Skip to main content
Log in

An Approximation of the \(M_2\) Closure: Application to Radiotherapy Dose Simulation

  • Published:
Journal of Scientific Computing Aims and scope Submit manuscript

Abstract

Particle transport in radiation therapy can be modelled by a kinetic equation which must be solved numerically. Unfortunately, the numerical solution of such equations is generally too expensive for applications in medical centers. Moment methods provide a hierarchy of models used to reduce the numerical cost of these simulations while preserving basic properties of the solutions. Moment models require a closure because they have more unknowns than equations. The entropy-based closure is based on the physical description of the particle interactions and provides desirable properties. However, computing this closure is expensive. We propose an approximation of the closure for the first two models in the hierarchy, the \(M_1\) and \(M_2\) models valid in one, two or three dimensions of space. Compared to other approximate closures, our method works in multiple dimensions. We obtain the approximation by a careful study of the domain of realizability and by invariance properties of the entropy minimizer. The \(M_2\) model is shown to provide significantly better accuracy than the \(M_1\) model for the numerical simulation of a dose computation in radiotherapy. We propose a numerical solver using those approximated closures. Numerical experiments in dose computation test cases show that the new method is more efficient compared to numerical solution of the minimum entropy problem using standard software tools.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16

Similar content being viewed by others

References

  1. Alldredge, G.W., Hauck, C.D., O’Leary, D.P., Tits, A.L.: Adaptive change of basis in entropy-based moment closures for linear kinetic equations. J. Comput. Phys. 258, 489–508 (2014)

    Article  MathSciNet  MATH  Google Scholar 

  2. Alldredge, G.W., Hauck, C.D., Tits, A.L.: High-order entropy-based closures for linear transport in slab geometry II: a comutational study of the optimization problem. SIAM J. Sci. Comput. 34(4), 361–391 (2012)

    Article  MathSciNet  MATH  Google Scholar 

  3. Alldredge, G.W., Li, R., Li, W.: Approximating the \({M_2}\) method by the extended quadrature method of moments for radiative transfer in slab geometry. Kin. Rel. Mod. 9(2), 237–249 (2016)

  4. Anile, A.M., Pennisi, S.: Thermodynamic derivation of the hydrodynamical model for charge transport in semiconductors. Phys. Rev. B 46, 13186–13193 (1992)

    Article  MATH  Google Scholar 

  5. Berntsen, J., Espelid, T.O., Genz, A.: Algorithm 698: Dcuhre: an adaptive multidemensional integration routine for a vector of integrals. ACM Trans. Math. Softw. 17(4), 452–456 (1991). http://netlib.org/toms/

  6. Berthon, C., Charrier, P., Dubroca, B.: An HLLC scheme to solve the \({M_1}\) model of radiative transfer in two space dimensions. J. Sci. Comput. 31(3), 347–389 (2007)

    Article  MathSciNet  MATH  Google Scholar 

  7. Berthon, C., Frank, M., Sarazin, C., Turpault, R.: Numerical methods for balance laws with space dependent flux: application to radiotherapy dose calculation. Commun. Comput. Phys. 10(5), 1184–1210 (2011)

    Article  Google Scholar 

  8. Borwein, J., Lewis, A.: Duality relationships for entropy-like minimization problems. SIAM J. Control Optim. 29(2), 325–338 (1991)

    Article  MathSciNet  MATH  Google Scholar 

  9. Borwein, J., Lewis, A.: Partially finite convex programming, part I: quasi relative interiors and duality theory. Math. Program. 57, 15–48 (1992)

    Article  MATH  Google Scholar 

  10. Borwein, J., Lewis, A.: Partially finite convex programming: part II. Math. Program. 57, 49–83 (1992)

    Article  MATH  Google Scholar 

  11. Brunner, T.A., Holloway, J.P.: One-dimensional riemann solvers and the maximum entropy closure. J. Quant. Spectrosc. Radiat. Transf. 69(5), 543–566 (2001)

    Article  Google Scholar 

  12. Caron, J., Feugeas, J.-L., Dubroca, B., Kantor, G., Dejean, C., Birindelli, G., Pichard, T., Nicolaï, Ph, d’Humières, E., Frank, M., Tikhonchuk, V.: Deterministic model for the transport of energetic particles. Application in the electron radiotherapy. Phys. Med. 31, 912–921 (2015)

    Article  Google Scholar 

  13. Chandrasekhar, S.: Radiative Transfer. Dover, New York (1950)

    MATH  Google Scholar 

  14. Chetty, I.J., Curran, B., Cygler, J.E., DeMarco, J.J., Ezzell, G., Faddegon, B.A., Kawrakow, I., Keall, P.J., Liu, H., Ma, C.-M.C., Rogers, D.W.O., Seuntjens, J., Sheikh-Bagheri, D., Siebers, J.V.: Report of the AAPM task group no. 105: issues associated with clinical implementation of Monte Carlo-based photon and electron external beam treatment planning. Med. Phys. 34(12), 4818–4853 (2007)

    Article  Google Scholar 

  15. Curto, R., Fialkow, L.A.: Recusiveness, positivity, and truncated moment problems. Houst. J. Math. 17(4), 603–634 (1991)

    MATH  Google Scholar 

  16. Curto, R., Fialkow, L.A.: A duality prood to Tchakaloff’s theorem. J. Math. Anal. Appl. 269, 519–536 (2002)

    Article  MathSciNet  MATH  Google Scholar 

  17. Dubroca, B., Feugeas, J.L.: Hirarchie des modles aux moments pour le transfert radiatif. C.R. Acad. Sci. Paris 329, 915–920 (1999)

    Article  Google Scholar 

  18. Dubroca, B., Frank, M.: An iterative method for transport equations in radiotherapy. Prog. Ind. Math. ECMI 2008, 407–412 (2010)

    MATH  Google Scholar 

  19. Duclous, R., Dubroca, B., Frank, M.: A deterministic partial differential equation model for dose calculation in electron radiotherapy. Phys. Med. Biol. 55, 3843–3857 (2010)

    Article  Google Scholar 

  20. Friedriechs, K.O., Lax, P.D.: Systems of conservation equations with a convex extension. Proc. Nat. Acad. Sci. USA 68(8), 1686–1688 (1971)

    Article  MathSciNet  MATH  Google Scholar 

  21. Grad, H.: On the kinetic theory of rarefied gases. Commun. Pure Appl. Math. 2(4), 331–407 (1949)

    Article  MathSciNet  MATH  Google Scholar 

  22. Guisset, S., Brull, S., d’Humières, E., Dubroca, B., Karpov, S., Potapenko, I.: Asymptotic-preserving scheme for the M1-Maxwell system in the quasi-neutral regime. Commun. Comput. Phys. 19(2), 301–328 (2016)

  23. Guisset, S., Moreau, J.G., Nuter, R., Brull, S., d’Humieres, E., Dubroca, B., Tikhonchuk, V.T.: Limits of the \({M_1}\) and \({M_2}\) angular moments models for kinetic plasma physics studies. J. Phys. A Math. Theor. 48(33), 335501 (2015)

    Article  MathSciNet  MATH  Google Scholar 

  24. Harten, A., Lax, P., Van Leer, B.: On upstream differencing and Gudonov-type schemes for hyperbolic conservation laws. SIAM Rev. 25(1), 35–61 (1983)

    Article  MathSciNet  MATH  Google Scholar 

  25. Hauck, C., McClarren, R.: Positive \(P_N\) closures. SIAM J. Sci. Comput. 32(5), 2603–2626 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  26. Hauck, C.D.: High-order entropy-based closures for linear transport in slab geometry. Commun. Math. Sci 9(1), 187–205 (2011)

    Article  MathSciNet  MATH  Google Scholar 

  27. Hauck, C.D., Levermore, C.D., Tits, A.L.: Convex duality and entropy-based moment closures: characterizing degenerate densities. SIAM J. Control Optim. 47, 1977–2015 (2007)

    Article  MathSciNet  MATH  Google Scholar 

  28. Hensel, H., Iza-Teran, R., Siedow, N.: Deterministic model for dose calculation in photon radiotherapy. Phys. Med. Biol. 51, 675–693 (2006)

    Article  Google Scholar 

  29. Junk, M.: Maximum entropy for reduced moment problems. Math. Mod. Methods Appl. Sci. 10(1001–1028), 2000 (1998)

    MathSciNet  Google Scholar 

  30. Kershaw, D.: Flux Limiting Nature’s Own Way. Technical Report, Lawrence Livermore Laboratory (1976)

  31. Larsen, E.W., Miften, M.M., Fraass, B.A., Bruinvis, I.A.: Electron dose calculations using the method of moments. Med. Phys. 24(1), 111–125 (1997)

    Article  Google Scholar 

  32. La Rosa, S., Mascali, G., Romano, V.: Exact maximum entropy closure of the hydrodynamical model for Si semiconductors: The 8-moment case. SIAM J. Appl. Math. 70(3), 710–734 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  33. Levermore, C.D.: Relating Eddington factors to flux limiters. J. Quant. Spectrosc. Radiat. Transf. 31, 149–160 (1984)

    Article  Google Scholar 

  34. Levermore, C.D.: Moment closure hierarchies for kinetic theories. J. Stat. Phys. 83(5–6), 1021–1065 (1996)

    Article  MathSciNet  MATH  Google Scholar 

  35. Lewis, E.E., Miller, W.F.: Computational Methods of Neutron Transport. American Nuclear Society (1993)

  36. Mallet, J., Brull, S., Dubroca, B.: An entropic scheme for an angular moment model for the classical Fokker–Planck–Landau equation of electrons. Commun. Comput. Phys. 15(2), 422–450 (2014)

    Article  MathSciNet  Google Scholar 

  37. Mallet, J., Brull, S., Dubroca, B.: General moment system for plasma physics based on minimum entropy principle. Kin. Rel. Mod. 8(3), 533–558 (2015)

    Article  MathSciNet  MATH  Google Scholar 

  38. Maple™. Technical report, Maplesoft, a division of Waterloo Maple Inc. (2016)

  39. Mayles, P., Nahum, A., Rosenwald, J.C. (eds.): Handbook of Radiotherapy Physics: Theory and Practice. Taylor & Francis, London (2007)

    Google Scholar 

  40. McClarren, R.G., Hauck, C.D.: Robust and accurate filtered spherical harmonics expansion for radiative transfer. J. Comput. Phys. 229, 5597–5614 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  41. McDonald, J., Torrilhon, M.: Affordable robust moment closures for cfd based on the maximum-entropy hierarchy. J. Comput. Phys. 251, 500–523 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  42. Mead, L.R., Papanicolaou, N.: Maximum entropy in the problem of moments. J. Math. Phys. 25(8), 2404–2417 (1984)

    Article  MathSciNet  Google Scholar 

  43. Moré, J.J., Garbow, B.S., Hillstrom, K.E.: User Guide for MINPACK-1 (1980). http://www.netlib.org/minpack/

  44. Olbrant, E., Frank, M.: Generalized Fokker-Planck theory for electron and photon transport in biological tissues: application to radiotherapy. Comput. Math. Methods Med. 11(4), 313–339 (2010)

    Article  MathSciNet  MATH  Google Scholar 

  45. Pichard, T., Alldredge, G.W., Brull, S., Dubroca, B., Frank, M.: The \({M_2}\) model for dose simulation in radiation therapy. J. Comput. Theor. Transport. 45(3), 174–183 (2016)

  46. Pichard, T., Aregba-Driollet, D., Brull, S., Dubroca, B., Frank, M.: Relaxation schemes for the \({M_1}\) model with space-dependent flux: application to radiotherapy dose calculation. Commun. Comput. Phys. (2014, to appear)

  47. Piessens, R., De Doncker-Kapenga, E., Überhuber, C.W.: QUADPACK: A Subroutine Package for Automatic Integration. Springer edition (1983). http://www.netlib.org/quadpack/

  48. Pomraning, G.C.: The Fokker–Planck operator as an asymptotic limit. Math. Mod. Methods Appl. Sci. 2(1), 21–36 (1991)

    Article  MathSciNet  MATH  Google Scholar 

  49. Schneider, F.: Kershaw closures for linear transport equations in slab geometry I: model derivation. J. Comput. Phys. 322(C), 905–919 (2016)

  50. Schneider, J.: Entropic approximation in kinetic theory. ESAIM Math. Model. Numer. 38(3), 541–561 (2004)

    Article  MathSciNet  MATH  Google Scholar 

  51. Sempau, J., Salvat, F., Fernández-Varea, J. M.: PENELOPE-2011: A Code System for Monte Carlo Simulation of Electron and Photon Transport (2011)

  52. Spezi, E., Lewis, G.: An overview of Monte Carlo treatment planning for radiotherapy. Radiat. Prot. Dos. 131(1), 123–129 (2008)

    Article  Google Scholar 

  53. Vikas, V., Hauck, C.D., Wang, Z.J., Fox, R.O.: Radiation transport modeling using extended quadrature method of moments. J. Comput. Phys. 246, 221–241 (2013)

    Article  MathSciNet  MATH  Google Scholar 

  54. Wareing, T.A., McGhee, J.M., Archambault, Y., Thompson, S.: Acuros XB advanced dose calculation for the Eclipse™ treatment planning system. Clinical Perspectives (2010)

  55. Zankowski, C., Laitinen, M., Neuenschwander, H.: Fast electron Monte Carlo for eclipse™. Technical Report, Varian Medical System

Download references

Acknowledgments

The authors are grateful to Gabriele Birindelli (Université de Bordeaux) and Kerstin Küpper (RWTH Aachen) for performing the PENELOPE simulations, and to Nuria Escobar Corral (Uniklinikum RWTH Aachen) for providing the density map in the chest for the last test case. T. Pichard’s Ph.D. was funded by IdEx Bordeaux and Aquitaine Region. This work was also partially funded by Aquitaine Region and FEDER fund through IOPRA interface.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to T. Pichard.

Appendix: Computation of Moments in \(H^i\)

Appendix: Computation of Moments in \(H^i\)

1.1 Computation of Moments in \(H^1\)

Consider \(\bar{\lambda }\in \mathscr {L}_1\), and the associated ansatz \(\psi _1\) (defined in (21a)) and its moments \((\psi ^1,\psi ^2)\) given by

$$\begin{aligned} \psi _1(\varOmega )= & {} \exp \left( \lambda _1\varOmega _1+\lambda _4\varOmega _1^2+\lambda _5\varOmega _2^2+\lambda _6\varOmega _3^2\right) , \\ (\psi ^1, \psi ^2)= & {} \left\langle (\varOmega ,\ \varOmega \otimes \varOmega ) \psi _1 \right\rangle \in H^1. \end{aligned}$$

One can remark that \(\psi _1\) is an even function of \(\varOmega _2\) or \(\varOmega _3\), and therefore the moment of \(\psi _1\) according to any odd polynomial is zero, in particular

$$\begin{aligned} \psi _2^1 = \psi _3^1 = \psi _{1,2}^2 = \psi _{1,3}^2 = \psi _{2,3}^2 = 0. \end{aligned}$$

With those computations, the moments \(\psi ^1\) and \(\psi ^2\) actually reads

$$\begin{aligned} \psi ^1 = \psi _1^1 e_1, \quad \psi ^2 = Diag\left( \psi _{1,1}^2,\psi _{2,2}^2,\psi _{3,3}^2\right) , \end{aligned}$$

and \(\psi ^1\) is therefore an eigenvector of \(\psi ^2\). Using Notations 1 leads to write \(N^1\) and \(N^2\) under the form (23a), and one may observe that the eigenvectors of \(N^2-N^1\otimes N^1\) are along the cartesian axis \(e_i\).

Using again evenness of \(\psi _1\), one obtains

$$\begin{aligned} \psi _{1,1,2}^2 = \psi _{1,1,3}^2 =\psi _{1,2,3}^2 =\psi _{2,2,2}^2 =\psi _{2,2,3}^2 =\psi _{2,2,3}^2 =\psi _{3,3,3}^2 = 0. \end{aligned}$$

Using the fact that \(tr(\varOmega \otimes \varOmega ) = 1\), one obtains that

$$\begin{aligned} \sum \limits _{j=1}^3 \psi _{i,j,j}^3 = \int _{S^2} \varOmega _i tr(\varOmega \otimes \varOmega ) \psi _1(\varOmega ) d\varOmega = \psi _{i}^1. \end{aligned}$$

This leads to write \(N^3\) under the form (24a).

Proposition 4

Consider realizable moments \((\psi ^0,\psi ^1,\psi ^2)\in \mathscr {R}_2\) such that \(\psi ^1\) is an eigenvector of \(\psi ^2\).

Then the rotated normalized moments \((N^1,N^2)\) given by (17) are in \(H^1\).

Proof

Under those hypothesis, the decomposition (17) can be simplified. Indeed, since \(\psi ^1\) is an eigenvector of \(\psi ^2\), then \(N^1\) is an eigenvector of \(N^2\). So a rotation R diagonalizing \(N^2\) will send \(N^1\) onto one of the cartesian axis (chose R such that \(N^1\) is along \(e_1\)).

Then this rotation also diagonalizes \(N^2-N^1\otimes N^1\) since it diagonalizes both \(N^2\) and \(N^1\otimes N^1\) and one can write \(N^1\) and \(N^2\) under the form (21a).

Finally one can prove that the unique exponential representation for moments \((N^1,N^2)\) satisfying (23a) is (21a) by using Theorem 1 with \(\bar{m}(\varOmega ) = (\varOmega _1, \ \varOmega _1^2, \ \varOmega _2^2,\ \varOmega _3^2).\) Indeed this theorem provide the existence of a unique function \(\psi \) of the form (21a) satisfying

$$\begin{aligned} (N_1^1,N_{1,1}^2,N_{2,2}^2,N_{3,3}^2) = \left\langle \bar{m}\psi \right\rangle . \end{aligned}$$

Computing the other moments of such a function (21a) read

$$\begin{aligned} \psi _2^1 = \psi _3^1 = \psi _{1,2}^2 = \psi _{1,3}^2 = \psi _{2,3}^2 = 0, \end{aligned}$$

i.e. it satisfies the other moment constraints. Then the unique function (10) satisfying all the moment constraints has the form (21a).\(\square \)

1.2 Computation of Moments in \(H^2\)

Consider \(\bar{\lambda }\in \mathscr {L}_2\), and the associated ansatz \(\psi _2\) (defined in (21b)) and its moments \((\psi ^1,\psi ^2)\) given by

$$\begin{aligned} \psi _2(\varOmega )= & {} \exp \left( \lambda _5+\lambda _1\varOmega _1+(\lambda _4-\lambda _5)\varOmega _1^2\right) , \\ (\psi ^1, \psi ^2)= & {} \left\langle (\varOmega ,\ \varOmega \otimes \varOmega ) \psi _2 \right\rangle \in H^2. \end{aligned}$$

The computations of the previous subsection hold. Since \(\psi _2\) does not depend on \(\varOmega _2\) nor \(\varOmega _3\), one deduces that

$$\begin{aligned} \psi _{2,2}^2= & {} \int _{S^2} \varOmega _2^2 \exp \left( \lambda _5+\lambda _1\varOmega _1+(\lambda _4-\lambda _5)\varOmega _1^2\right) d\varOmega = \frac{\psi ^0}{2},\\ \psi _{3,3}^2= & {} \int _{S^2} \varOmega _3^2 \exp \left( \lambda _5+\lambda _1\varOmega _1+(\lambda _4-\lambda _5)\varOmega _1^2\right) d\varOmega = \frac{\psi ^0}{2}, \end{aligned}$$

in particular, \(\psi _{2,2}^2 = \psi _{3,3}^2\). Using Notations 1 leads to write \(N^1\) and \(N^2\) under the form (23b).

Similarily, one has

$$\begin{aligned} \psi _{1,2,2}^2= & {} \int _{S^2} \varOmega _1 \varOmega _2^2 \exp \left( \lambda _4+\lambda _1\varOmega _1\right) d\varOmega = \frac{\psi _1^1}{2},\\ \psi _{1,3,3}^2= & {} \int _{S^2} \varOmega _1 \varOmega _3^2 \exp \left( \lambda _4+\lambda _1\varOmega _1\right) d\varOmega = \frac{\psi _1^1}{2}, \end{aligned}$$

This leads to write \(N^3\) under the form (24b).

Proposition 5

Consider realizable moments \((\psi ^0,\psi ^1,\psi ^2)\in \mathscr {R}_2\) such that \(\psi ^1\) is an eigenvector of \(\psi ^2\) and \(\psi _{2,2}^2=\psi _{3,3}^2\).

Then the rotated normalized moments \((N^1,N^2)\) given by (17) are in \(H^2\).

The proof is identical to the one of Proposition 4 with \(\bar{m}(\varOmega ) = (1,\varOmega _1,\varOmega _1^2)\).

1.3 Computation of Moments in \(H^3\)

Consider \(\bar{\lambda }\in \mathscr {L}_3\), and the associated ansatz \(\psi _3\) (defined in (21c)) and its moments \((\psi ^1,\psi ^2)\) given by

$$\begin{aligned} \psi _3(\varOmega )= & {} \exp \left( \lambda _4+\lambda _1\varOmega _1\right) , \\ (\psi ^1, \psi ^2)= & {} \left\langle (\varOmega ,\ \varOmega \otimes \varOmega ) \psi _3 \right\rangle \in H^2. \end{aligned}$$

The computations of the previous subsections hold. In this case, the ansatz \(\psi _3\) is the \(M_1\) ansatz defined in (13), and therefore \(\psi _{1,1}^2\) is the Eddington factor \(\chi _2\) defined in Sect. 2.3. Using Notations 1 leads to write \(N^1\) and \(N^2\) under the form (23c).

The form of \(N^3\) is not simplified compared to the previous case.

Proposition 6

Consider realizable moments \((\psi ^0,\psi ^1,\psi ^2)\in \mathscr {R}_2\) such that \(\psi ^1\) is an eigenvector of \(\psi ^2\), \(\psi _{2,2}^2=\psi _{3,3}^2\) and \(\psi _{1,1}^2 = \psi ^0 \chi _2\left( |\psi ^1|/\psi ^0\right) \).

Then the rotated normalized moments \((N^1,N^2)\) given by (17) are in \(H^2\).

The proof is identical to the one of Proposition 4 with \(\bar{m}(\varOmega ) = (1,\ \varOmega _1)\).

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Pichard, T., Alldredge, G.W., Brull, S. et al. An Approximation of the \(M_2\) Closure: Application to Radiotherapy Dose Simulation. J Sci Comput 71, 71–108 (2017). https://doi.org/10.1007/s10915-016-0292-8

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10915-016-0292-8

Keywords

Mathematics Subject Classification

Navigation