Strong valid inequalities for fluence map optimization problem under dose-volume restrictions

Tuncel, Ali T.; Preciado, Felisa; Rardin, Ronald L.; Langer, Márk; Richard, Jean‐Philippe P.

doi:10.1007/s10479-010-0759-1

Cited by 13 publications

(18 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This fluctuation limits the applicability of automated planning methods which rely heavily on ROIs (including optimization ROIs) and require highly standardized ROI labelling. There are three related challenges: i) RT planning is a multicriteria optimization problem with many possible solutions achieving different outcomes some of which are clinically acceptable (a Pareto front [7]), while most are not; ii) DVOs are sums over the dose to a given ROI and thus don't capture spatial information where it is desired to have the dose shaped a certain way within the region (example below) as a result meeting the DVOs may not yield an acceptable plan and the DVOs get tweaked or new ones iteratively added by the user; and iii) the optimization problem is non-convex, NP-hard [8] and sensitive to initialization, and as a result DVOs and new optimization ROIs are routinely added to "guide" the solution toward a better optima. Optimization ROIs are created adhoc and are therefore highly specific for a particular patient.…”

Section: A Problem Description and Hypothesismentioning

confidence: 99%

Contextual Atlas Regression Forests: Multiple-Atlas-Based Automated Dose Prediction in Radiation Therapy

McIntosh

Purdie

2016

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

Radiation therapy is an integral part of cancer treatment, but to date it remains highly manual. Plans are created through optimization of dose volume objectives that specify intent to minimize, maximize, or achieve a prescribed dose level to clinical targets and organs. Optimization is NP-hard, requiring highly iterative and manual initialization procedures. We present a proof-of-concept for a method to automatically infer the radiation dose directly from the patient's treatment planning image based on a database of previous patients with corresponding clinical treatment plans. Our method uses regression forests augmented with density estimation over the most informative features to learn an automatic atlas-selection metric that is tailored to dose prediction. We validate our approach on 276 patients from 3 clinical treatment plan sites (whole breast, breast cavity, and prostate), with an overall dose prediction accuracies of 78.68%, 64.76%, 86.83% under the Gamma metric.

show abstract

Section: A Problem Description and Hypothesismentioning

confidence: 99%

Contextual Atlas Regression Forests: Multiple-Atlas-Based Automated Dose Prediction in Radiation Therapy

McIntosh

Purdie

2016

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

show abstract

“…Quadratic-programming is common as the minimization of the deviation from the prescribed target dose [33][34][35][36]. Linear programming (LP) was applied to radiotherapy planning by Bahr et al as early as 1968 [37], and it has remained popular [38][39][40][41][42][43][44]. The advantage of linear programming is that LP solvers have predictable average case performance compared to the nonlinear formulations.…”

Section: Optimization Methodsmentioning

confidence: 99%

Intensity modulated radiation therapy with field rotation—a time-varying fractionation study

Dink

Langer

Rardin

et al. 2012

Health Care Manag Sci

Self Cite

View full text Add to dashboard Cite

This paper proposes a novel mathematical approach to the beam selection problem in intensity modulated radiation therapy (IMRT) planning. The approach allows more beams to be used over the course of therapy while limiting the number of beams required in any one session. In the proposed field rotation method, several sets of beams are interchanged throughout the treatment to allow a wider selection of beam angles than would be possible with fixed beam orientations. The choice of beamlet intensities and the number of identical fractions for each set are determined by a mixed integer linear program that controls jointly for the distribution per fraction and the cumulative dose distribution delivered to targets and critical structures. Trials showed the method allowed substantial increases in the dose objective and/or sparing of normal tissues while maintaining cumulative and fraction size limits. Trials for a head and neck site showed gains of 25%-35% in the objective (average tumor dose) and for a thoracic site gains were 7%-13%, depending on how strict the fraction size limits were set. The objective did not rise for a prostate site significantly, but the tolerance limits on normal tissues could be strengthened with the use of multiple beam sets.

show abstract

“…Integer variables are needed because a DVC limits the dose applied for a certain number of voxels. Determining exactly how many of the voxels should meet DVCs is a difficult combinatorial problem that has multiple local optima and is nonconvex and nondeterministic polynomial time hard [4,5]. …”

Section: Introductionmentioning

confidence: 99%

“…However, MIP models are too difficult to solve to be practically useful because of their nonconvex and nondeterministic polynomial time hard characteristics. Tuncel, Preciado, Rardin, Langer, and Richard [5] introduced a family of disjunctive valid inequalities to the MIP formulation of the FMO problem under DVCs. A heuristic algorithm based on the geometric distance sorting technique is proposed by Lan, Li, Ren, Zhang, and Min [21] for solving a Linear constrained, quadratic objective MIP formulation of the FMO with DVCs.…”

Section: Introductionmentioning

confidence: 99%

An Automatic Approach for Satisfying Dose-Volume Constraints in Linear Fluence Map Optimization for IMPT

Zaghian¹,

Lim²,

Liu³

et al. 2014

JCT

View full text Add to dashboard Cite

Prescriptions for radiation therapy are given in terms of dose-volume constraints (DVCs). Solving the fluence map optimization (FMO) problem while satisfying DVCs often requires a tedious trial-and-error for selecting appropriate dose control parameters on various organs. In this paper, we propose an iterative approach to satisfy DVCs using a multi-objective linear programming (LP) model for solving beamlet intensities. This algorithm, starting from arbitrary initial parameter values, gradually updates the values through an iterative solution process toward optimal solution. This method finds appropriate parameter values through the trade-off between OAR sparing and target coverage to improve the solution. We compared the plan quality and the satisfaction of the DVCs by the proposed algorithm with two nonlinear approaches: a nonlinear FMO model solved by using the L-BFGS algorithm and another approach solved by a commercial treatment planning system (Eclipse 8.9). We retrospectively selected from our institutional database five patients with lung cancer and one patient with prostate cancer for this study. Numerical results show that our approach successfully improved target coverage to meet the DVCs, while trying to keep corresponding OAR DVCs satisfied. The LBFGS algorithm for solving the nonlinear FMO model successfully satisfied the DVCs in three out of five test cases. However, there is no recourse in the nonlinear FMO model for correcting unsatisfied DVCs other than manually changing some parameter values through trial and error to derive a solution that more closely meets the DVC requirements. The LP-based heuristic algorithm outperformed the current treatment planning system in terms of DVC satisfaction. A major strength of the LP-based heuristic approach is that it is not sensitive to the starting condition.

show abstract

Strong valid inequalities for fluence map optimization problem under dose-volume restrictions

Cited by 13 publications

References 24 publications

Contextual Atlas Regression Forests: Multiple-Atlas-Based Automated Dose Prediction in Radiation Therapy

Contextual Atlas Regression Forests: Multiple-Atlas-Based Automated Dose Prediction in Radiation Therapy

Intensity modulated radiation therapy with field rotation—a time-varying fractionation study

An Automatic Approach for Satisfying Dose-Volume Constraints in Linear Fluence Map Optimization for IMPT

Contact Info

Product

Resources

About