X-ray or gamma ray systems or devices – Specific application – Absorption
Reexamination Certificate
2000-11-13
2002-06-25
Bruce, David V. (Department: 2882)
X-ray or gamma ray systems or devices
Specific application
Absorption
C378S064000
Reexamination Certificate
active
06411675
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to radiation therapy planning for the treatment of tumors or for stereotactic radiosurgery and more particularly to the optimization of the radiation dose delivered to a patient by inverse treatment planning.
2. Description of the Background Art
The delivery of radiation for the treatment of tumors or obliteration of other abnormalities in human patients has undergone important changes in the last few years. The most important change has been the use of fine pencil beams of radiation of fixed intensity that can be scanned in a manner similar to a raster, remaining in each position of the scan for different lengths of time. With groups of pencil beams directed to the tumor or abnormality from different angles around the body, it is possible to deliver a radiation dose distribution that is effective in the treatment of the tumor or abnormality while sparing organs or tissues that are in the path of the radiation beams. That technology is known as Fluence Modulated Radiation Treatment (FMRT) or Fluence Modulated Radiation Surgery (FMRS). For historical reasons, it is also commonly known as Intensity Modulated Radiation Treatment/Surgery (IMRT/S) although the latter terminology is misleading since it implies modulated beam intensities.
The problem of therapy planning for FMRT/S is a difficult one: the fluences of a large number of pencil beams from many angles have to be calculated so that the sum of all their effects results in a uniform dose to the volume that contains the tumor or abnormality. Excessive dose to some portions of that volume can result in medical complications and under-dosing a region can lead to poor tumor control or recurrence. At the same time, the calculated fluences must fulfill some restrictive requirements to the radiation dose that is allowable in the different sensitive tissues traversed by the beams. The requirements placed by a therapist may be medically desirable but, to a smaller or larger extent, will be inconsistent with the physical laws that govern the absorption of radiation by the different tissues in the patient. The calculation of beam fluences, then, requires a methodology that allows for those inconsistencies and still delivers a set of beam flux results that achieve the most important of the desired results, as defined by the therapist. The process of obtaining that set of results is usually termed “optimization”. The volume that includes the tumor or other abnormalities to be treated is normally termed “Planning Treatment Volume”, or PTV. There can be more than one PTV in a treatment plan, but, without loss of generality, this document will refer to a PTV in the singular. Similarly, the volume that includes sensitive tissues or organs is normally termed “Organ at Risk”, or OAR. There can be more than one OAR, but the term will be used in the singular, without loss of generality, except when giving specific examples with multiple OAR's.
Optimization is invariably carried out by maximizing or minimizing a target or cost function incorporating the requirements and limitations placed by the therapist on the solution of the problem. Once a target or cost function has been defined, an algorithm is called upon to find the values of the beam fluence variables that will maximize or minimize that function, as the case may require.
Current art in optimization for FMRT/S can be divided into technologies that use:
a) a mathematically derived iterative formula to maximize or minimize the target or cost function. Those methods need an analytic target or cost function, so that it has explicit first partial derivatives with respect to the beam fluence variables.
b) a stochastic method that repeatedly tries maximizing or minimizing a target or cost function by testing whether a small change in the fluence of a randomly selected beam results in a change of that function in the desired direction.
Current art for the two types of technologies can be summarized as:
a1) using a Dynamically Penalized Likelihood target function which maximizes the likelihood that the resulting beam fluence values will yield the desired dose to the PTV, subject to a maximum dose specified for the OAR (see U.S. Pat. No. 5,602,892, J. Llacer, “Method for Optimization of Radiation Therapy Planning”, Feb. 11, 1997; and J. Llacer, “Inverse radiation treatment planning using the Dynamically Penalized Likelihood method”, Med. Phys., 24, (11) pp 1751-1764, 1997).
a2) using a quadratic cost function which is to be minimized, with additional quadratic penalty functions imposed to apply restrictions. A gradient method speeded up by scaling the gradient with the inverse of the diagonal elements of the Hessian matrix is used for that minimization. It allows for specifying the maximum dose to the OAR and also the fraction of the OAR volume that is allowed to receive more than a certain dose (see T. Bortfeld, J. Bürkelbach, R. Boesecke and W. Schlegel, “Methods of image reconstruction from projections applied to conformation radiotherapy”, Pys. Med. Biol., 1990, Vol 35, No. 10, 1423-1434; T. Bortfeld, J. Bürkelbach and W. Schlegel, “Three-dimensional solution to the inverse problem in conformation radiotherapy”, Advanced Radiation Therapy Tumor Response Monitoring and Treatment Planning, Breit Ed., Springer, pp 503-508, 1992; T. Bortfeld, W. Schlegel and B. Rhein, “Decomposition of pencil beam kernels for fast dose calculations in three-dimensional treatment planning”, Med. Phys. 20 (2), Pt. 1, pp 311-318, 1993; and T. Bortfeld, J. Stein and K. Preiser, “Clinically relevant intensity modulation optimization using physical criteria”, Proceedings of the XII
th
International Conference on the Use of Computers in Radiation Therapy (ICCR), Leavitt & Starkschall, eds., Springer, pp 1-4, 1997).
a3) using a quadratic cost function with constraints that limit the space of the allowable solutions to those that are non-negative (there cannot be negative beam fluences) and lead to doses to the OAR that are not above a maximum and/or result in no more than a specified fraction of the OAR volume receiving more than a certain dose. A modified form of the Conjugate Gradient method is used for minimization of the cost function (see S. V. Spirou and C-S. Chui, “A gradient inverse planning algorithm with dose-volume constraints”, Med. Phys. 25 (3), pp 321-333, 1998).
b1) using the Simulated Annealing technique, a stochastic method, to minimize a variety of proposed cost functions, including some possibly important biological functions like Tumor Control Probability and Normal Tissue Complication Probability (see S. Webb, “Optimizing radiation therapy inverse treatment planning using the simulated annealing technique”, Int. Journal of Imaging Systems and Tech., Vol. 6, pp 71-79, 1995, which summarizes the extensive work over many years by that author and co-workers) and
b2) using non-analytic cost functions that describe the fractions of PTV and OAR volumes that are to receive no more or less than a certain range of doses. The minimization of those cost functions is carried out by the simulated annealing method, (see U.S. Pat. No. 6,038,283, M. P. Carol, R. C. Campbell, B. Curran, R. W. Huber and R. V. Nash, “Planning method and apparatus for radiation dosimetry”, Mar. 14, 2000).
The above methods are either in use at several institutions in the U.S. and elsewhere or being prepared for inclusion in Radiation Therapy Planning commercial software packages by several Companies.
An advantage attributed to the stochastic method of U.S. Pat. No. 6,038,283 is that it allows the therapist to define cost functions that prescribe the fractions of PTV and OAR volumes that are to receive no more or less than a certain range of doses. In order to provide that advantage, that method has to use non-analytic cost functions without explicit first partial derivatives. Testing whether a small change in the fluence of a randomly selected beam will result in an increase or decrease of the cost function involves complex calculations that slow down the op
LandOfFree
Stochastic method for optimization of radiation therapy... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Stochastic method for optimization of radiation therapy..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Stochastic method for optimization of radiation therapy... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2909375