System and methods for accelerating simulation of radiation treatment

• US 9,044,601 B2
• Filed: 10/19/2009
• Issued: 06/02/2015
• Est. Priority Date: 10/20/2008
• Status: Active Grant

• Current Assignee: Fred J. Currell, Marcus Mendenhall
• Original Assignee: Fred J. Currell, Marcus Mendenhall
• Inventors: Mendenhall, Marcus, Currell, Fred
• Primary Examiner(s): Silver, David
• Assistant Examiner(s): Hann, Jay B
• Application Number:
  US13/125,132
• Publication Number:
  US 20110202324A1
• Time in Patent Office: 2,052 Days
• Field of Search: 703/2, 703/6, 703/11, 703/12, 600/436, 378/64, 378/65, 378/70, 378/82
• US Class Current: 1/1
• CPC Class Codes: A61B 6/03 Computed tomography [CT] echo-tomography A61B8/14
• CPC Class Codes: A61N 2005/1034 Monte Carlo type methods; particle tracking
• CPC Class Codes: A61N 5/1031 using a specific method of dose optimization
• CPC Class Codes: G06F 30/20 Design optimisation, verification or simulation optimisation, verification or simulation of circuit designs G06F30/30

Abstract

A method for estimating a radiation dose and distribution for a target volume is provided. The method includes receiving a simulated dose array (606) describing a plurality of simulated dose values for a plurality of voxels in the target volume, generating an energy deposition coefficient function for the plurality of voxels (608), and obtaining a raw fluence array based at least on the simulated dose array and the energy deposition coefficient function (612). The method also includes generating an adjusted fluence array (613) based on the raw fluence array and at least one adjustment criteria, and generating an adjusted dose array (622) for the target volume based on the adjusted fluence array and the energy deposition coefficient function.

Claims

1. A method for estimating a radiation dose and distribution for a target volume, the method comprising:

receiving a simulated dose array, the dose array describing a plurality of simulated dose values for a plurality of voxels in the target volume;

generating a non-uniform energy deposition coefficient function for the plurality of voxels;

obtaining a raw fluence array based at least on the simulated dose array and the energy deposition coefficient function;

generating an adjusted fluence array based on the raw fluence array and at least one adjustment function; and

generating an adjusted dose array for the target volume based on the adjusted fluence array and the energy deposition coefficient function;

wherein the adjusted dose array is of the form

D( r)=μ

( r)[ . . . ],

where μ

( r) is the energy deposition coefficient function which is defined by a set of coefficients which describe mean energy deposition per fluence at a specified voxel, D( r) is the adjusted dose array at a specified voxel, and r is the specified voxel.

2. The method of claim 1, further comprising:

receiving a composition array describing one or more compositions associated with each of the plurality of voxels in the target volume;

wherein the generating the energy deposition coefficient function further comprises computing energy deposition coefficient values for each of the compositions in the target volume.

3. The method of claim 2, wherein:

the computing further comprises deriving the energy coefficient values such that differences in fluence values are minimized for the plurality of voxels along a border between a first of the compositions and a second of the compositions.

4. The method of claim 1, wherein:

the adjustment function comprises a smoothing function that is a weighted convolution with a kernel having an anisotropy, wherein the anisotropy is based on the direction of the flux of particles used for generating the simulated dose array.

5. The method of claim 4, wherein:

the raw fluence array has a plurality of dimensions, and the adjustment function specifies a kernel width for each of the plurality of dimensions; and

the applying the kernel function further comprises;

extending the raw fluence array in each of the dimensions by at least 1½

times the kernel width for each of the plurality of dimensions to construct a padded fluence array;

constructing a padded kernel array based on the kernel function and the dimension of the padded fluence array; and

multiplying the padded fluence array by the padded kernel array to produce values for adjusted kernel array.

6. The method of claim 5, wherein:

the multiplying comprises a fast Fourier transform multiplication of the padded fluence array and the padded kernel array.

7. The method of claim 1, wherein:

for a given voxel, the adjusted dose array is generated subsequent to the generation of the energy deposition coefficient function and the generation of adjusted fluence array, and the energy deposition coefficient function for the given voxel and the adjusted fluence array for the given voxel are used to calculate the adjusted dose array for the given voxel.

8. The method of claim 1, wherein:

the adjusted dose array is based on a product of the adjusted fluence array and the energy deposition coefficient function.

9. The method of claim 1, wherein:

the at least one adjustment function is a smoothing convolution kernel.

10. The method of claim 9, wherein:

the smoothing convolution kernel is Gaussian and is applied to voxels for which the smoothing convolution kernel is non-zero.

11. The method of claim 1, wherein:

the at least one adjustment function is an anisotropic kernel.

12. The method of claim 1, wherein:

the at least one adjustment function is an inhomogeneous kernel.

13. The method of claim 1, wherein:

the at least one adjustment function comprises a number of kernels of different sizes.

14. A radiation treatment system, comprising:

a computer processing system including at least one processor and associated memory;

wherein the memory is configured to store information that represents a simulated dose array, the simulated dose array describing a plurality of simulated dose values for a plurality of voxels in a target volume comprising one or more different compositions; and

wherein the processor is configured to;

generate a non-uniform energy deposition coefficient function for the plurality of voxels as defined by the information stored in the memory;

generate a raw fluence array based at least on the simulated dose array as defined by the information stored in the memory and the energy deposition coefficient function;

generate an adjusted fluence array based on the raw fluence array and at least one adjustment function; and

generate an adjusted dose array for the target volume based on the adjusted fluence array and the energy deposition coefficient function;

wherein the adjusted dose array is of the form

D( r)=μ

( r)[ . . . ],

where μ

( r) is the energy deposition coefficient function which is defined by a set of coefficients which describe mean energy deposition per fluence at a specified voxel,

D(r) is the adjusted dose array at a specified voxel, and r is the specified voxel.

15. The system of claim 14, wherein:

the information stored in the memory further includes a composition array describing one or more compositions associated with each of the plurality of voxels in the target volume; and

the processor is further configured during the generating of the energy deposition coefficient function to compute energy deposition coefficient values for each of the compositions in the target volume.

16. The system of claim 15, wherein:

the energy coefficient values are computed such that differences in fluence values are minimized for the plurality of voxels along a border between a first of the compositions and a second of the compositions.

17. The system of claim 14, wherein:

the adjustment function comprises a smoothing function that is a weighted convolution with a kernel having an anisotropy, wherein the anisotropy is based on the direction of the flux of particles used for generating the simulated dose array.

18. The system of claim 17, wherein:

the raw fluence array has a plurality of dimensions, and the adjustment function specifies a kernel width for each of the plurality of dimensions; and

wherein the applying the kernel function further comprises;

extending the raw fluence array in each of the dimensions by at least 1½

times the kernel width for each of the plurality of dimensions to generate a padded fluence array;

constructing a padded kernel array based on the kernel function the dimension of the padded fluence array; and

multiplying the padded fluence array by the padded kernel array to produce values for adjusted kernel array.

19. The system of claim 18, wherein:

the multiplying comprises a fast Fourier transform multiplication of the padded fluence array and the padded kernel array.

20. A method for estimating a radiation dose and distribution for a target volume, comprising:

receiving a simulated dose array, the dose array describing a plurality of simulated dose values for a plurality of voxels in the target volume;

generating a non-uniform energy deposition coefficient function for the plurality of voxels;

obtaining a raw fluence array based at least on the simulated dose array and the energy deposition coefficient function;

generating an adjusted fluence array based on the raw fluence array and at least one adjustment function; and

generating an adjusted dose array for the target volume based on the adjusted fluence array and the energy deposition coefficient function;

wherein, for a given voxel, the adjusted dose array is generated subsequent to the generation of the energy deposition coefficient function and the generation of adjusted fluence array, and the energy deposition coefficient function for the given voxel and the adjusted fluence array for the given voxel are used to calculate the adjusted dose array for the given voxel;

wherein, the adjusted dose array for the given voxel is based on a product of the adjusted fluence array for the given voxel and the energy deposition coefficient function for the given voxel.

21. The method of claim 20, wherein:

the energy deposition coefficient function is defined by a set of coefficients that describe mean energy deposition per fluence at a specified voxel.