Neighborhood search approaches to beam orientation optimization in intensity modulated radiation therapy treatment planning

The intensity modulated radiation therapy (IMRT) treatment planning problem consists of several subproblems which are typically solved sequentially. We seek to combine two of the subproblems: the beam orientation optimization (BOO) problem and the fluence map optimization (FMO) problem. The BOO problem is the problem of selecting the beam orientations to deliver radiation to the patient. The FMO problem is the problem of determining the amount of radiation intensity, or fluence, of each beamlet in each beam. The solution to the FMO problem measures the quality of a beam set, but the majority of previous BOO studies rely on heuristics and approximations to gauge the quality of the beam set. In contrast with these studies, we use an exact measure of the treatment plan quality attainable using a given beam set, which ensures convergence to a global optimum in the case of our simulated annealing algorithm and a local optimum in the case of our local search algorithm. We have also developed a new neighborhood structure that allows for faster convergence using our simulated annealing and local search algorithms, thus reducing the amount of time required to obtain a good solution. Finally, we show empirically that we can generate clinically acceptable treatment plans that require fewer beams than in current practice. This may reduce the length of treatment time, which is an important clinical consideration in IMRT.     

Coupling three-field formulation and meshless mixed Galerkin methods using radial basis functions

In this work, we solve the elliptic partial differential equation by coupling the meshless mixed Galerkin approximation using radial basis function with the three-field domain decomposition method. The formulation has been adopted to increase the efficiency of the numerical technique by decreasing the error and dealing with the ill conditioning of the linear system caused by the radial basis function. Convergence analysis of the coupled technique is treated and numerical results of some solved examples are given at the end of this paper.     

Mixed methods for fourth-order elliptic and parabolic problems using radial basis functions

By extending Wendlands meshless Galerkin methods using RBFs, we develop mixed methods for solving fourth-order elliptic and parabolic problems by using RBFs. Similar error estimates as classical mixed finite element methods are proved. AMS subject classification 35G15, 65N12     

Numerical solutions of KdV equation using radial basis functions

Numerical solution of the Korteweg-de Vries equation is obtained by using the meshless method based on the collocation with radial basis functions. Five standard radial basis functions are used in the method of the collocation. The results are compared for the numerical experiments of the propagation of solitons, interaction of two solitary waves and breakdown of initial conditions into a train of solitons.     

Fast evaluation of radial basis functions: methods for two-dimensional polyharmonic splines

Received on 23 February 1995. Revised on 7 May 1996. This paper concerns the fast evaluation of radial basis functions.It describes the mathematics of a methos for splines of theform where p is a low-degree polynomial. Such functions are veryuseful for the interpolation of scattered data, but can be computationallyexpensive to use when N is large. The method described is ageneralization of the fast multipole method of Greengard andRokhlin for the potential case (m=0), and reduces the incrementalcost of a single extra evaluation from O(N) operations to O(1)operations. The paper develops the required series expansionsand uniqueness results. It pays particular attention to therate of convergence of the series approximations involved, obtainingimproved estimates which explain why numerical experiments revealfaster convergence than predicted by previous work for the potential(m=0) and thin-plate spline (m=1) cases.     

Parallel radial basis function methods for the global optimization of expensive functions

We introduce a master–worker framework for parallel global optimization of computationally expensive functions using response surface models. In particular, we parallelize two radial basis function (RBF) methods for global optimization, namely, the RBF method by Gutmann [Gutmann, H.M., 2001a. A radial basis function method for global optimization. Journal of Global Optimization 19(3), 201–227] (Gutmann-RBF) and the RBF method by Regis and Shoemaker [Regis, R.G., Shoemaker, C.A., 2005. Constrained global optimization of expensive black box functions using radial basis functions, Journal of Global Optimization 31, 153–171] (CORS-RBF). We modify these algorithms so that they can generate multiple points for simultaneous evaluation in parallel. We compare the performance of the two parallel RBF methods with a parallel multistart derivative-based algorithm, a parallel multistart derivative-free trust-region algorithm, and a parallel evolutionary algorithm on eleven test problems and on a 6-dimensional groundwater bioremediation application. The results indicate that the two parallel RBF algorithms are generally better than the other three alternatives on most of the test problems. Moreover, the two parallel RBF algorithms have comparable performances on the test problems considered. Finally, we report good speedups for both parallel RBF algorithms when using a small number of processors.     

On the formulation of credit barrier model using radial basis functions

Albanese et al in 2003 and Avellaneda and Zhu in 2001 develop the framework of credit barrier model. They provide special solutions to the model in case of simple stochastic structure. The technical aspect of the model remains wide open for general stochastic structure that is crucial when the model is required to calibrate with aggregate amount of empirical data. This paper provides a technical solution to this problem with the use of radial basis functions (RBF). This paper discusses the numerical implementation of the credit barrier model using the RBF method. It also demonstrates that the RBF method is numerically tractable in this problem and allows in the model richer stochastic structure capable of capturing relevant market information.     

Finding local optima of high-dimensional functions using direct search methods

This paper focuses on a subclass of box-constrained, non-linear optimization problems. We are particularly concerned with settings where gradient information is unreliable, or too costly to calculate, and the function evaluations themselves are very costly. This encourages the use of derivative free optimization methods, and especially a subclass of these referred to as direct search methods. The thrust of our investigation is twofold. First, we implement and evaluate a number of traditional direct search methods according to the premise that they should be suitable as local optimizers when used in a metaheuristic framework. Second, we introduce a new direct search method, based on Scatter Search, designed to remedy the lack of a good derivative free method for solving problems of high dimensions. Our new direct search method has convergence properties comparable to those of existing methods in addition to being able to solve larger problems more effectively.     

Some variants of the method of fundamental solutions: regularization using radial and nearly radial basis functions

The method of fundamental solutions and some versions applied to mixed boundary value problems are considered. Several strategies are outlined to avoid the problems due to the singularity of the fundamental solutions: the use of higher order fundamental solutions, and the use of nearly fundamental solutions and special fundamental solutions concentrated on lines instead of points. The errors of the approximations as well as the problem of ill-conditioned matrices are illustrated via numerical examples.     

Determination of a control parameter in a one-dimensional parabolic equation using the method of radial basis functions

In this work, the method of radial basis functions is used for finding the solution of an inverse problem with source control parameter. Because a much wider range of physical phenomena are modelled by nonclassical parabolic initial-boundary value problems, theoretical behavior and numerical approximation of these problems have been active areas of research. The radial basis functions (RBF) method is an efficient mesh free technique for the numerical solution of partial differential equations. The main advantage of numerical methods which use radial basis functions over traditional techniques is the meshless property of these methods. In a meshless method, a set of scattered nodes are used instead of meshing the domain of the problem. The results of numerical experiments are presented and some comparisons are made with several well-known finite difference schemes.     

Construction of a local and global Lyapunov function for discrete dynamical systems using radial basis functions

The basin of attraction of an asymptotically stable fixed point of the discrete dynamical system given by the iteration x_n+1=g(x_n) can be determined through sublevel sets of a Lyapunov function. In Giesl [On the determination of the basin of attraction of discrete dynamical systems. J. Difference Equ. Appl. 13(6) (2007) 523–546] a Lyapunov function is constructed by approximating the solution of a difference equation using radial basis functions. However, the resulting Lyapunov function is non-local, i.e. it has no negative discrete orbital derivative in a neighborhood of the fixed point. In this paper we modify the construction method by using the Taylor polynomial and thus obtain a Lyapunov function with negative discrete orbital derivative both locally and globally.     

Numerical solution of the nonlinear Klein–Gordon equation using radial basis functions

The nonlinear Klein–Gordon equation is used to model many nonlinear phenomena. In this paper, we propose a numerical scheme to solve the one-dimensional nonlinear Klein–Gordon equation with quadratic and cubic nonlinearity. Our scheme uses the collocation points and approximates the solution using Thin Plate Splines (TPS) radial basis functions (RBF). The implementation of the method is simple as finite difference methods. The results of numerical experiments are presented, and are compared with analytical solutions to confirm the good accuracy of the presented scheme.     

Closed form representations for a class of compactly supported radial basis functions

In this paper we re-examine Wendland’s strategy for the construction of compactly supported positive definite radial basis functions. We acknowledge that this strategy can be modified to capture a much larger range of functions, including the so-called missing Wendland functions which have been the subject of a recent paper by Schaback (Adv Comput Math 34:67–81, 2011). Our approach is to focus on a general integral representation of such functions and we will show how a careful evaluation of this integral leads to new closed form expressions for both Wendland’s original functions and the missing ones. The resulting expressions are easy to code and so provide the potential user with a quick way of accessing a desired example for a given application.     

Optimal control of a parabolic distributed parameter system via radial basis functions

This paper attempts to present a meshless method to find the optimal control of a parabolic distributed parameter system with a quadratic cost functional. The method is based on radial basis functions to approximate the solution of the optimal control problem using collocation method. In this regard, different applications of RBFs are used. To this end, the numerical solutions are obtained without any mesh generation into the domain of the problems. The proposed technique is easy to implement, efficient and yields accurate results. Numerical examples are included and a comparison is made with an existing result.     

Using eigenstructure of the Hessian to reduce the dimension of the intensity modulated radiation therapy optimization problem

Optimization is of vital importance when performing intensity modulated radiation therapy to treat cancer tumors. The optimization problem is typically large-scale with a nonlinear objective function and bounds on the variables, and we solve it using a quasi-Newton sequential quadratic programming method. This study investigates the effect on the optimal solution, and hence treatment outcome, when solving an approximate optimization problem of lower dimension. Through a spectral decompostion, eigenvectors and eigenvalues of an approximation to the Hessian are computed. An approximate optimization problem of reduced dimension is formulated by introducing eigenvector weights as optimization parameters, where only eigenvectors corresponding to large eigenvalues are included. The approach is evaluated on a clinical prostate case. Compared to bixel weight optimization, eigenvector weight optimization with few parameters results in faster initial decline in the objective function, but with inferior final solution. Another approach, which combines eigenvector weights and bixel weights as variables, gives lower final objective values than what bixel weight optimization does. However, this advantage comes at the expense of the pre-computational time for the spectral decomposition. A preliminary version of this paper was presented at the AAPM 46th annual meeting, held July 25–29, 2004 in Pittsburgh, PA.     

A coupled column generation,mixed integer approach to optimal planning of intensity modulated radiation therapy for cancer

Approximately 40% of all U.S. cancer cases are treated with radiation therapy. In Intensity-Modulated Radiation Therapy (IMRT) the treatment planning problem is to choose external beam angles and their corresponding intensity maps (showing how the intensity varies across a given beam) to maximize tumor dose subject to the tolerances of surrounding healthy tissues. Dose, like temperature, is a quantity defined at each point in the body, and the distribution of dose is determined by the choice of treatment parameters available to the planner. In addition to absolute dose limits in healthy tissues, some tissues have at least one dose-volume restriction that requires a fraction of its volume to not exceed a specified tighter threshold level for damage. There may also be a homogeneity limit for the tumor that restricts the allowed spread of dose across its volume. We formulate this planning problem as a mixed integer program over a coupled pair of column generation processes -- one designed to produce intensity maps, and a second specifying protected area choices for tissues under dose-volume restrictions. The combined procedure is shown to strike a balance between computing an approximately optimal solution and bounding its maximum possible suboptimality that we believe holds promise for implementations able to offer the power and flexibility of mixed-integer linear programming models on instances of practical scale.A portion of the work of Dr. Langer, Mr. Thai and Dr. Preciado-Walters was supported by National Science Foundation grant ECS-0120145 and National Cancer Institute 1R41CA91688-01. Dr. Rardin is participated while on rotation as Program Director for Operations Research and Service Enterprise Engineering at the National Science Foundation.     

On the convergence of a class of algorithms using linearly independent search directions

Powell has shown that the cyclic coordinate method with exact searches may not converge to a stationary point. In this note we consider a more general class of algorithms for unconstrained minimization, and establish their convergence under the assumption that the objective function has a unique minimum along any line.     

A meshless method for numerical solution of the one-dimensional wave equation with an integral condition using radial basis functions

The hyperbolic partial differential equation with an integral condition arises in many physical phenomena. In this paper, we propose a numerical scheme to solve the one-dimensional hyperbolic equation that combines classical and integral boundary conditions using collocation points and approximating the solution using radial basis functions (RBFs). The results of numerical experiments are presented, and are compared with analytical solution and finite difference method to confirm the validity and applicability of the presented scheme.     

Numerical solution of the system of second-order boundary value problems using the local radial basis functions based differential quadrature collocation method

In this research, we propose a numerical scheme to solve the system of second-order boundary value problems. In this way, we use the Local Radial Basis Function Differential Quadrature (LRBFDQ) method for approximating the derivative. The LRBFDQ method approximates the derivatives by Radial Basis Functions (RBFs) interpolation using a small set of nodes in the support domain of any node. So the new scheme needs much less computational work than the globally supported RBFs collocation method. We use two techniques presented by Bayona et al. (2011, 2012) [29], [30] to determine the optimal shape parameter. Some examples are presented to demonstrate the accuracy and easy implementation of the new technique. The results of numerical experiments are compared with the analytical solution, finite difference (FD) method and some published methods to confirm the accuracy and efficiency of the new scheme presented in this paper.