System and method for estimating and manipulating estimated radiation dose

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
7Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1. A method for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel, the method comprising:
 (a) determining a dose modification voxel for which it is desired to modify the dose value and determining a corresponding magnitude of desired dose modification;
(b) receiving the achievable dose distribution estimate, the dose modification voxel and the corresponding magnitude of desired dose modification as input to a processor;
(c) for each of a plurality of beams;
(i) characterizing, using the processor, each beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing a projection of the beamlet into space; and
(ii) identifying, using the processor, one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel;
(d) modifying, using the processor, the intensity values of the one or more identified dosechange beamlets based on the magnitude of desired dose modification, wherein modifying the intensity values of the one or more identified dosechange beamlets comprises, for each of the one or more identified dosechange beamlets, modifying the intensity value of each identified dosechange beamlet by a corresponding amount, the corresponding amount based on the magnitude of the desired dose modification;
(e) updating, using the processor, the achievable dose distribution estimate to account for the modified intensity values of the identified dosechange beamlets; and
(f) outputting, from the processor, the updated achievable dose distribution estimate.
1 Assignment
0 Petitions
Accused Products
Abstract
Methods are provided for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject. One such method comprises: determining a dose modification voxel for which it is desired to modify the dose value and a corresponding magnitude of desired dose modification; for each of a plurality of beams: (i) characterizing the beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing the projection of the beamlet into space; and (ii) identifying one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel; modifying the intensity values of at least one of the dosechange beamlets; and updating the achievable dose distribution estimate to account for the modified intensity values of the at least one of the dosechange beamlets.
13 Citations
View as Search Results
Method and Apparatus for Measuring, Verifying, and Displaying Progress of Dose Delivery in Scanned Beam Particle Therapy  
Patent #
US 20150087885A1
Filed 09/22/2014

Current Assignee
Pyramid Technical Consultants Inc.

Sponsoring Entity
Pyramid Technical Consultants Inc.

Intensitymodulated ion therapy  
Patent #
US 9,555,265 B2
Filed 11/18/2015

Current Assignee
LOMA Linda University Medical Center

Sponsoring Entity
LOMA Linda University Medical Center

Method and apparatus for measuring, verifying, and displaying progress of dose delivery in scanned beam particle therapy  
Patent #
US 9,731,149 B2
Filed 09/22/2014

Current Assignee
Pyramid Technical Consultants Inc.

Sponsoring Entity
Pyramid Technical Consultants Inc.

Method and Apparatus for Measuring, Verifying, and Displaying Progress of Dose Delivery in Scanned Beam Particle Therapy  
Patent #
US 20170246479A1
Filed 05/11/2017

Current Assignee
Pyramid Technical Consultants Inc.

Sponsoring Entity
Pyramid Technical Consultants Inc.

Systems and methods for intensity modulated radiation therapy  
Patent #
US 9,884,206 B2
Filed 07/22/2016

Current Assignee
CarmelHaifa University Economic Corp. Ltd.

Sponsoring Entity
CarmelHaifa University Economic Corp. Ltd.

Method and apparatus for measuring, verifying, and displaying progress of dose delivery in scanned beam particle therapy  
Patent #
US 10,092,777 B2
Filed 05/11/2017

Current Assignee
Pyramid Technical Consultants Inc.

Sponsoring Entity
Pyramid Technical Consultants Inc.

Method and apparatus for measuring, verifying, and displaying progress of dose delivery in scanned beam particle therapy  
Patent #
US 10,456,598 B2
Filed 08/29/2018

Current Assignee
Pyramid Technical Consultants Inc.

Sponsoring Entity
Pyramid Technical Consultants Inc.

Planning system, method and apparatus for conformal radiation therapy  
Patent #
US 7,831,289 B2
Filed 10/07/2004

Current Assignee
Best Medical International Incorporated

Sponsoring Entity
Best Medical International Incorporated

PLANNING SYSTEM, METHOD AND APPARATUS FOR CONFORMAL RADIATION THERAPY  
Patent #
US 20100183121A1
Filed 03/29/2010

Current Assignee
Best Medical International Incorporated

Sponsoring Entity
Best Medical International Incorporated

METHODS AND APPARATUS FOR THE PLANNING AND DELIVERY OF RADIATION TREATMENTS  
Patent #
US 20080298550A1
Filed 06/03/2008

Current Assignee
Varian Medical Systems International AG

Sponsoring Entity
Karl Otto

Method of modulating laseraccelerated protons for radiation therapy  
Patent #
US 20070034812A1
Filed 06/01/2006

Current Assignee
Fox Chase Cancer Center

Sponsoring Entity
Fox Chase Cancer Center

Treatment planning software and corresponding user interface  
Patent #
US 20060274885A1
Filed 09/30/2005

Current Assignee
Accuray Incorporated

Sponsoring Entity
Accuray Incorporated

RADIATION TREATMENT PLANNING SYSTEM AND COMPUTER PROGRAM PRODUCT  
Patent #
US 20120136677A1
Filed 06/08/2010

Current Assignee
Deutsches Krebsforschungszentrum Stiftung des Offentlichen Rechts

Sponsoring Entity
Deutsches Krebsforschungszentrum Stiftung des Offentlichen Rechts

49 Claims
 1. A method for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel, the method comprising:
(a) determining a dose modification voxel for which it is desired to modify the dose value and determining a corresponding magnitude of desired dose modification; (b) receiving the achievable dose distribution estimate, the dose modification voxel and the corresponding magnitude of desired dose modification as input to a processor; (c) for each of a plurality of beams; (i) characterizing, using the processor, each beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing a projection of the beamlet into space; and (ii) identifying, using the processor, one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel; (d) modifying, using the processor, the intensity values of the one or more identified dosechange beamlets based on the magnitude of desired dose modification, wherein modifying the intensity values of the one or more identified dosechange beamlets comprises, for each of the one or more identified dosechange beamlets, modifying the intensity value of each identified dosechange beamlet by a corresponding amount, the corresponding amount based on the magnitude of the desired dose modification; (e) updating, using the processor, the achievable dose distribution estimate to account for the modified intensity values of the identified dosechange beamlets; and (f) outputting, from the processor, the updated achievable dose distribution estimate.  View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41)
 42. A method for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel, the method comprising:
(a) determining a dose modification voxel for which it is desired to modify the dose value and determining a corresponding magnitude of desired dose modification; (b) receiving the achievable dose distribution estimate, the dose modification voxel and the corresponding magnitude of desired dose modification as input to a processor; (c) for each of a plurality of beams; (i) characterizing, using the processor, each beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing a projection of the beamlet into space; and (ii) identifying, using the processor, one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel; (d) modifying, using the processor, the intensity values of the one or more identified dosechange beamlets based on the magnitude of desired dose modification, wherein modifying the intensity values of the one or more identified dosechange beamlets comprises, for each of the one or more identified dosechange beamlets, modifying the intensity value of each identified dosechange beamlet by a corresponding amount, the corresponding amount based on the magnitude of the desired dose modification; and (e) updating, using the processor, the achievable dose distribution estimate to account for the modified intensity values of the one or more identified dosechange beamlets and, after updating, using the processor, the achievable dose distribution estimate, determining, using the processor, one or more radiation delivery parameters based on the updated achievable dose estimate, the one or more radiation delivery parameters suitable for use by the radiation delivery apparatus for treatment of the subject; (f) outputting, from the processor, the updated achievable dose distribution estimate and the one or more radiation delivery parameters; wherein determining, using the processor, one or more radiation delivery parameters based on the updated achievable dose estimate comprises performing an iterative optimization process; wherein performing the iterative optimization process comprises estimating an optimized dose distribution over a threedimensional range of voxels resulting from an optimization beam having a spatially varying twodimensional optimization intensity distribution characterized by a twodimensional array of optimization beamlets wherein each optimization beamlet is associated with a corresponding optimization intensity value, the estimating comprising; associating an optimization ray line with each optimization beamlet, the optimization ray line representing a projection of the optimization beamlet into space; convolving the twodimensional optimization intensity distribution with a twodimensional optimization dose estimate kernel to obtain a twodimensional optimization convolved intensity distribution, the twodimensional optimization convolved intensity distribution comprising an optimization convolved intensity value for each optimization beamlet; and for each optimization beamlet in the twodimensional array of optimization beamlets;
identifying voxels in the three dimensional range of voxels that are intersected by the optimization ray line associated with the optimization beamlet; and
adding an optimization dose contribution to the intersected voxels, the added optimization dose contribution based on the optimization convolved intensity value of the optimization beamlet.
 43. A system for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel, the system comprising a computerbased controller connectable to control a radiation delivery apparatus, the controller configured to:
determine a dose modification voxel for which it is desired to modify the dose value and a corresponding magnitude of desired dose modification; for each of a plurality of beams; characterize each beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing a projection of the beamlet into space; and identify one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel; modify the intensity values of the one or more identified dosechange beamlets based on the magnitude of the desired dose modification, wherein the controller is configured to modify the intensity values of the one or more identified dose change beamlets by, for each of the one or more identified dosechange beamlets, modifying the intensity value of each identified dosechange beamlet by a corresponding amount, the corresponding amount based on the magnitude of the desired dose modification; update the achievable dose distribution estimate to account for the modified intensity values of the identified dosechange beamlets.  View Dependent Claims (44, 45, 46, 47, 48, 49)
1 Specification
This application claims priority from, and claims the benefit under 35 USC 119(e) of, U.S. application No. 61/398,286 filed 22 Jun. 2010 which is hereby incorporated herein by reference.
The invention relates to planning and delivery of radiation. Particular embodiments provide methods and systems for planning and delivering does distributions for radiation therapy.
Radiation is used in the treatment of cancer as well as some other medical conditions. When radiation interacts with tissue, energy from the radiating particles is transferred and deposited within the tissue. The energy is normally deposited in the vicinity of the transfer. The maximum deposition is normally close to the point of interaction. The energy deposited causes damage to cells that may eventually lead to cell death. The quantity of energy deposited is normally described as radiation dose and has the units of Gray (Gy). 1 Gray is equal to 1 Joule per kilogram of medium. The primary goal of radiation treatment is to eradicate cancerous cells in a subject by depositing sufficient radiation dose.
Radiation dose can damage or kill both cancerous and healthy tissue cells. It is typical that some healthy tissue will receive radiation dose during a radiation treatment. For example, a radiation beam originating from a radiation source and projecting through a subject will deposit radiation dose along its path. Any healthy tissue located within the path will normally receive some radiation dose. Additionally, some radiation dose will typically be deposited outside of the beam path into healthy tissue due to radiation scatter and other radiation transport mechanisms. One of the challenges of radiation therapy is to deposit dose in cancerous tissue while minimizing dose received by healthy tissue. Furthermore, some healthy tissues are more sensitive to radiation dose than others making it more important to avoid radiation dose in those tissues.
Modern radiation delivery systems are capable of delivering complex dose distributions. There is a desire for the radiation therapy clinician to be capable of evaluating, determining and/or optimizing tradeoffs between delivering dose to a tumor and minimizing dose delivered to healthy tissue. Current techniques for evaluating these tradeoffs (treatment plan optimization, for example) are cumbersome and disconnect the operator from quick and direct manipulation and evaluation of achievable dose distributions. There is a desire for improvement of systems and methods for estimating achievable dose distributions and possibly improving the evaluation of tradeoffs between radiation dose to cancerous and healthy tissue.
The following embodiments and aspects thereof are described and illustrated in conjunction with systems, tools and methods which are meant to be exemplary and illustrative, not limiting in scope. In various embodiments, one or more of the abovedescribed problems have been reduced or eliminated, while other embodiments are directed to other improvements.
One aspect of the invention provides a method for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel. The method comprises: determining a dose modification voxel for which it is desired to modify the dose value and a corresponding magnitude of desired dose modification; for each of a plurality of beams: (i) characterizing the beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value and a ray line representing the projection of the beamlet into space; and (ii) identifying one or more dosechange beamlets which have associated ray lines that intersect the dose modification voxel; modifying the intensity values of at least one of the dosechange beamlets; and updating the achievable dose distribution estimate to account for the modified intensity values of the at least one of the dosechange beamlets.
Another aspect of the invention provides a method for permitting manipulation of an achievable dose distribution estimate deliverable by a radiation delivery apparatus for proposed treatment of a subject, the achievable dose distribution estimate defined over a threedimensional range of voxels with a dose value for each voxel. The method comprises: determining a dose modification voxel for which it is desired to modify the dose value and a corresponding magnitude of desired dose modification; for each of a plurality of beams: (i) characterizing the beam as a twodimensional array of beamlets, wherein each beamlet is associated with a corresponding intensity value; (ii) associating a ray line with each voxel, the ray line projecting from the voxel and intersecting the two dimensional array of beamlets; and (iii) identifying a dosechange beamlet to be the beamlet intersected by the ray line associated with the dose modification voxel; modifying the intensity values of one or more of the dosechange beamlets; and updating the achievable dose distribution estimate to account for the modified intensity values of the one or more of the dosechange beamlets.
Another aspect of the invention provides a method for estimating a dose distribution over a threedimensional range of voxels resulting from a beam having a spatially varying twodimensional intensity distribution characterized by a twodimensional array of beamlets wherein each beamlet is associated with a corresponding intensity value. The method comprises: associating a ray line with each beamlet, the ray line representing the projection of the beamlet into space; convolving the twodimensional intensity distribution with a twodimensional dose estimate kernel to obtain a twodimensional convolved intensity distribution, the twodimensional convolved intensity distribution comprising a convolved intensity value for each beamlet; and for each beamlet in the twodimensional array of beamlets: identifying voxels in the threedimensional range of voxels that are intersected by the ray line associated with the beamlet; and adding a dose contribution to the intersected voxels, the dose contribution based on the convolved intensity value of the beamlet.
Other aspects of the invention provide systems comprising one or more controllers configured to perform the methods of various embodiments of the invention. Other aspects of the invention provide computer program products carrying instructions embodied in a nontransitory computerreadable medium, the instructions when executed by a suitable processor cause the processor to perform the methods of various embodiments of the invention. Other aspects of the invention provide methods, systems and/or computer program products that use the methods of various embodiments for planning and/or delivery of radiation treatment to subjects.
In addition to the exemplary aspects and embodiments described above, further aspects and embodiments will become apparent by reference to the drawings and by study of the following detailed descriptions.
Exemplary embodiments are illustrated in referenced figures of the drawings. It is intended that the embodiments and figures disclosed herein are to be considered illustrative rather than restrictive. In drawings which illustrate nonlimiting embodiments:
Throughout the following description specific details are set forth in order to provide a more thorough understanding to persons skilled in the art. However, well known elements may not have been shown or described in detail to avoid unnecessarily obscuring the disclosure. Accordingly, the description and drawings are to be regarded in an illustrative, rather than a restrictive, sense.
In the exemplary
Radiation emitted toward subject 110 is permitted to pass through enclosure 117 and through a beamshaping system 118. Beamshaping system 118 may comprise one or more collimators which may be used to define a beam of radiation that emanates from radiation source 111 and into subject 110. The collimators of beamshaping system 118 may be motorized and their position and/or movement may be controlled (e.g. by a suitably configured computer control system or the like). The collimators of beamshaping system 118 may be controllably configured so that the shape of the radiation beam entering subject 110 preferentially intersects cancerous tissue. In the illustrated example radiation delivery apparatus 10 of
The
Another example of a technique for using radiation delivery apparatus 10 in a manner which attempts to provide desired dose to a tumor of diseased tissue while minimizing dose to healthy tissue involves collimating each radiation beam (i.e. the beam from each relative configuration of gantry 112 and table 113) so that the projection of the tumor from the view of the radiation source (Beam'"'"'s eye view) closely approximates the outline of the tumor. In this way dose to healthy tissue surrounding the tumor will be reduced. Collimation system 118 (e.g. MLC 118A) may be used to collimate the individual beams. This collimation technique may be improved by selectively (partially or fully) blocking portions of a radiation beam (from a first direction) that intersect both tumor and sensitive healthy tissue and then compensate for the blocked portion of the tumor by selectively (partially or fully) unblocking portions of one or more radiation beams from one or more other directions. A radiation beam (from a particular direction) comprising spatially nonuniform transmitted portions may be referred to as “Intensity Modulated” in reference to the spatially varying intensity of radiation across the twodimensional beam projection. Intensity modulation can further improve the difference between dose received by healthy tissue and dose received by tumor, particularly in circumstances where some healthy tissue is of relatively high importance (e.g. healthy organs) and/or is in relatively close proximity to the target tissue and it is desired to impart even less dose to such highly important tissues.
In accordance with other techniques, it may be beneficial to dynamically move one or more components of radiation delivery apparatus 10 during the delivery of radiation. For example, collimation system 118 can change the shape of a radiation beam while source 117 is emitting radiation, thereby providing dynamically varying collimation shapes for intensity modulation. Additionally or alternatively, gantry 112 and table 113 can move relative to one another while source 117 is emitting radiation, thereby providing continuously varying beam directions (in contrast to a finite number of discrete beam directions). Some techniques may involve dynamically varying the position of radiation source 117 through a motion trajectory (e.g. relative movement of gantry 112 and table 113) while simultaneously dynamically varying the collimated beam shape and/or the intensity of radiation source 117.
Radiation delivery apparatus 10 represents only one example of a radiation delivery apparatus. Other types of radiation delivery apparatus may be used to deliver therapeutic radiation to a subject. A number of nonlimiting examples of radiation delivery apparatus include CyberKnife™ (Accuracy Incorporated), Tomotherapy™ (Tomotherapy Incorporated) and Gammaknife™ (Elekta AB).
Using radiation delivery apparatus (like exemplary apparatus 10, 12 of
Since the size, shape and position of a tumor with respect to the surrounding healthy tissue are different for each subject, a diagnostic imaging procedure is typically used prior to (or as a part of) treatment planning for the purposes of determining the spatial locations of diseased and healthy tissue. Computed Tomography (CT), Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) are common imaging methods used for this diagnostic imaging process. The result of CT, MRI and PET imaging may comprise 3dimensional images which contain anatomical and functional information. In some embodiments, diagnostic imaging may involve procuring 4dimensional images, which incorporate time varying image information (e.g. to account for breathing).
The locations of diseased and healthy tissue may be identified on these images prior to (or as a part of) treatment planning. Identification of the locations of diseased and healthy tissue can be performed manually although methods for automatic and semiautomatic identification may also be used. Treatment planning may involve using well known methods to model the radiation dose resulting from a radiation beam. The dose that a subject would receive from a particular radiation beam may be evaluated by overlaying the dose distribution modeled for the particular radiation beam on the subject'"'"'s images. The dose that a subject would receive from a proposed treatment plan (e.g. a plurality of radiation beams) may be evaluated by superposing dose distributions modeled for the individual beams of the plan and overlaying the superposed dose distribution on the subject'"'"'s images.
A proposed treatment plan may be characterized or otherwise specified by a set of radiation delivery parameters. As used in this specification and the accompanying claims, radiation delivery parameters may comprise: beam configuration parameters which may relate to the geometric positioning of the radiation source with respect to the subject (e.g. numbers of beams, directions of beams, radiation energy of beams, motion of beams (e.g. for continuously varying beams) and/or the like) and/or beam delivery parameters which may relate to the characteristics of one or more beam configurations (e.g. collimation shape(s) and corresponding collimation system configuration(s), motion of collimation shape(s) and corresponding collimation system configuration(s) (e.g. for continuously varying collimation shapes), radiation intensity(s), and/or the like). By estimating/modeling the dose distributions from multiple different treatment plan proposals (as characterized by multiple corresponding sets of radiation delivery parameters), the multiple different treatment plan proposals can be compared against one another. Once a proposed treatment plan is selected (e.g. because it is determined to be superior to others or is otherwise determined to be satisfactory), the radiation delivery parameters associated with the selected treatment plan may be transferred to the radiation delivery apparatus for delivery of the selected treatment plan to the subject.
Treatment planning for intensity modulated radiation therapy (IMRT) may be more complex because of the permissible spatial variation of intensity distribution across a twodimensional cross section of each beam or each portion of a beam. Because of the spatial varying intensity distributions of beams associated with IMRT, IMRT treatment planning typically involves dividing each beam into a twodimensional matrix of spatially varying intensity portions which may be referred to as beamlets. Each beamlet may be effectively treated as a separate beam element that follows a ray line from the source into the subject. In a typical nonlimiting IMRT plan, there may be 5 to 9 beams each with a matrix that may comprise more than 100 beamlets.
IMRT planning by a human observer is normally considered impractical due to the large number of beamlets. Several computer algorithms have been developed to determine the spatially varying intensity distributions of each beam in IMRT plans—e.g. the intensities for each of the beamlets in each portion of each beam of the IMRT plan. These algorithms typically involve iterative optimization. For example, at each iteration, a particular set of radiation delivery parameters is proposed, the corresponding dose distribution is calculated (e.g. modeled) and the corresponding dose distribution is evaluated by comparing a quality metric associated with the corresponding dose distribution to some objective. The next iteration may then attempt to propose a set of radiation delivery parameters whose dose distribution is superior to the previous iteration (when evaluated in relation to the same objective). The iterative process normally repeats until an optimization termination criterion is achieved. The iterative optimization process may be relatively time consuming, computationally expensive and/or temporally inefficient, because of the need to model/calculate a dose distribution at each iteration.
Block 131 involves specifying dose objectives for target and healthy tissues. Ideally, the dose objectives might be: (i) 100% of the tumor volume receives at least the prescription dose; (ii) 0% of the tumor volume receives dose greater than the prescription dose; and (iii) 0% of the healthy tissue volume receive any dose. Such an ideal objective is not realizable in practice. Instead, achieving a prescription dose to the tumor must be balanced against providing a low dose to healthy tissue structures. Also, there are commonly many healthy tissue structures of concern that vary in importance which can make the number of possible tradeoffs cumbersome. By way of nonlimiting example: a healthy tissue objective may comprise maximum 20% of the healthy tissue structure volume receives 30% of the dose prescribed to the target; and a target tissue objective may comprise minimum of 90% of the tumor target volume receives 95% of the dose prescribed to the target. Block 131 may also involve specifying a dose quality metric which may be used to evaluate proposed treatment plan iterations against the block 131 dose distribution objectives. Such dose quality metrics may additionally or alternatively be used to indirectly specify dose distribution objectives.
Method 14 then proceeds to block 132 which involves performing an iterative optimization process to arrive at a set of radiation delivery parameters. A typical iterative optimization process which may be implemented as a part of block 132 was discussed above. At the conclusion of the block 132 iterative optimization process, method 14 proceeds to block 133 which involves evaluating the treatment plan (e.g. the calculated/modeled dose distribution) which results from using the radiation delivery parameters output by the block 132 optimization. The block 133 evaluation may be performed by one or more human operators. In some embodiments, however, the block 133 evaluation may be automated. For example, constraints could be developed that specify a minimum number of objectives that must be achieved in order to achieve a positive block 133 evaluation. The number of achieved objectives might be more important for some target structures or for some specified healthy tissue structures. Even in the case of an automated block 133 evaluation, the resulting dose distribution would likely be evaluated by a clinician before actually delivering radiation.
If the block 131 dose distribution objectives are not achievable in practice, then the block 132 optimization may fail to determine an acceptable treatment plan. Conversely, if the block 131 dose distribution objectives are too easily achieved, the treatment plan specified by the block 132 optimization may not achieve the best tradeoffs that could be realized in practice. In these circumstances (block 133 NO output path), method 14 may loop back to block 131, where a different optimization may be performed with different dose distribution objectives. Typically, to ensure that the dose distribution achieved in treatment planning method 14 is close to optimal, it is desirable to perform multiple block 132 optimizations with multiple sets of block 131 dose distribution objectives. Once it is determined in block 133 that a particular plan is optimal (block 133 YES output path), then method 14 may proceed to block 134 where the radiation delivery parameters may be transferred to the radiation delivery apparatus for delivery to the subject.
Other methods of providing complex treatment plans and corresponding dose distributions may involve similar iterative optimization. Any combination of radiation delivery parameters may be used in an optimization process for deriving a treatment plan. Such complex treatment plans and corresponding dose distributions may include, by way of nonlimiting example, socalled direct aperture optimization techniques, radiation treatment techniques involving dynamic variation of the direction of radiation (e.g. movement of the radiation source during treatment) and/or dynamic variation of collimation shapes and/or intensities within particular beams during treatment and radiation treatment techniques using radiation delivery apparatus such as CyberKnife™ (Accuracy Incorporated), Tomotherapy (Tomotherapy Incorporated) and Gammaknife™ (Elekta AB). The iterative optimizations involved in treatment planning for all of these radiation delivery techniques suffer from similar drawbacks as those discussed above for the iterative optimization associated with IMRT treatment planning. More particularly, such optimizations are time consuming and computationally expensive because of the need for calculating/modeling a dose distribution at each iteration.
One aspect of the invention provides systems and methods for permitting manipulation of achievable dose distribution estimates. In particular embodiments, estimated dose distributions and associated dose quality metrics may be manipulated without the cumbersome and computationally expensive calculations involved in simulating dose for specific radiation delivery parameters (e.g. without the need for iterative optimization). These methods and systems may be simple to use and may permit operator manipulation of estimated dose distributions and associated dose quality metrics. By way of nonlimiting example, an operator may select a graphical representation of a dose quality metric using a computer mouse or similar computer pointing device, drag it to the left or right, up or down as desired. As another nonlimiting example, an operator may modify a graphical representation of a dose distribution using a computer mouse or similar computer pointing device, to “paint” or “erase” dose from a region of subject anatomy. As these operatordirected manipulations are made, the achievable dose distribution estimate and corresponding dose quality metrics may be updated in near realtime.
The range of physically achievable dose distributions may be limited. Particular embodiments involve the imposition of limits or restrictions on available manipulations, so that the estimated dose distributions (after operator manipulation) are at least approximately achievable. In this way operators are able to rapidly explore tradeoffs between dose delivery to target tissues (e.g. tumor(s)) and healthy tissues (e.g. organ(s)) while ensuring that the subject will ultimately receive a dose distribution substantially similar to the estimated one.
Method 16 commences in block 141 which involves delineating target (e.g. diseased) tissue and healthy tissue within the subject using whatever diagnostic image information may be available. Block 141 may be substantially similar to block 130 described above for method 14 (
At the conclusion of block 142 (e.g. where the operator is satisfied with the manipulated version of the achievable dose distribution or otherwise), method 16 proceeds to optional block 143 which involves determining radiation delivery parameters capable of permitting a radiation delivery apparatus to deliver the estimate of achievable dose as output from block 142. Block 143 may involve performing an iterative optimization process or the like to derive radiation delivery parameters (e.g. beam configuration parameters and/or beam delivery parameters). The block 143 iterative optimization may involve processes similar to those described in blocks 131 and 132 of treatment planning process 14 described above (
After optimization to obtain the radiation delivery parameters in block 143, method 16 proceeds to block 144 which involves transferring the block 143 radiation delivery parameters to the controller of a radiation delivery apparatus. These radiation delivery parameters may then be used by the controller of the radiation delivery apparatus in block 145 to cause the radiation delivery apparatus to deliver radiation to the subject in accordance with the radiation treatment plan corresponding to the radiation delivery parameters.
As discussed briefly above, block 142 of radiation treatment method 16 involves generating, and permitting operator manipulation of, an achievable dose distribution.
Method 18 commences with initialization in block 218. The block 218 initialization may involve: establishing a calculation grid over a region of interest in the delineated image information; defining a configuration of beams; defining an initial intensity distribution of beamlets for each beam; generating an initial estimate of an achievable dose distribution using the beam configuration and beamlet intensity distributions; and, optionally, determining an initial estimate of one or more dose quality metrics based on the initial estimated dose distribution.
Returning to
Initialization method 20 then proceeds to block 44 which involves defining an initial intensity distribution of beamlets for each of the block 42 beams. Such initial intensity distributions may be similar to initial values for parameters similar to beam delivery parameters discussed above.
Block 44 also involves assigning initial intensity values to the individual beamlets 164 for each beam 159—i.e. initializing the intensity distribution 165 for each beam 159. The initial intensity distributions 165 may be defined in block 44 using a variety of different techniques. By way of nonlimiting example:
 Intensity distribution 165 is zero for all beamlets 164 and all beams 159. This initial intensity assignment corresponding to zero dose for all structures.
 Intensity distribution 165 is random over each grid 162 of beamlets 164 for each beam 159, intensity of each beamlet 164 is random over all intensity distributions 165 for all beams 159 or some other suitable scheme involving at least some form of random assignment of intensities to corresponding beamlets 162.
 Intensity distributions 165 for all or a subset of beams 159 are assigned so that an approximate sphere of dose encompasses target structure(s) 153.
 Intensity distributions 165 for all or a subset of beams 159 are assigned such that beamlets 162 corresponding to ray lines 163 that intersect target structure(s) 153 are assigned an initial positive intensity value (e.g. unity) and beamlets 162 are otherwise assigned a different initial value (e.g. zero or some other relatively low value).
 Intensity distributions 165 for all or a subset of beams 159 are rescaled (e.g. by an equal amount) so that the resultant estimated dose distribution would have a maximum dose corresponding to some dose threshold (e.g. a dose threshold equal to a highest prescription dose for target structure(s) 153).
 Intensity distributions 165 for all or a subset of beams 159 are proportional or correlated with a desired (e.g. prescription) dose for target structure(s) 153.
 Intensity distributions 165 are generated for all or a subset of beams 159 where beamlets 162 corresponding to ray lines 163 that intersect target structure(s) 153 are assigned an intensity value proportional to or correlated with a desired dose for each such target structure 153.
 A margin (e.g. offset) may be added to the intensity distributions 165 (or to individual beamlets 162 corresponding to ray lines 163 that intersect target structure(s) 153) to ensure proper coverage of the target structure(s) 153 by the resulting dose distribution.
 For target structures 153 that overlap in a given beam 159 (e.g. a ray line 163 intersects multiple target structures 153), the intensity of the corresponding beamlet 164 may be assigned an intensity proportional to or correlated with the desired dose for the target structure with the highest desired dose and/or the corresponding beamlet 164 may have its assigned intensity weighted by the target structure 153 that has the highest desired dose.
 For target structures that overlap in a given beam 159, the intensity of the corresponding beamlet 164 may be assigned an intensity proportional to or correlated with the desired dose for the target structure with the lowest desired dose and/or the corresponding beamlet 164 may have its assigned intensity weighted by the target structure 153 that has the lowest desired dose.
Returning to
Method 20 may then proceed to block 48 which involves optionally determining one or more initial dose quality metrics based on the block 46 initial dose distribution estimate. Dose quality metrics determined in block 48 may generally comprise any function of the estimated dose distribution. Some dose quality metrics include:
 The average dose to a structure;
 Dose volume histogram(s)—often referred to as DVHs;
 Rate of dose falloff outside target structure(s) 153;
 Dose conformity indices—e.g. how closely the prescription dose matches the shape of target structure(s) 153);
 Radiobiological objective(s)—e.g. tumor control probability, normal tissue complication probability, equivalent uniform dose, and/or the like.
DHVs represent one popular and widely used dose quality metric. A DVH is a graphical plot of structure volume (target tissue or healthy tissue) on the Yaxis versus dose on the Xaxis. It is common to use cumulative DVHs (which are typically referred to simply as ‘DVHs’) when evaluating treatment plans.
Returning to the
 Keyboard entry (e.g. typing a spatial location (e.g. calculation grid coordinates) of the desired dose modification voxel and the amount of the desired dose modification or typing a desired modification to a dose quality metric).
 Manipulation of graphical representations of dose distributions and/or dose quality metrics via a graphical user interface.
 Specifying a location on a graphical representation of the subject using a mouse or similar computer pointing device and, using a mouse button or keyboard input to indicate whether to increase or decrease the dose at that location and/or to indicate the magnitude of the desired dose modification.
 Other computer input devices (trackball, touch screen, voice command, video command and/or the like) may also be used
 inverting the dose quality metric function, so that dose magnitude and dose coordinates become function(s) of the dose quality metric; and
 calculating the required dose distribution modification (magnitude and location) by applying the operatorindicated modifications to the dose quality metric value in the inverted dose metric function.
In some embodiments, where desired dose modification input is received in the form of manipulation of dose quality metrics, techniques other than inversion may be used to predict desired dose modification voxels and corresponding dose modification magnitudes. In one particular example, a change to a DVH is received as input which comprises a change to a point on a DVH curve which may be identified by a corresponding dose (D_selected) and a corresponding volume (V_selected). The dose modification voxels may then be identified (in block 220) using a number of techniques including: identifying voxels to be dose modification voxels if the voxels have values falling within D_selected+/−Δ, where Δ may be a fixed value (which may be operatorselectable), a fraction of D_selected (which may be an operatorselectable fraction), a value determined by calibration or empirical evidence and/or the like. If no voxels are identified to be dose modification voxels, then Δ may be expanded and voxels may be reidentified until at least one dose modification voxel is identified. If a large number of voxels (e.g. all voxels for that structure or a number of voxels greater than a threshold number or a threshold percentage of the voxels in a structure) are identified to be dose modification voxels, then Δ may be reduced and voxels may be reidentified. In some embodiments, all voxels inside the structure for which the DVH is changed may be identified as dose modification voxels. By way of nonlimiting example, the magnitudes of dose modifications may be determined on the basis of: a fraction of the D_selected value; a fraction of the maximum dose for the structure to which the DVH corresponds; a fraction of a prescription dose assigned to the structure to which the DVH corresponds; a fraction of the maximum prescription dose assigned to all structures; a fraction correlated with (e.g. proportional to) the amount of mouse or similar computer pointing device movement by an operator; a operatorselected quantity; a fixed quantity which may be an operatorconfigurable parameter or may be a “hard coded” constant; a combination of the above; and/or the like.
Block 220 involves determining the coordinates (e.g. voxel location(s)) and magnitudes of desired dose changes. The coordinates determined in block 220 to be associated with desired dose changes may be referred to as the desired dose modification coordinate(s)/voxel(s) and the associated magnitudes may be referred to as the desired dose modification magnitude(s). Typically, although not necessarily, such desired dose modification coordinate(s) and magnitude(s) are determined on the basis of operator input, but could additionally or alternative be generated based on other forms of input (e.g. computergenerated automated test input and/or the like). Such operator input may involve direct specification of desired dose modification coordinates and magnitudes or indirect specification of desired dose modification coordinates and magnitudes through specification of desired changes to one or more dose quality metrics. The examples shown in
In some embodiments, block 220 may involve determining secondary dose modification coordinates/voxels and corresponding dose modification magnitudes in addition to the primary dose modification coordinates/voxels and corresponding dose modification magnitudes determined in accordance with the techniques described above. Secondary dose modification voxels may be defined in a marginal region proximate to the primary dose modification voxels. By way of nonlimiting example, such secondary dose modification voxels could be determined to be in a marginal region less than or equal to a threshold number of voxels away from the primary dose modification coordinates. The threshold number of voxels that define the marginal region may be operatorconfigurable, may be a system parameter which may be determined by one or more suitable calibration procedures, may be a system parameter which may be determined based on empirical testing and/or data and/or the like.
The secondary dose modification magnitudes of the secondary dose modification voxels may be less than the primary dose modification magnitudes of their corresponding primary dose modification voxels. For example, the secondary dose modification magnitudes may be a fraction a of the primary dose modification magnitudes (where 0<=a<=1). This fraction may be a function of the distance between a given secondary dose modification voxel and its corresponding primary dose modification voxel—e.g. within a marginal region of 3 voxels around a primary dose modification voxel, the fraction a may be relatively high for the secondary dose modification voxels that are nearest neighbors to the primary dose modification voxel; lower for the secondary dose modification voxels that are spaced by one voxel from the primary dose modification voxel; and lowest for the secondary dose modification voxels that are spaced by two voxels from the primary dose modification voxel.
In other respects (e.g. for the purposes of other procedures involved in the methods and systems described herein), secondary dose modification voxels determined as a part of block 220 may be treated, for the most part, in the same manner as primary dose modification voxels determined in block 220. Accordingly, both primary dose modification voxels and secondary dose modification voxels determined in block 220 may be referred to simply as dose modification voxels.
Returning to
Method 18 then proceeds to block 223, which involves adjusting the intensities of the dosechange beamlets 164 identified in block 221. For example, if it is determined in block 220 that the dose corresponding to a particular desired dose modification voxel is to be decreased, then block 223 will typically involve decreasing intensities of the corresponding dosechange beamlets 164. Conversely, if it is determined in block 220 that the dose corresponding to a particular desired dose modification voxel is to be increased, then block 223 will typically involve increasing intensities of the corresponding dosechange beamlets 164.
Changes to the intensities of the dosechange beamlets 164 in block 223 may be effected using a wide variety of techniques. By way of nonlimiting example:
 identical magnitude intensity changes or identical percentage intensity changes may be applied to each dosechange beamlet 164;
 identical intensity changes may be applied to one or more subsets of dosechange beamlets 164. Such subsets of dosechange beamlets 164 could correspond, for example, to consecutive dosechange ray lines 163, to dosechange beamlets 164 from every n^{th }(e.g. every second or third) beam 159, to randomly selected dosechange beamlets 164 and/or the like;
 different magnitude or percentage intensity changes may be applied for all or a subset of dosechange beamlets 164; and/or
 the like.
The magnitudes of the block 223 changes to the intensities of the dosechange beamlets 164 may be a function of (e.g. correlated with or proportional to) the block 220 desired dose modification magnitude(s). As discussed briefly above, initialization method 20 may optionally include a procedure (block 47) which establishes an approximate relationship between the intensities of beamlet(s) 164 having ray line(s) 163 that intersect a particular voxel (e.g. dosechange beamlets 164 that intersect a dose modification voxel) and the corresponding dose delivered to the particular voxel. While this approximate relationship is shown as being determined in block 47 of the illustrated embodiment, this is not necessary and this approximate relationship may be determined as a part of one or more other procedures, including, possibly, separate procedures. This approximate relationship may be used as a part of the block 223 determination of changes to the intensities of the dosechange beamlets 164. The approximate relationship between the intensities of dosechange beamlets 164 and the magnitude of a dose change to a dose modification voxel may have a form D=gni where D is the dose (or dose change) for the dose modification voxel, i is an approximate intensity (or intensity change) value for a dosechange beamlet 164, n represents the number of dosechange beamlets 164 for the dose modification voxel and g is a scaling variable. The scaling variable g may be determined in block 47. Thus, for a particular magnitude D of block 220 desired dose modification, the adjustments i to the intensity values of the dosechange beamlets 164 may be determined in accordance with this approximate relationship.
Restrictions may be applied to the block 223 intensity changes to the dosechange beamlets 164. Such restrictions may be related to practical considerations—e.g. to more accurately reflect dose distributions which are achievable in practice. By way of nonlimiting example, such intensitychange restrictions may include:
 the intensities of the dosechange beamlets 164 (after the block 223 changes) must be greater than a minimum threshold (e.g. negative intensities are not permitted);
 the intensities of the dosechange beamlets 164 (after the block 223 changes) must be less than a maximum threshold;
 the intensities of the dosechange beamlets 164 (after the block 223 changes) must be controlled such that spatial variations are limited over the twodimensional extent of a corresponding dose distribution 165 or a dose distribution grid 162 (see
FIG. 6 ). For example, such a restriction may limit the maximum (magnitude or percentage) change in intensity between immediately adjacent (or some range of adjacent) beamlets 164 in a particular dose distribution 165; and/or  the like.
If it is determined, in block 223, that a prospective intensity change to one or more dosechange beamlets 164 would violate an intensitychange restriction, then a variety of strategies may be employed in block 223 to overcome such a violation. By way of nonlimiting example, such strategies may involve:
 rejecting the block 220 desired dose modification;
 adjusting the block 220 desired dose modification magnitude;
 the one or more beams 159 corresponding to dosechange beamlets 164 resulting in intensity limit violations may be omitted from subsequent dose estimation (discussed further below);
 the intensity changes to one or more beamlets 164 associated with secondary dose modification voxels may be omitted or the marginal region of secondary dose modification voxels may be reduced;
 the magnitudes of one or more intensity changes applied to one or more corresponding dosechange beamlets 164 may be modified until the intensitychange restriction is no longer violated;
 a combination of two or more of the above strategies; and/or
 the like.
Once it is assured that no intensitychange restrictions are violated by the prospective block 223 intensity changes, then the intensities of the dosechange beamlets 164 are modified as discussed above.
In some embodiments, block 221 may involve identifying secondary dosechange beamlets 164 and block 223 may involve adjusting the intensity values of secondary dosechange beamlets 164 in addition to the identification and intensity value adjustment of the primary dosechange beamlets 164 in accordance with the techniques discussed above. Secondary dosechange beamlets 164 may be identified in marginal regions proximate to the primary dosechange beamlets 164. By way of nonlimiting example, such secondary dosechange beamlets 164 could be identified to be in a marginal region less than or equal to a threshold number of beamlets away from a primary dosechange beamlet. The threshold number of beamlets that define the marginal region may be operatorconfigurable, may be a system parameter which may be determined by one or more suitable calibration procedures, may be a system parameter which may be determined based on empirical testing and/or data and/or the like.
The block 223 intensity adjustments to the secondary dosechange beamlets 164 may be less than corresponding adjustments to the primary dosechange beamlets 164. For example, the block 223 intensity adjustments to secondary dosechange beamlets 164 may be a fraction a of the intensity adjustments to the primary dosechange beamlets (where 0<=a<=1). This fraction may be a function of the distance between a given secondary dosechange beamlet and its corresponding primary dosechange beamlet—e.g. within a marginal region of 3 beamlets around a primary dosechange beamlet, the fraction a may be relatively high for the secondary dosechange beamlets that are nearest neighbors to the primary dosechange beamlet; lower for the secondary dosechange beamlets that are spaced by one beamlet from the primary dosechange beamlet; and lowest for the secondary dosechange beamlets that are spaced by two beamlets from the primary dosechange beamlet.
In other respects (e.g. for the purposes of other procedures involved in the methods and systems described herein), secondary dosechange beamlets 164 may be treated, for the most part, in the same manner as primary dosechange beamlets 164. Accordingly, both primary dosechange beamlets and secondary dosechange beamlets may be referred to simply as dosechange beamlets.
Method 18 (
The block 224 estimation of the changes in the dose distribution may involve the use of known dose estimation techniques such as, Monte Carlo, collapsed cone convolution, pencil beam, anisotropic analytical algorithm, Boltzman equation solvers and/or the like. The block 224 estimation of dose distribution may involve one or more of the rapid dose distribution estimation techniques described herein. It is generally desirable that the block 224 dose estimation technique be computationally efficient. Block 224 may involve using dose estimation techniques that are able to update dose estimates resulting from intensity changes to particular dosechange beamlets 164 along individual dosechange ray lines 163 which may be more computationally efficient. The resulting changes in the dose estimates may then be added (or subtracted in the case of a dose reduction) to the existing dose distribution estimate. In this regard, block 224 need only involve updating the achievable dose distribution for the dosechange beamlets 164 whose intensities were modified in block 223—i.e. it is not necessary to recalculate the entire dose distribution estimate in block 223. Typically, the number of dosechange beamlets 164 having their intensities modified in block 223 will be a relatively small subset of the beamlets 164 associated with a given beam 159. In accordance with some dose estimation techniques, the block 224 dose estimation update may therefore consume a relatively small amount of computational resources compared to a full dose distribution estimate.
Block 52 of the illustrated embodiment involves identifying the dosechange beamlets 164 which correspond to the current dose modification voxel. Such dosechange beamlets 164 may be those identified in block 221 (
Method 50 of the illustrated embodiment then proceeds to block 58 which involves summing the block 56 estimated dose contributions for all of the dosechange beamlets 164 corresponding to the current dose modification voxel to obtain an estimated dose distribution update for the current dose modification voxel. In block 60, the block 58 dose distribution update for the current dose modification voxel may optionally be scaled to provide a scaled dose update grid corresponding to the current dose modification voxel. The term dose update grid may be used interchangeably with dose update distribution or dose modification distribution. As discussed above, an approximate relationship may be established (e.g. in block 47 of initialization method 20) between the dose change to a dose modification voxel and the corresponding intensity changes to the dosechange beamlets 164 and, in accordance with this approximate relationship, changes to the intensities of the dosechange beamlets 164 may be established in block 223. However, this relationship is only approximate. Consequently, the block 58 sum of the dose change contributions for each of the dosechange beamlets 164 may not yield the desired magnitude (e.g. the block 220 desired magnitude) of dose change to the dose modification voxel. In such cases, the block 58 sum of the dose change contributions for each of the dosechange beamlets 164 may be scaled in block 60 to obtain a scaled dose update grid which achieves the desired magnitude of dose change (e.g. the block 220 desired dose change magnitude). The block 60 scaling of the dose update grid may also be accompanied by corresponding scaling to the intensities of the dosechange beamlets. The dosechange beamlets may be scaled by a similar factor. For example, if the block 60 scaling of the dose change grid involves a scaling factor s, then the intensities of the dosechange beamlets may be scaled by the same scaling factor s.
Since there may be an approximate relationship established between the dose change to the dose modification voxel and the corresponding intensity changes to the dosechange beamlets 164, the block 60 scaling may be minimal. In some embodiments, block 60 scaling is not used. In some instances, even where block 60 scaling is used, it may be desirable to limit the amount of block 60 scaling. For example, it may be undesirable to scale in a manner which may result in one or more beamlet intensities that violate beam restrictions, such as any of the beam restrictions discussed above. In some cases (e.g. because scaling is not used or is limited in amount), it may not be possible to achieve the desired dose change magnitude (e.g. the block 220 dose modification magnitude) in the dose modification voxel. This circumstance is permissible.
The block 60 doseupdate grid 234 may represent an amount of dose to add to (or subtract from) an overall achievable dose distribution in block 64. As discussed above, an initial overall achievable dose distribution may be determined in block 46 (
It will be appreciated by those skilled in the art that doseestimation update method 50 of the
Method 50 of the illustrated embodiment involves obtaining individual intensity changes to dosechange beamlets (in block 54), estimating dose changes (in blocks 56, 58, 60, 62) and then adding (or subtracting) dose changes to the existing dose distribution (in block 62) to obtain the block 224 (
In one example of such an embodiment, the parts of an existing dose distribution which could be discarded comprise the parts of the existing dose distribution contributed by the previous values of dosechange beamlets 164, in which case the estimated replacement parts of the updated dose distribution estimate would be the dose contributions from the new intensity values of the dosechange beamlets 164. In another example of such an embodiment, the parts of an existing dose distribution which could be discarded comprise the parts of the existing dose distribution contributed by the intensity distributions 165 of beams 159 (e.g. intensity distributions 165 having one or more dosechange beamlets 164), in which case the replacement parts of the updated dose distribution estimate would be the dose contributions from the modified intensity distributions 165 of beams 159 (e.g. intensity distributions 165 modified by updated intensity values for one or more dosechange beamlets 164). In another example of such an embodiment, the entire existing dose distribution could be discarded and a replacement dose distribution could be estimated based on dose contributions from all of the updated intensity distributions 165 of all of the beams 159. These techniques of discarding parts of existing dose distributions and estimating replacement parts for the updated dose distribution estimate may be particularly useful where they are used in conjunction with the convolution technique for rapid estimation of achievable dose distribution which is discussed further below.
At the conclusion of method 50 and/or block 224 (
In some embodiments, method 18 and/or portions thereof may be performed in loops. For example, at the conclusion of blocks 224 and/or 225, method 18 may loop back to block 220 to permit additional dose changes (e.g. operator input of additional desired dose changes). In some embodiments, dose estimation updates (block 224) and/or updates to dose quality metrics (block 225) may be performed and/or displayed periodically. The periods between such computation and/or display updates (which need not be temporal periods) may be defined using a variety of techniques. By way of nonlimiting example, updates may be performed:
 each time that a block 220 desired dose modification is requested or otherwise determined;
 after multiple block 220 desired dose modifications are requested or otherwise determined;
 after a time interval;
 if several block 220 desired dose modifications are requested, it may be desirable to display a result which includes some subset of the requested dose modifications;
 after an achievable dose distribution and/or dose quality metric has changed by a threshold amount;
 a combination of the above (e.g. after a time interval or a threshold number of desired dose modifications are requested); and/or
 the like.
Some of the procedures of method 18 may overlap with one another. For example, an operator may request multiple desired dose modifications (block 220) prior to the completion of the remainder of method 18. As desired dose modifications are communicated by the operator (or otherwise obtained in block 220), the rest of method 18 may be carried out, so that modifications may be continuously applied and the achievable dose distribution may be continuously updated. It may occur that one or more further desired dose modifications are requested prior to completion of the rest of method 18 for a previous desired dose modification update. Further desired dose modifications may be places in a queue so that, once method 18 is completed for a particular desired dose modification, method 18 may be completed for the next desired dose modification in the queue. In this way all desired dose modification request changes will eventually be processed. In other embodiments, further desired dose modification requests may only be permitted after method 18 has completed for a previous desired dose modification request. Additional example schemes for addressing further desired dose modification requests while method 18 is being carried out for previous desired dose modification requests include, but are not limited to, rejection of every 2nd, 3rd or N^{th }request (where N is an integer) while the remaining requests are placed in a queue.
During method 18 (e.g. as a part of block 224 and/or method 50 (
Restrictions (e.g. thresholds) may be specified for the dose distribution estimate itself and/or for one or more dose quality metrics. As part of method 18, proposed updates to the achievable dose distribution and/or proposed updates to one or more dose quality metrics may be evaluated with respect to one or more such restrictions. In the event that such a restriction is violated, a variety of actions may be taken. By way of nonlimiting example:
 the block 220 desired dose modification request corresponding to the restriction violation may be rejected;
 the magnitude of the block 220 desired dose modification request may be adjusted so that the restriction is no longer violated;
 one or more additional changes to the block 220 desired dose modification request may be made in attempt to overcome the violation of the restriction; and/or
 the like.
In the event that one or more additional changes to the block 220 desired dose modification request are changes in attempt to overcome violation of the restriction, the following exemplary procedure may be used:
 restriction violating coordinates (e.g. voxels) corresponding to the voxels in the dose distribution that violate the restriction may be determined. Such restriction violating coordinates may be determined in a manner similar to the determination of the block 220 dose modification coordinates (e.g. for dose quality metrics).
 dose modifications may be applied to the restriction violating coordinates and a resulting dose distribution estimates may be obtained in accordance with method 18 of
FIG. 4B .  after the new dose distribution estimates are calculated for the changes to the restriction violating coordinates, the new dose distribution estimates may be evaluated for restriction violations.
 If one or more restrictions remains violated, then the procedure can be repeated until there are no longer restriction violations.
At the conclusion of method 18 (any loops or any portions thereof), method 18 yields an achievable dose distribution and/or an estimated dose quality metric. Preferably, such achievable dose distribution and/or estimated dose quality metric will meet the operator'"'"'s treatment objectives. The achievable dose distribution and/or dose quality metrics may be output for use by another method or system. Such methods or systems may comprise, for example, a computerized database, a treatment plan optimization system, a radiation delivery apparatus and/or as an input to any other system or device used in radiation treatment. The corresponding beamlet intensities and/or dose restrictions associated with the
By way of nonlimiting example, as discussed above, method 18 of
If iterative optimization is used in block 143, one or more outputs from the block 142 (method 18) generation and manipulation of estimated dose may be used to aid in the derivation of the block 143 radiation delivery parameters. By way of nonlimiting example:
 the block 224 achievable dose distribution and/or block 225 estimated dose quality metric(s) may be used in block 143 to define the optimization objectives (e.g. cost function or the like) of the optimization process;
 the beamlet intensities determined in block 223 and corresponding to the final block 224 achievable dose distribution may be used to determine the beam intensity required of the radiation delivery apparatus;
 other metrics derived from the block 224 achievable dose distribution may be used to define optimization objectives (e.g. a cost function or the like) of the block 143 optimization process; and/or
 the like.
Block 143 does not necessarily require the performance of an optimization process. In some embodiments, the output(s) of the block 142 generation and manipulation of achievable dose distributions (e.g. block 224 achievable dose distributions, block 225 dose quality metrics and/or block 223 beamlet intensities) may lead be convertible directly to radiation delivery parameters of sufficient accuracy. Such direct derivation of radiation delivery parameters (i.e. without iterative optimization) in block 143 may occur, for example, where the block 224 achievable dose distributions are calculated according to a sufficiently accurate estimation technique and various restrictions (e.g. on beamlet intensities and/or dose estimates) are sufficiently robust.
Radiation treatment method 16 may then proceed to block 144, where the block 143 radiation delivery parameters may be transferred to a radiation delivery apparatus. In block 145, a controller associated with the radiation delivery apparatus (equipped with the radiation delivery parameters) may then cause the radiation delivery apparatus to deliver radiation to the subject. The radiation received by the subject is preferably similar to the achievable dose distribution output predicted in blocks 142 (
Systems and methods according to various embodiments described herein involve estimating dose distributions based on one or more beamlet intensities. Nonlimiting examples of estimating dose distributions include: the block 218 initialization procedure of method 18 (
One aspect of the invention provides different methods for estimation of achievable dose distributions. Such methods may be used to perform the dose distribution estimation procedures in any of the other methods and systems described herein. In some embodiments, methods for estimating achievable dose distributions are provided which are relatively rapid in comparison to currently available dose estimation techniques, such as Monte Carlo, collapsed cone convolution, pencil beam, anisotropic analytical algorithm, Boltzman equation solvers and/or the like. Dose distributions estimated in accordance with the inventive methods described herein may be referred to as rapid dose distribution estimates to contrast them with traditional dose distribution estimates obtained using known techniques. Methods of estimating achievable dose distributions according to various embodiments of the invention may involve simplifications based on ray lines 163 (and corresponding beamlets 164 of intensity distributions 165) emanating from radiation source 161 and knowledge of how such ray lines 163 interact with calculation grids which are used to map threedimensional space in subject 160 (see
 using simplified models of radiation scatter and/or radiation transport as compared to traditional dose distribution estimation techniques;
 omitting the effects of inhomogeneous subject density;
 omitting attenuation of beams 159 as a function of depth in subject 160;
 ignoring the distance from the radiation source to the subject 160;
 a combination of the above; and/or
 the like.
Given a set of beams 159 with known intensity distributions 165, rapid estimates of achievable dose distributions (and corresponding dose quality metrics) determined in accordance with some of the inventive methods described herein may not yield the same achievable dose distributions (and corresponding dose quality metrics) as traditional dose estimation methods. However, where suitable limits are imposed on beams 159, corresponding intensity distributions 165 and the intensities of individual beamlets 164, rapid dose distribution estimation techniques described herein may yield dose distribution estimates that are reasonably close to those that are physically deliverable. Suitable examples of limitations on beams 159, corresponding intensity distributions 165 and the intensities of individual beamlets 164 are described above. In some embodiments, where rapid estimation of achievable dose distributions are used during the process of treatment planning and/or delivery (e.g. method 16 of
Rapid dose distribution estimation method 70 commences in block 71 which involves an inquiry into whether there are more beams to be considered in the rapid dose distribution estimation. In the first loop through method 70, the block 71 inquiry will typically be positive (block 71 YES update path) and method 70 will proceed to block 72. Before proceeding to block 72, block 71 may also involve selecting one beam to be the current beam for this iteration of the method 70 rapid dose distribution estimation loop. In block 72, method 70 involves performing a convolution operation which comprises convolving the twodimensional intensity distribution i(x,y) associated with the current beam with a twodimensional dose estimate kernel k(x,y) to obtain a twodimensional convolved intensity distribution f(x,y). The coordinates x,y may be defined in the plane of grid 162 of the intensity distribution 165 (see
The dose estimate kernel k(x,y) may be intended to approximate the amount of radiation scatter and energy transport resulting from radiation interacting with tissue. In some embodiments, the dose estimation kernel k(x,y) comprises a point spread function.
In some embodiments, the dose estimation kernel k(x,y) comprises a linear combination of a plurality of point spread functions. In one exemplary embodiment, the dose estimate kernel k(x,y) comprises a linear combination of one or more 2dimensional Gaussian functions:
k(x,y)=A_{1}e^{−(x}^{2}^{−y}^{2}^{)/σ}^{1}^{2}+A_{2}e^{−(x}^{2}^{−y}^{2}^{)/σ}^{2}^{2}+ . . . +A_{n}e^{−(x}^{2}^{−y}^{2}^{)/σ}^{n}^{2} (1)
where A_{i }are magnitude variables of the various Gaussian functions and σ_{i }are variables representative of the radial spread of the various Gaussian functions. The variables A_{i }and σ_{i }may be operatorconfigurable, may be system parameters which may be determined by one or more suitable calibration procedures, may be system parameters which may be determined based on empirical testing and/or data and/or the like. The variables A_{i }and σ_{i }may depend on the type of radiation. It will be appreciated by those skilled in the art that the equation (1) point spread function is merely an example point spread function and that dose estimation kernel k(x,y) may comprise a variety of other point spread functions and/or linear combinations of point spread functions. In some embodiments, the dose estimation kernel k(x,y) (or one or more parameters thereof) may be experimentally determined (e.g. from calibration type measurements and/or the like). In some embodiments, the dose estimation kernel k(x,y) (and/or its parameters) may be stored in accessible memory (e.g. in a look up table or the like).
Convolution operations can be computationally intensive and can consume relatively large amounts of processing resources. To reduce this burden on computational resources, the block 72 convolution may involve converting the twodimensional intensity distribution i(x,y) and the twodimensional dose estimate kernel k(x,y) to the Fourier domain. Advantageously, a convolution operation in the spatial domain may be implemented as a multiplication operation in the Fourier domain. The twodimensional intensity distribution i(x,y) and the twodimensional dose estimate kernel k(x,y) may be converted to the Fourier domain using any of a wide variety of known computational techniques for performing Fourier transforms (e.g. fast Fourier transforms (FFT) and/or the like). The convolved intensity distribution f(x,y) may therefore be calculated according to:
f(x,y)=IFT[FT[i(x,y)]×FT[k(x,y)]] (2)
where FT[●] and IFT[●] are respectively Fourier transform and inverse Fourier transform operators. It will be appreciated that, in absence of a change to the kernel function k(x,y), the Fourier transform of the kernel function FT[k(x,y)] need only be calculated once and the result may be stored (e.g. in a look up table in accessible memory and/or the like).
Method 70 then proceeds to block 74 which involves a loop for each convolved beamlet 164 in the current beam 159. For each convolved beamlet 164 in the current beam 159: block 76 involves projecting the ray line 163 corresponding to the convolved beamlet 164 into the calculation grid and identifying voxels in the calculation grid which are intersected by the ray line 163 and block 78 involves adding the convolved intensity value for the convolved beamlet 164 to each voxel identified in block 76. Method 70 then loops back to block 71 to determine whether there are more beams to be considered in the method 70 rapid dose estimation. After one or more loops through blocks 7278, the block 71 inquiry will be negative, terminating method 70.
Rapid dose distribution estimation method 70 involves the principal of superposition. As discussed above in connection with dose estimation update method 50, dose estimation techniques which involve the principal of superposition may operate on changes (e.g. determining changes to dose distributions that result from changes to beam and/or beamlet intensities) or on absolute values (e.g. discarding dose contributions from previous values of beam and/or beamlet intensities and estimating new dose contributions based on the new absolute values of the beam and/or beamlet intensities).
The abovedescribed method 70 for rapid dose distribution estimation may be augmented by incorporating an estimate for attenuation of radiation as it passes through the subject. For this purpose, in some embodiments, an additional attenuation function a(d) may be applied to account for such attenuation. Such an attenuation function a(d) may cause the block 72 convolved intensity values to decrease with distance d along the ray lines 163 which they are projected in block 76, so that their respective block 78 dose contributions decrease with distance d along their respective ray lines 163. By way of nonlimiting example, an attenuation function a(d) may be multiplied to the dose contribution values in block 78 before such dose contribution values are added to each voxel intersected by ray lines 163. A variety of different decreasing attenuation functions a(d) are suitable to model this attenuation and the choice of particular attenuation function a(d) may depend on the characteristics of the radiation. In one particular embodiment, an attenuation function a(d) may be provided by an exponential function of the form:
a(d)=B e^{−kd} (3)
where B is a magnitude variable, k is a variable characterizes the rate of attenuation with depth and d represents the distance along a particular ray line 163. In some embodiments, the variable d may represent the depth of penetration into the body of the subject—i.e. there is negligible attenuation prior to the radiation impinging on the body of the subject. The variables B and k may be operatorconfigurable, may be system parameters which may be determined by one or more suitable calibration procedures, may be system parameters which may be determined based on empirical testing and/or data and/or the like. Like the Fourier transform of the kernel function, the values of the attenuation factor (e.g. equation 3) may be precalculated and stored in a look up table or the like.
In some embodiments, beam configurations are contemplated which involve a plurality of beams 159 at a corresponding variety of locations along a trajectory which involves a 360° rotation of the radiation source with respect to the subject. In some embodiments, it is contemplated that the radiation source will move continuously with respect to the subject about the 360° trajectory. In such embodiments, the 360° trajectory can be approximated by a plurality of sample beams. For the purpose of this description, sample beams on a trajectory where it is contemplated that the radiation source will move continuously relative to the subject can be treated in the same manner as discrete beams 159.
An additional property of the 360° trajectory 240 and the corresponding dose update grid 242 is that the same dose update grid may be translated to any desired dose modification voxel. An example of this translatability is shown in
A 360° trajectory 240 is used as an example for illustration purposes. Beam configurations with other trajectories (e.g. other angular rotation ranges and/or other) motions may also be used. Furthermore, it may be desirable to use different angular ranges of beams when permitting manipulation of achievable dose (e.g. in block 142). For angular ranges of beams smaller than 360°, the corresponding angular range of ray lines 163 intersecting a dose distribution coordinate (e.g. a dose modification voxel) will be reduced. In some embodiments, dose update grids for alternative motion ranges may also be precalculated and stored (e.g. in a look up table in an accessible memory) for subsequent retrieval. A library of dose update grids referred to as radial modification dose (RMD) distributions may be stored in accessible memory. Such a library of RMDs may include RMDs for N incremental motion ranges from 0°N°, for example, where N has incremental values going from 0°360°. Specific angular ranges of beams may be determined by subtracting two dose modification distributions stored in the library. For example:
RMD(θφ)=RMD(0φ)−RMD(0θ) (4)
Where θφ indicates an angular range of beams going from angle θ to angle φ. Using a subtraction operation like that of equation (4) reduces the number of actual RMDs that must be stored in memory. A graphical example is shown in
Processor 304 of the illustrated embodiment has access to a computerreadable memory 308 which may house software instructions 310. In other embodiments, processor 304 may obtain instructions 310 from one or more other sources. When executed by processor 304, software instructions 310 may cause processor 304 to perform one of more of the methods described herein (e.g. radiation delivery method 16 (
In still other embodiments, one or more of the methods described herein may be performed by controller 302A (or some other suitable processor) that is part of radiation delivery apparatus 302. In such embodiments, radiation delivery apparatus 302 may comprise (or otherwise have access to) suitable memory which may comprise suitable computer software instructions.
The examples set out below represent nonlimiting examples of methods, systems and various features of methods and systems according to various embodiments. These nonlimiting examples are for illustrative purposes only and are not intended to represent limiting features unless otherwise claimed below.
Setup Initialization
 Import subject image and segmented tissue structure information (e.g. tumor target, healthy tissue structures (e.g. organs), other tissue of interest, target prescription)
 Select dose estimate kernel (e.g. point spread function (PSF)) based on beam energy and associated scattering characteristic(s)
 Calculate Fast Fourier Transform (FFT) of PSF and store in accessible memory for later use
 Create voxel matrix (calculation grid) overlaid with image data from subject including tissue structure information
 Identify voxels located within any structure(s) of interest
 identify voxels partially within a structure (e.g. at the edge of the structure) and assign partial volume values (e.g. in a range of 01) to those voxels for the purpose of calculating dose quality metrics based on such voxels
 Select beam configuration (e.g. number, location and direction of beams)
 Desired beam configuration may include a number of fixed beams (e.g. 5 to 9 beams or other numbers of fixed beams)
 If desired beam configuration involves continuously moving beams, then a number of beam samples (e.g. 10 or more beam samples) along the trajectory may be used to represent dynamic source motion
 Initialize beamlet intensities
 Calculate estimated dose distribution resulting from initial beamlet intensities
 Calculate initial dose volume histogram (DVH) (and/or other initial dose quality metric(s))
 Display initial DVH(s) and/or initial dose distribution overlaid with image data from subject including tissue structure information
 dose may be represented via cross sectional view(s) of the image data
 dose can be represented as a color wash
 dose can be represented as an isoline plot (line contours of constant dose)
 dose can be represented using any other suitable display technique
 threedimensional dose display (socalled “dose cloud” displays) could additionally or alternatively be used
 dose may be represented via cross sectional view(s) of the image data
Beamlet Initialization
 Create twodimensional beamlet matrix for each beam
 For each target structure:
 Target structure is projected onto the beamlet matrix
 A margin may be added to the projection to ensure proper coverage of the target structure by the resulting dose distribution
 Assign an intensity value to beamlets corresponding to target structure.
Such intensity value may be correlated with (e.g. proportional to) the prescription dose for the target structure
 Optional:
 For target structures that overlap in a given beam projection, weight the beamlet intensities in the overlapping area with the overlapping target structure that has the highest prescription dose; or
 For target structures that overlap in a given beam projection, weight the beamlet intensities in the overlapping area with the overlapping target that has the lowest prescription dose
 Optional:
Rapid Estimation of Achievable Dose Distribution
 Estimate dose contribution—For each beam:
 Convolve beamlet intensity matrix by the dose estimate kernel (e.g. PSF)
 Perform convolution by: determining the FFT of the beamlet matrix; multiplying by the precalculated FFT of the dose estimate kernel; and determining the Inverse FFT of the result
 Convolve beamlet intensity matrix by the dose estimate kernel (e.g. PSF)
 Determine intersection of ray lines from beamlet matrix with voxels—For each voxel where the dose has not been previously determined:
 For each beam:
 Determine the location where any ray line passing through the voxel intersects with the beamlet matrix
 Store the beamlet intersection location in accessible memory for use in subsequent dose calculations (so that the beamlet intersection location only needs to be determined once)
 Achievable dose estimation may be limited to voxels of interest. For example:
 For DVH calculation, achievable dose estimation may be limited to voxels within target and healthy tissue structure(s)
 In the case where a dose plane cross section is being displayed, achievable dose estimation may be limited to voxels in that plane
 For each beam:
 Calculate total dose—For each voxel:
 Extract the dose contribution—For each beam:
 Extract the convolved intensity value from the convolved intensity distribution at the location of the previously determined beamlet matrix intersection
 Sum the contribution from each beam
 Extract the dose contribution—For each beam:
 Estimate dose contribution—For each beam:
Permit Dose Volume Histogram Manipulation
 Operator selects a point on a DVH curve of a particular structure and requests a change of the curve shape at that point
 Identify the dose and volume coordinate values of the selected point (D_selected, V_selected)
 The request may be communicated by selecting the DVH with a mouse or similar computer pointing device click and then moving the mouse or similar computer pointing device
 movement left and/or down indicates may indicate a dose reduction (desired dose change is negative)
 movement right and/or up may indicate a dose increase (desired dose change is positive)
 Determining a magnitude of the desired dose change may involve, for example:
 A fraction of the D_selected value
 A fraction of the maximum dose for that structure
 A fraction of a prescription dose assigned to that structure
 A fraction of the maximum prescription dose assigned to all structures
 A fraction correlated with (e.g. proportional to) the amount of mouse or similar computer pointing device movement
 A quantity set by the operator
 A fixed quantity which may be an operatorconfigurable parameter or may be a “hard coded” constant
 A combination of the above
 For the selected structure, voxels having dose values close to D_selected may be identified as dose modification voxels
 Voxels may be identified as dose modification voxels if the voxels have values falling within D_selected+/−Δ
 Δ may be a fixed value (which may be operatorselectable) or a fraction of D_selected (which may be an operatorselectable fraction)
 If no voxels are identified to be dose modification voxels then Δ may be expanded and voxels may be reidentified
 This process repeats until at least one voxel is identified as a dose modification voxel
 If a large number of voxels (e.g. all voxels for that structure or a number of voxels greater than a threshold number or a threshold percentage of the voxels in a structure) are identified to be dose modification voxels, then A may be reduced and voxels may be reidentified
 In some embodiments, all voxels inside the structure for which the DVH is changed may be identified as dose modification voxels
 Voxels may be identified as dose modification voxels if the voxels have values falling within D_selected+/−Δ
 Optional: identification of secondary dose modification voxels in marginal region around primary dose modification voxels and assignment of suitable magnitude of dose modification to secondary dose modification voxels
 Dose modification is performed for voxels identified as dose modification voxels (see Dose Modification below)
 Once Dose Modification is performed, DVH is recalculated based on new estimated dose
 Check whether any dose restrictions are violated (see Dose Manipulation Under Dose Restrictions)
 Display is updated with new DVH
 Operator selects a point on a DVH curve of a particular structure and requests a change of the curve shape at that point
Dose Modification
 For each voxel identified to be a dose modification voxel (Vox_identified) and for each beam:
 Determine the location(s) where ray line(s) passing through Vox_identified intersect with the beamlet matrix
 Modify intensity values of beamlet(s) at intersection point(s) (dosechange beamlets) using a default quantity
 The default quantity may be correlated with (e.g. proportional to) the magnitude of the desired dose change and may also depend on the number of dosechange beamlets who ray lines intersect Vox_identified
 Optional: identify and determine intensity value adjustment for secondary dosechange beamlets in a marginal region around primary dosechange beamlets
 Modified beamlets are evaluated for possible intensity restriction violations. By way of nonlimiting example:
 Intensity values that are less than a minimum threshold
 Corrected by increasing violating beamlet intensities to minimum threshold
 Intensity values greater than a maximum threshold
 Corrected by decreasing violating beamlet intensities to maximum threshold
 Variations in intensity over the two dimensional extent of the beamlet matrix may also be limited (e.g. a maximum variation of a fixed percentage from one beamlet to an adjacent beamlet)
 Corrected by increasing or decreasing the intensities of modified beamlets so that variation restriction is no longer violated
 Intensity values that are less than a minimum threshold
 Dose is estimated at voxels for all beams using Rapid Estimation of Achievable Dose Distribution
 Dose may be calculated for only a subset of voxels. For example:
 Voxels of the selected structure only
 Voxels of all structures
 Voxels of all structures plus currently displayed cross sectional plane
 Dose may be calculated for only a subset of voxels. For example:
 Estimated dose for each dose modification voxel (Vox_identified) is compared to dose requested by operator:
 Determine difference between the mean estimated dose and mean operatorrequested dose for all dose modification voxels Vox_identified
 Rescale magnitude of intensity modification for all dosechange beamlets is rescaled so that mean dose for each dose modification voxel (Vox_identified) is equal to mean operatorrequested dose requested by operator unless the rescaling of the dosechange beamlets results in a violation of the beamlet intensity restrictions
 For each voxel identified to be a dose modification voxel (Vox_identified) and for each beam:
Dose Manipulation Under Dose Restrictions
 Restrictions on the shape of DVH curves (or other restrictions on the dose distribution itself or on other dose quality metrics) may be designated by the operator or may be otherwise incorporated into the procedure (e.g. as global parameters of the system/method). For example:
 No voxels within a structure can be below a specified dose threshold
 No voxels within a structure can exceed a specified dose threshold
 A specified percentage of volume of a structure cannot exceed a specified dose threshold
 A specified percentage of volume of a structure cannot exceed a specified dose threshold
 When a dose manipulation is performed (through DVH or otherwise) the dose distribution is evaluated for potential dose restriction violations
 Violations are detected by comparing the estimated dose distribution and/or dose quality metrics (e.g. DVH) with the dose restrictions
 Voxels contributing to the violation are identified
 The magnitude of dose change required for each violating voxel to satisfy the restriction is calculated
 Perform Dose Modification (see above) using the violating voxels and voxel dose change requirements as inputs
 Repeat as necessary until there are no further dose restriction violations
 Restrictions on the shape of DVH curves (or other restrictions on the dose distribution itself or on other dose quality metrics) may be designated by the operator or may be otherwise incorporated into the procedure (e.g. as global parameters of the system/method). For example:
The following nonlimiting example is meant to provide further understanding of how various aspects and features of the invention could be used in practice.
The beam configuration used in this particular example comprises a 360° degree rotation of the radiation source with respect to the subject. Such a beam configuration could be implemented using the radiation delivery apparatus of
Once the initial dose estimate is established, manipulation of the estimated dose distribution may be permitted (e.g. by an operator). The operator may desire to establish adequate volumetric coverage of target structure 281 with the prescription dose. This may be accomplished by increasing the minimum dose to target tissue structure 281 through manipulation of the DVH.
Now that the operator has established an acceptable dose distribution for target structure 281, it may be desirable to place restrictions on DVH manipulation so that the acceptable dose distribution is maintained during further dose modification. For example, such a restriction could be that 100% of the target structure 281 should receive 57 Gy and 0% of the target structure 281 should receive more than 63 Gy. The operator may then proceed with modifying the DVH 315 corresponding to healthy tissue structure 282 by selecting a point on DVH 315 and dragging the mouse or similar computer pointing device to the left 317 (for dose reduction). Once the dose modification voxels corresponding to this desired DVH modification have been determined, a dose modification is applied to each dose modification voxel. In this case restrictions on the intensity changes necessitate the use of a smaller number of beamlets.
With the healthy tissue dose reduction illustrated in
Further dose modifications may be imposed by the operator by repeating in a similar fashion the process described above. For example, further reductions in the healthy tissue structure dose may be applied under the constraints 321 and 314 (
After the dose manipulation is complete the radiation delivery parameters for the beam configuration may be determined. The radiation delivery parameters may then be transferred to the control system and computer of the radiation delivery apparatus. The radiation may then be delivered to the subject, thereby delivering a dose distribution in the subject substantially similar to that derived from the dose manipulation.
Embodiments of the present invention include various operations, which are described herein. These operations may be performed by hardware components, software, firmware, or a combination thereof.
Certain embodiments may be implemented as a computer program product that may include instructions stored on a machinereadable medium. These instructions may be used to program a generalpurpose or specialpurpose processor to perform the described operations. A machinereadable medium includes any mechanism for storing information in a form (e.g., software, processing application) readable by a machine (e.g., a computer). The machinereadable medium may include, but is not limited to, magnetic storage medium (e.g., floppy diskette); optical storage medium (e.g., CDROM); magnetooptical storage medium; readonly memory (ROM); randomaccess memory (RAM); erasable programmable memory (e.g., EPROM and EEPROM); flash memory; or another type of medium suitable for storing electronic instructions.
Additionally, some embodiments may be practiced in distributed computing environments where the machinereadable medium is stored on and/or executed by more than one computer system. In addition, the information transferred between computer systems may either be pulled or pushed across the communication medium connecting the computer systems.
Computer processing components used in implementation of various embodiments of the invention include one or more generalpurpose processing devices such as a microprocessor or central processing unit, a controller, graphical processing unit (GPU), cell computer, or the like. Alternatively, such digital processing components may include one or more specialpurpose processing devices such as a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or the like. In particular embodiments, for example, the digital processing device may be a network processor having multiple processors including a core unit and multiple microengines. Additionally, the digital processing device may include any combination of generalpurpose processing device(s) and specialpurpose processing device(s).
Although the operations of the method(s) herein are shown and described in a particular order, the order of the operations of each method may be altered so that certain operations may be performed in an inverse order or so that certain operation may be performed, at least in part, concurrently with other operations. In another embodiment, instructions or suboperations of distinct operations may be in an intermittent and/or alternating manner.
Where a component (e.g. a software module, processor, assembly, device, circuit, etc.) is referred to above, unless otherwise indicated, reference to that component (including a reference to a “means”) should be interpreted as including as equivalents of that component any component which performs the function of the described component (i.e. that is functionally equivalent), including components which are not structurally equivalent to the disclosed structure which performs the function in the illustrated exemplary embodiments of the invention.
As will be apparent to those skilled in the art in light of the foregoing disclosure, many alterations and modifications are possible in the practice of this invention without departing from the spirit or scope thereof. For example:
 The abovedescribed embodiments describe several features which may be characteristics of each beam 159. By way of nonlimiting example, the procedures of block 7278 for each loop of method 70 (
FIG. 13 ) are performed for each beam 159 andFIG. 12 shows a number of beams 159. In some embodiments, radiation delivery may involve continuous movement of the radiation source with respect to the subject over a given trajectory. In such embodiments, the trajectory of movement of the radiation source with respect to the subject may be described by a number of discrete sample beams from along trajectory. Such sample beams may be sampled sufficiently closely to one another such that the sum of the dose contributions of the sample beams is representative of the sum of the dose contributions from the continuously moving radiation source. Such sample beams may be used in a manner similar to that described herein for conventional discrete beams and unless specifically described as being discrete, any references to beam(s) herein should be understood to include sample beams.  Referring to
FIG. 4A , it will be appreciated by those skilled in the art that the number of beam 159 is not required to be the same in blocks 142 and 143. For example, it can be computationally more efficient to have a relatively small number of beams 159 when permitting manipulation of the estimated dose in block 142. Then, in block 143, a larger number of beams 159 can be used to accurately determine the radiation delivery parameters. In some embodiments, the number and/or location of beams 159 may be one of the variables that is permitted to change in a block 143 optimization process.  The discussion presented above suggests that there is a one to one relationship between beamlets 164 and ray lines 163—i.e. each beamlet 164 has a corresponding ray line 163. Such ray lines 163 may pass through the center of their corresponding beamlets 164. Using such ray lines 163, a relationship can be established between particular beamlets and particular voxels—e.g. dosechange beamlets can be identified as those beamlets that have corresponding ray lines that project through a dose modification voxel (see discussion of block 221 above). It will be appreciated, however, that beamlets actually have a twodimensional shape. In some embodiments, it may be desirable to model this two dimensional beamlet shape by accounting for the impact of a beamlet on voxels other than merely the particular voxels through which its corresponding ray line passes. For example, in some rapid dose estimation embodiments (e.g. in some embodiments of method 70 of
FIG. 13 ) the ray lines associated with particular convolved intensity values may impinge on particular voxels at locations away from the centers of the particular voxels and the dose contributions from those convolved intensity values may be fractionally divided as between the particular voxels and one or more of their neighboring voxels. The fractional division of dose contributions between the particular voxels and neighboring voxels may depend on the location where the ray line impinges on the particular voxels. For example, most of the fractional dose contribution may be added to a particular voxel intersected by a ray line and the remaining dose contribution may be divided between the neighboring voxels, with neighboring voxels relatively close to the location of intersection receiving a greater dose contribution than neighboring voxels that are relatively far from the location of intersection.  In some rapid dose estimation embodiments (e.g. in some embodiments of method 70 of
FIG. 13 ), it may be the case that multiple ray lines 163 from a single beam of convolved intensities impinge on a single voxel. In such cases, the dose contribution added to the voxel may be some form of average or interpolation of the dose contributions that would be predicted by each of the individual convolved intensity values.  The discussion presented above suggests that there is a one to one relationship between beamlets 164 and ray lines 163—i.e. each beamlet 164 has a corresponding ray line 163 which may pass through its center. However, in other embodiments, for each beam, ray lines could additionally or alternatively have a one to one relationship with voxels—i.e. each voxel could have a corresponding ray line that extends from the center of the voxel toward the radiation source location for that particular beam. In such embodiments, the block 221 (
FIG. 4B ) process for identification of dosechange beamlets could involve tracing rays from the block 220 dose modification voxels onto the twodimensional grids 162 of beamlets 164—i.e. dosechange beamlets could be identified as those through which a ray line projects. In some embodiments, it may be desirable to model the threedimensional voxel shape by accounting for the impact of a voxel on beamlets other than merely the particular beamlets through which its corresponding ray line passes. For example, in some embodiments of method 18 (FIG. 4B ), the ray lines associated with particular dose modification voxels may impinge on particular beamlets at locations away from the centers of the particular beamlets and the block 223 intensity value adjustments may be fractionally divided as between the particular beamlets and one or more of their neighboring beamlets. The fractional division of intensity adjustments between the particular beamlets and neighboring beamlets may depend on the location where the ray line impinges on the particular beamlets. For example, most of the fractional intensity value adjustment may be made to a particular beamlet intersected by a ray line and the remaining intensity value adjustment may be divided between the neighboring beamlets, with neighboring beamlets relatively close to the location of intersection receiving a greater fraction of the intensity value adjustment than neighboring beamlets that are relatively far from the location of intersection.  In some embodiments, ray lines for each beam may have a one to one association with voxels and ray lines may extend from the center of voxels to the radiation source location for that particular beam. In such embodiments, it may be the case (e.g. in block 221 of method 18 (
FIG. 4B )) that multiple ray lines from multiple voxels impinge on a single beamlet. In such cases, the block 223 intensity value adjustment to the beamlet may be some form of average or interpolation of the intensity value adjustments that would be applied by each of the individual voxels.  One technique for rapid estimation of achievable dose described above involves convolving a twodimensional intensity distribution i(x,y) and the twodimensional dose estimate kernel k(x,y) and then projecting the convolved intensity values of the resultant twodimensional convolved intensity distribution f(x,y) along corresponding ray lines by adding the convolved intensity values as dose contributions to voxels intersected by the ray lines. In some embodiments, it may be desirable to compute convolved intensity distributions f_{1}(x,y), f_{2}(x,y), . . . for a number of different dose estimate kernels k_{1}(x,y), k_{2}(x,y), . . . , where the different dose estimate kernels can be used to model different tissue densities and their corresponding different scattering patterns. Then, when estimating achievable dose for a particular voxel (e.g. when adding convolved intensity values as dose contributions to the particular voxel), the tissue density of the particular voxel p can be used to select convolved intensity values from a convolved intensity distribution f_{p}(x,y) computed using a dose estimate kernel k_{p}(x,y) which corresponds to the tissue density.
 The abovedescribed embodiments describe several features which may be characteristics of each beam 159. By way of nonlimiting example, the procedures of block 7278 for each loop of method 70 (
In the foregoing description, the invention has been described with reference to specific exemplary embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention. The description and drawings are, accordingly, to be regarded in an illustrative sense rather than a restrictive sense. While a number of exemplary aspects and embodiments have been discussed above, those of skill in the art will recognize certain modifications, permutations, additions and subcombinations thereof. It is therefore intended that the following appended claims and claims hereafter introduced are interpreted to include all such modifications, permutations, additions and subcombinations as are within their true spirit and scope.