable of improving robustness of PTV D90% against uncertainty in dwell positions for both models.

Mixed integer programming improves comprehensibility and plan quality in inverse optimization of prostate HDR brachytherapy.

Current inverse treatment planning methods that optimize both catheter positions and dwell times in prostate HDR brachytherapy use surrogate linear or quadratic objective functions that have no direct interpretation in terms of dose-volume histogram (DVH) criteria, do not result in an optimum or have long solution times. We decrease the solution time of the existing linear and quadratic dose-based programming models (LP and QP, respectively) to allow optimizing over potential catheter positions using mixed integer programming. An additional average speed-up of 75% can be obtained by stopping the solver at an early stage, without deterioration of the plan quality. For a fixed catheter configuration, the dwell time optimization model LP solves to optimality in less than 15 s, which confirms earlier results. We propose an iterative procedure for QP that allows us to prescribe the target dose as an interval, while retaining independence between the solution time and the number of dose calculation points. This iterative procedure is comparable in speed to the LP model and produces better plans than the non-iterative QP. We formulate a new dose-volume-based model that maximizes V(100%) while satisfying pre-set DVH criteria. This model optimizes both catheter positions and dwell times within a few minutes depending on prostate volume and number of catheters, optimizes dwell times within 35 s and gives better DVH statistics than dose-based models. The solutions suggest that the correlation between the objective value and the clinical plan quality is weak in the existing dose-based models.

Convex reformulation of biologically-based multi-criteria intensity-modulated radiation therapy optimization including fractionation effects.

Finding fluence maps for intensity-modulated radiation therapy (IMRT) can be formulated as a multi-criteria optimization problem for which Pareto optimal treatment plans exist. To account for the dose-per-fraction effect of fractionated IMRT, it is desirable to exploit radiobiological treatment plan evaluation criteria based on the linear-quadratic (LQ) cell survival model as a means to balance the radiation benefits and risks in terms of biologic response. Unfortunately, the LQ-model-based radiobiological criteria are nonconvex functions, which make the optimization problem hard to solve. We apply the framework proposed by Romeijn et al (2004 Phys. Med. Biol. 49 1991-2013) to find transformations of LQ-model-based radiobiological functions and establish conditions under which transformed functions result in equivalent convex criteria that do not change the set of Pareto optimal treatment plans. The functions analysed are: the LQ-Poisson-based model for tumour control probability (TCP) with and without inter-patient heterogeneity in radiation sensitivity, the LQ-Poisson-based relative seriality s-model for normal tissue complication probability (NTCP), the equivalent uniform dose (EUD) under the LQ-Poisson model and the fractionation-corrected Probit-based model for NTCP according to Lyman, Kutcher and Burman. These functions differ from those analysed before in that they cannot be decomposed into elementary EUD or generalized-EUD functions. In addition, we show that applying increasing and concave transformations to the convexified functions is beneficial for the piecewise approximation of the Pareto efficient frontier.

Derivative-free generation and interpolation of convex Pareto optimal IMRT plans.

In inverse treatment planning for intensity-modulated radiation therapy (IMRT), beamlet intensity levels in fluence maps of high-energy photon beams are optimized. Treatment plan evaluation criteria are used as objective functions to steer the optimization process. Fluence map optimization can be considered a multi-objective optimization problem, for which a set of Pareto optimal solutions exists: the Pareto efficient frontier (PEF). In this paper, a constrained optimization method is pursued to iteratively estimate the PEF up to some predefined error. We use the property that the PEF is convex for a convex optimization problem to construct piecewise-linear upper and lower bounds to approximate the PEF from a small initial set of Pareto optimal plans. A derivative-free Sandwich algorithm is presented in which these bounds are used with three strategies to determine the location of the next Pareto optimal solution such that the uncertainty in the estimated PEF is maximally reduced. We show that an intelligent initial solution for a new Pareto optimal plan can be obtained by interpolation of fluence maps from neighbouring Pareto optimal plans. The method has been applied to a simplified clinical test case using two convex objective functions to map the trade-off between tumour dose heterogeneity and critical organ sparing. All three strategies produce representative estimates of the PEF. The new algorithm is particularly suitable for dynamic generation of Pareto optimal plans in interactive treatment planning.