CN101209368A - Method for patient target region automatically positioning in radiation therapy - Google Patents

Abstract

The invention discloses an automatic positioning method of the target of a patient in radiotherapy, the method makes use of a light source and a flat panel detector to carry out the rotary scanning around the target of the patient, so as to obtain the image data of the target of the patient; the invention applies a virtual PI line reconstruction algorithm to carry out the reconstruction and processing on data to obtain the image of the target of the patient; and the treatment is carried out by comparison results of the input image data and the adjustment of the position of the patient. As the invention adopts a true three-dimensional image reconstruction algorithm which is based on the flat panel detector, the spatial resolution of the true three-dimensional image is isotropic, the invention does not need the interpolation, and can directly realize the processing of the image on the three-dimensional space, thus greatly improving the imaging accuracy and precision and solving the problem that the imaging is not precise due to the interpolation error during the integration process into a quasi three-dimensional image when in automatic positioning of the target of the patient of the existing radiotherapy, and the problem can further affect the precise positioning of focus and the accurate implementation of the treatment plan.

Description

The automatic localized method of patient target region in a kind of radiotherapy

Technical field

The present invention relates to a kind of method of patient location, especially relate to the localized method of a kind of control patient target region.

Background technology

Remote stereotactic radiotherapy mainly is divided into 60Co stereotactic radiotherapy and electron linear accelerator stereotactic radiotherapy two big classes.In radiotherapy process influence radiotherapy in the treatment effect all multifactor, to the determining and be to implement accurately and the key of image-guided radiotherapy of patient's pendulum position, therapeutic community position, also be the key technical problem that emphasis will solve in the radiotherapy apparatus from now on to the control of the motion (the target area motion that the motion of can't help, internal organs motion drive) of patient in the therapeutic process.Accurate heavy dose of radiotherapy, the requirement of this respect is high more.

Along with progress of science and technology, it mainly is the fast imaging development of technology, make the radiotherapy process begin overall process to implementation plan from formulating radiotherapy planning, and the evaluation to performance can both be main information source after the implementing plan with the CT image, the data set with individual character is the basis.The enforcement of these technology has improved the accuracy and the curative effect of oncotherapy greatly, is the developing direction in whole radiotherapy field.Except the human anatomy information that the CT imaging provides, physiological parameter such as body metabolism, blood flow, especially in-house these physiological parameters of tumor target area and adnexa thereof all can accurately be measured, these data provide the bioactive real information based on the tumor of voxel and surrounding tissue thereof, these information are not only to the correct typing of tumor with count for much by stages, to judging inside tumor structure also have very important significance (for example define necrotic area, general oxygen district and highly soak into distinguish etc.).These information also should be concluded among accurately the determining of the selection of oncotherapy center and treatment plan.

What describe in the imaging technique that adopts in the existing radiotherapy such as the U.S.'s tomography radiotherapy technology patent (US6438202) is to use the technology of fault imaging to realize, this technology is divided into tumor focus and surrounding tissue thereof on the different sectional slices (slice), after the information of tomography inside has been asked on average, having lost some information of focus and surrounding tissue thereof can not recover again, section is integrated in the process of quasi-three-dimensional images, because interpolated error is former thereby caused the image inaccuracy, and then influences the accurate location of focus and the accuracy of treatment plan.

Summary of the invention

The invention provides the automatic localized method of patient target region in a kind of radiotherapy, to solve in the existing radiotherapy the automatic localized method of patient target region because of in being integrated into the process of quasi-three-dimensional images, causing the imaging inaccuracy and then influence the accurate location of focus and the accuracy problem of treatment plan owing to the interpolated error reason.

In order to solve above technical problem, the technical scheme that the present invention takes is;

The automatic localized method of patient target region is characterized in that in a kind of radiotherapy, may further comprise the steps:

(1) input image data;

(2) being provided with in therapy equipment can be around the light source and the flat panel detector of the synchronous rotation of patient;

(3) utilize light source and flat panel detector view data, write down corresponding data around patient target region rotation sweep acquisition patient target region;

(4) data are rebuild and handled, obtain the image of patient target region;

(5) view data of the view data of target area and input is compared, if two picture positions do not conform to, execution in step 6 is if two images conform to execution in step 7;

(6) adjust patient's position according to the result of two image comparisons, execution in step 2;

(7) continue treatment.

Data in the described step (3) comprise: light source is to the distance A of detector, back projection data P, the starting point λ of light source scanning track ₁(with the initial angle θ of runing rest ₁Corresponding), the terminal point λ of light source scanning ₂(with the angle θ in the final position of runing rest ₂Corresponding), the scan angle theta of correspondence during light source scanning track starting point ₁, the scan angle theta of correspondence during the light source scanning terminal point ₂

Described step (4) is handled the algorithm for reconstructing that is based on string of a musical instrument section of use to data, promptly when light source from the position 1 when scanning position 2, the track of light source is an arc

Corresponding string of a musical instrument section λ ₁λ ₂The image of the object part that is covered is rebuild with following formula:

$f(x_{\mathrm{\π}},{\mathrm{\λ}}_1,{\mathrm{\λ}}_2)=-\frac{1}{2\mathrm{\π}}\underset{{\mathrm{\λ}}_1}{\overset{{\mathrm{\λ}}_2}{\∫}}\mathrm{d\λ}\frac{A^2(u_d,v_d)}{{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}^2}H\left\{\frac{P^{\′}({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A^2({u_d}^{\′},{v_d}^{\′})}\right\}$

$-\frac{1}{2\mathrm{\π}}{\left[\frac{A(u_d,v_d)}{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}H\right\{\frac{P({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A({u_d}^{\′},{v_d}^{\′})}\left\}\right]}_{{\mathrm{\λ}}_2}^{{\mathrm{\λ}}_1}$

Wherein

With

Be respectively right

With

The Xi Er baud conversion, in the formula Processing is projected in string of a musical instrument section x _{π 1}x _{π 2}On data carry out filtering, then at whole λ ₁λ ₂Integration in the scope, f (x _π, λ ₁, λ ₂) be the grey value profile of rebuilding in the subject image afterwards.

In the described step (5), the view data of target area and the view data of input are compared by the barycenter of target area, determine two image centroids and obtain corresponding position difference.

In the described step (6),,, move carrying patient's therapeutic bed, realize the adjustment of patient location according to the view data of target area and the view data difference relatively of input by control system.

After having adopted technique scheme, employing is based on the true 3D image reconstruction algorithm of flat panel detector, the spatial resolution of true 3-D view is isotropic, do not need to carry out interpolation, can directly realize treatment of picture at three dimensions, therefore the accuracy and the precision of imaging have been improved greatly, solved when patient target region is located automatically in the existing radiotherapy, because of in being integrated into the process of quasi-three-dimensional images, causing the imaging inaccuracy and then influencing the accurate location of focus and the problem of the accurate execution of treatment plan owing to the interpolated error reason.

Description of drawings

Fig. 1 is a radiotherapy imaging device sketch map of the present invention;

Fig. 2 is virtual PI line algorithm for reconstructing sketch map;

Fig. 3 is MLC leaf position figure;

Fig. 4 is that the therapeutic process target area is detected and positioning flow figure automatically;

Fig. 5 is that inverting of therapeutic process dose delivery and treatment plan are revised flow chart automatically;

Fig. 6 is the planar sketch map of incident beam fluence of dosage inversion algorithm definition;

1. light sources among the figure, 2. collimator apparatus, 3. human body, 4. therapeutic bed, 5. detector 6. plane of incidences

The invention will be further described below in conjunction with accompanying drawing.

The specific embodiment

As shown in Figure 1, be used for the device of the process imaging of radiotherapy, comprise: parts such as light source, collimator apparatus, flat panel detector and program control patient's bed are formed; Wherein, light source comprises that the X-ray light source that γ-light source and electron linear accelerator produce is used for radiotherapy or imaging radiation with other, and collimator apparatus comprises multi-leaf collimator, collimator.In remote stereotactic radiotherapy, light source (LS) around the focus rotation, dynamic multi-leaf collimator of programme controlled two dimension (2DMLC) and flat panel detector (PD), form the conformal intensity modulated radiotherapy unit of image-guided (IGIMRT:Imaging GuidedIntensity Modulated Radiotherapy) that integrates with remote radiotheraphy device, when light source and detector during synchronously around therapeutic community's rotation, its track can be in same plane, allow an inclination alpha between the ring of rotation and the vertical plane, the size at α angle, by the radiotherapy planning regulation, the principle of selection is to avoid tissue and responsive internal organs responsive in the human body as far as possible.Light source and detector, also can change the α angle and realize more optimal source and detector rotational trajectory in the rotation of same α angle around therapeutic community.The angular range, theta of rotation also can be selected according to the optimum range of implementing radiotherapy planning, can select arbitrarily in 0 to 360 degree scope.Avoid the sensitive organization of human body or internal organs when following radiotherapy as much as possible, be dispersed in dosage in the normal structure as far as possible, shorten treatment time as far as possible, and satisfy the principle of target area therapeutic dose again, select α and θ angle.Use is based on the algorithm for reconstructing of string of a musical instrument section, and no matter how many α and θ angles be, can realize accurate image reconstruction.Three kinds of possible situations are arranged here, first kind is after the α angle is determined, in whole therapeutic process, no longer change, in fact source and detector still move in annulus, but realize that in the intravital position of people the treatment region imaging is just passable according to tumor, this can be by selecting θ angle (comprising the end position of selecting original position and scanning).As shown in Figure 2, when light source lambda from λ ₁Spot scan is to λ ₂During point, the track of light source is an arc

Corresponding string of a musical instrument section λ ₁λ ₂The image of the object part that is covered can both be rebuild out, and straightway λ ₁λ ₂Rebuild used geometry, algorithm for reconstructing:

$f(x_{\mathrm{\π}},{\mathrm{\λ}}_1,{\mathrm{\λ}}_2)=-\frac{1}{2\mathrm{\π}}\underset{{\mathrm{\λ}}_1}{\overset{{\mathrm{\λ}}_2}{\∫}}\mathrm{d\λ}\frac{A^2(u_d,v_d)}{{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}^2}H\left\{\frac{P^{\′}({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A^2({u_d}^{\′},{v_d}^{\′})}\right\}$

$-\frac{1}{2\mathrm{\π}}{\left[\frac{A(u_d,v_d)}{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}H\right\{\frac{P({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A({u_d}^{\′},{v_d}^{\′})}\left\}\right]}_{{\mathrm{\λ}}_2}^{{\mathrm{\λ}}_1}---\left(1\right)$

Wherein

With

Be respectively right With

The Xi Er baud conversion.A is the distance of light source to detector in the formula, and P is the back projection data, first formula on formula (1) the equation left side

Processing is projected in string of a musical instrument section x _{π 1}x _{π 2}On data carry out filtering, then at whole λ ₁λ ₂Integration in the scope.

The f (x here _π, λ ₁, λ ₂) be the grey value profile of rebuilding in the subject image afterwards, x _π=x _{π 1}x _{π 2}Be the length of string of a musical instrument section, its end points is respectively x _{π 1}And x _{π 2}λ ₁Be the starting point of light source scanning track, the scan angle of its correspondence is θ ₁, λ ₂The terminal point of light source scanning, corresponding scan angle is θ ₂Therefore, θ=| θ ₁-θ ₂|.Formula (1) is pervasive filtering back projection formula.Verified, when light source from λ ₁Spot scan is to λ ₂During point, string of a musical instrument section can be full of fully by the whole object of imaging, and this volume needs not to be enclosed areas, can satisfy the local imaging needs to the patient.

As shown in Figure 3, the shape of 2D multi-leaf collimator (MLC) on some angle θ defined by the shape that light source and the tangent envelope of focus outer contour (for example the outer contour with clinical target area: CTV is as the criterion) are cut on 2DMLC.This shape is before radiotherapy, is stipulated by this patient's radiotherapy treatment planning.Therefore, the control sequence of 2D multi-leaf collimator (MLC) is in the radiotherapy planning implementation, and patient's image and radiotherapy planning satisfy on this position and see that from beam direction (BEV) tumor shape in the past is adaptive the positioning requirements of MLC blade during according to diagnosis.And the shape that 2DMLC constitutes and on this direction treatment target area (PTV) shape of radiotherapy planning definition in full accord.Along with the variation of θ angle, 2DMLC will be according to the method described above motion by blade constitute and this angle on the adaptive shape in oncotherapy target area, this is that the program of control blade movement is according to the radiotherapy planning parameter setting.

As shown in Figure 4, the automatic localized method of patient target region in a kind of radiotherapy may further comprise the steps:

(1) input image data;

(2) being provided with in therapy equipment can be around the light source and the flat panel detector of the synchronous rotation of patient;

(3) utilize light source and flat panel detector view data, write down corresponding data around patient target region rotation sweep acquisition patient target region;

(4) data are rebuild and handled, obtain the image of patient target region;

(5) view data of the view data of target area and input is compared, if two picture positions do not conform to, execution in step 6 is if two images conform to execution in step 7;

(6) adjust patient's position according to the result of two image comparisons, execution in step 2;

(7) continue treatment.

In the step (5), the view data of target area and the view data of input are compared by the barycenter of target area, determine two image centroids and obtain corresponding position difference.

In the step (6),,, move carrying patient's therapeutic bed, realize the adjustment of patient location by control system according to the view data of target area and the view data difference relatively of input.

As shown in Figure 5, the method for inversing patient target region dosage in the invention process process may further comprise the steps:

(1) input image data;

(2) being provided with in therapy equipment can be around the light source and the flat panel detector of the synchronous rotation of patient;

(3) utilize light source and flat panel detector to center on the dose data that the patient target region rotation sweep obtains the data of patient target region image and sees through patient target region, write down corresponding data;

(4) data of patient target region image are rebuild and handled, adjust patient's position data according to result;

(5) dose data that sees through patient target region is handled, obtained the dose distribution data that patient target region is accepted;

(6) the dose distribution data of target area acceptance and the dose distribution data of input are compared, if two data do not conform to, execution in step 7 is if two data conform to execution in step 8;

(7) result according to the comparison of two dose distribution data adjusts the input data;

(8) continue treatment.

In the step (4), the view data of target area and the view data of input are compared by the barycenter of target area, determine two image centroids and obtain corresponding position difference, according to this position difference, adjust patient's position data, by control system, move carrying patient's therapeutic bed, realize the adjustment of patient location.

In the step (6), the dose data that patient target region is accepted and the dose data of input are compared, and determine whether the difference of two dose distribution meets the demands, difference according to dose distribution, by control system, change the shape of 2D multi-leaf collimator, realize the adjustment of the dosage that the target area is accepted.

According to requirements for quality control, before radiotherapy machine is carried out radiotherapy planning, to check whether radiotherapy planning is correct with phantom, this step is by the irradiation phantom, and compare with result that radiotherapy planning calculates under the identical condition of irradiation phantom, if error illustrates that radiotherapy planning is feasible in the range of error that allows, otherwise will revise radiotherapy planning till meeting the demands, this belongs to the process of common radiotherapy planning checking; In the radiotherapy of formal image-guided, at first will be according to the automatic localized method of patient target region in the radiotherapy, patient is positioned, up to patient being moved on to centers such as treatment, this process will dynamically realize in whole radiotherapy process, but correcting the motion of position deviation, is interior among a small circle motion.

In the actual therapeutic process, correcting position deviation in real time needs the time, because in whole therapeutic process, exectorial process is divided into the time period, and in a period of time, though all moving in therapy apparatus and position, target area, quantity of motion can be very not big.Though the dosage field that the motion by patient's bed can the dynamic adaptation radiotherapy planning requires distributes, but because fill order needs the time, deviation between the dosage field that this process still might cause the dosage field distribution of actual execution and plan to determine distributes, this deviation compensates correction by the technical measures such as treatment time of control multi-leaf collimator collimator or control therapy apparatus, this is position deviation data (the Δ x that needs input new, Δ y, Δ z), by Rapid Dose Calculation, at this dosage field distribution bias of MLC distribution compensation after resetting based on multi-leaf collimator.

The dosage refutation process is to calculate the dosage deviation by the dosage inversion algorithm, according to the change of deviation adjustment MLC position, makes that by control station the position of dynamic multi-leaf collimator obtains revising.

On the flat panel detector be with the pixel unit record, through and the X-ray information of human body after interacting, both be used for entering locating information behind the therapeutic bed as patient, also be used for determining because the deviation information of the therapeutic community position that patient's physiological movement drives, and in the intravital dose distribution of inverting patient, proofread and correct because the kinetic dose distribution control information in treatment target area.

In the above-mentioned flow process, every positional information all is that the calculating by pervasive image reconstruction formula (1) obtains true 3-D view, by calculating the centroid position coordinate that obtains tumor, by obtaining the tumor treatment area information cutting apart of image, then by the setting to the parameter of patient's bed realizes from movable pendulum position and location to the therapy system control station; And all are because the deviation between the dose distribution of the dose distribution that causes of position deviation and treatment plan all is owing to cause in shape mismatch between the treatment target area of radiotherapy planning target area and actual execution, after calculating position deviation, reach real-time gauged target by the control parameter that changes radiotherapy machine and multi-leaf collimator (MLC).Localized information still be the gauged information of dosage all be to realize at true three dimensions based on voxel, exist data volume big, the computing complexity needs powerful computer system and realizes implementing to calculate and show.

The information of utilizing imaging to obtain at first realizes treatment position, target area being moved from movable pendulum position and therapeutic process of patient monitored and automatic shift, and therefore the dose error that causes compensates by the dosage inverting.

Real-time positioning information as position, patient target area, can be by the realization of local region of interest fast imaging technology, promptly because formula (1) can be in imaging in any θ angular range, if this angular range right string can to cover tumor region just passable.Its method is at the initial value θ that at first determines the θ angle according to the position of tumor ₁, then according to the requirement that covers tumor region fully, determine the angular range of scanning, determine the θ that scanning stops ₂The angle, by the algorithm for reconstructing of formula (1), the corresponding image of acquisition is by the mass centre and the outer contour of image processing techniques acquisition tumor.As therapeutic community's position offset (Δ x, Δ y, Δ z), and patient's therapeutic community is moved on to Δ x=Δ y=, the position of Δ z=0 with the motion of tumor target area barycenter.Therefore, position offset is in the three dimensions definition, and in the radiotherapy process, these group data are monitored all the time, thereby provide the direction of motion, movement velocity and absolute position shifted amount (the Δ x of center, tumor target area in whole therapeutic process _{I, j, k, t}, Δ y _{I, j, k, t}, Δ z _{I, j, k, t}), the footnote here is illustrated in the position, and (i, j is on position k), in the displacement array of moment t.The monitoring of position be common according to Cartesian coordinates (X, Y, Z) in definition, consistent with customary way in the medical imaging.

The present invention is to the dosage field inversion algorithm of radiotherapy planning definition based on voxel, and the deviation of the dose distribution that causes moving owing to the target area is carried out real-Time Compensation.The implication of compensation is that acquisition is the new dosage field distribution of unit with the voxel, and this dosage field distribution has comprised the result after the correction difference dosage.Method is according to new dosage field distribution requirement, calculating by the radiotherapy planning system, become the control parameter of the control therapy apparatus and the motion of the relevant blade of 2DMLC, setting by new argument and radiotherapy planning calculate the treatment plan that makes new advances in real time or become the radiotherapy planning that satisfies the therapy apparatus current location by the correction to original treatment plan, so the control parameter of sending the bundle parameter and the motion of the relevant blade of 2DMLC of radiotherapy machine all will be upgraded again.Though this process realizes in dynamically, finish up to whole therapeutic process, but set a time T, as the cycle of undated parameter according to the speed needs of Data Update, the value of T will be according to the ability of concrete radiotherapy planning system, and the factors such as speed of computer Rapid Dose Calculation decide.In order to accelerate computational speed, the target area characteristics of motion modeling that need cause patient's physiological movement, for example to breathe and heart beating etc. clocklike physiological parameter estimate, estimate the influence of their motion to target area therapeutic community position.Can by with radiotherapy before through optimizing and was input to the comparison of therapeutic community's position data of stipulating in the initial radiotherapy planning in the therapy apparatus, the quantitative data of acquisition model parameter.The process of input data set that obtains the kinematic parameter of treatment head of control therapy apparatuss and 2DMLC blade movement by these data is identical with the method to set up that begins to carry out radiotherapy planning.These group data also are used for the Inversion Calculation of dosage in the patient body.

As shown in Figure 6, to the dosage inversion algorithm, the present invention will be with the data that collect on the detector, and be stored in advance in the computer with Monte Carlo (Monte Carlo) method calculate the detector response function that obtains obtain with near treatment source one side with the vertical plane of treatment cone-beam central shaft on the energy fluence distribution of incident beam of (being defined as plane of incidence IP:Incident Plane).This IP plane is a virtual plane that needs setting in order to calculate, and the size of pixel on the IP plane will be selected according to the probability of dosage inversion accuracy and COMPUTER CALCULATION speed.The purpose of dosage inverting is that will to obtain on the IP plane with the pixel be the data Ψ of the incident beam of unit _Ij(E _{I, j}, Φ _{I, j}), the E here _{I, j}For incident beam in the position (i, the power spectrum of j) locating, Φ _{I, j}(E _{I, j}) be the fluence on the same position, Φ _{I, j}Can be set at the mean fluence of incident beam, also can be set at the fluence of a certain energy bite inner rays.Footnote (i, j) position coordinates at expression pixel place wherein.Our task is exactly the situation according to the radiotherapy machine radiation source, and the dose data D of detector measurement _d(m, n) (represent with the gray value that measures on the detector, cross through scaleover, the footnote here (m n) expression be pixel location coordinate on the detector plane, at last by determining Ψ on the IP plane _Ij(E _{I, j}, Φ _{I, j}) at each pixel dosage D of detector plane _d(m, contribution weight w n) _Ij, determine Ψ fully _Ij(E _{I, j}, Φ _{I, j}) value, calculate by following formula:

$D_d(m,n)=\underset{i,j}{\mathrm{\Σ}}{w}_{i,j}\·{\mathrm{\Λ}}_{\mathrm{ij}}^{\mathrm{mn}}(E_{i,j},{\mathrm{\Φ}}_{i,j})---\left(2\right)$

The w here _{I, j}Footnote (i, j) and Ψ _Ij(E _{I, j}, Φ _{I, j}) footnote (i, j) value is in full accord, Λ _Ij ^Mn(E _{I, j}, Φ _{I, j}) (i, j) unit's of locating fluence incident beam is in detector probe unit (m, n) response on for plane of incidence IP goes up the position.

In case we know Ψ _Ij(E _{I, j}, Φ _{I, j}) after the value, just can adopt the computational methods of sophisticated any forward radiological dose to obtain the intravital dose distribution of patient by the data of these virtual light sources.In order to obtain Ψ _IjValue, we are with covering moral Caro (Monte Carlo) method, (intersection point of the outer surface of close detector one side of E (i, j)) and corresponding line of incidence and human body is to the distance D of detector at the intravital equivalent radiation journey ER of people the line of incidence of various energy in the energy range of incident radiation source power spectrum covering in advance _rCalculate, constitute ER-D _rThe paired array list of searching exists in the computer in advance.For the versatility of ensuring method, we need ER (E (i, j)) changes into the equivalent length of water, the formula below using:

ER(E(i，j))＝ρ ₁×L ₁+ρ ₂×L ₂+…+ρ _k×L _k (3)

The ρ here ₁, ρ ₂..., ρ _kWhen passing the intravital material of people of process in the process that is detected the device record behind the human body, according to the electron density value with respect to water of the different urstoffs of dividing of electron density, L for some particles (perhaps a certain beam line) ₁, L ₂..., L _kBe the length of this urstoff on directions of rays.Formula (3) all is converted into the length of equivalent water to the material of different densities on the path of ray process, calculates and relatively with convenient.Because calculated in advance comes out to be stored in the ER-D in the computer _rData volume is always limited in the data look-up table, also needs ER-D during actual the use _rCarry out interpolation.Calculated in advance is stored in ER and the D in the computer _rLength, the physical length that can calculate by tissue is inconsistent, adopts the general linear interpolation just can obtain the value of needs.

Above-mentioned dosage inverting also obtains the process that the dose difference based on voxel distributes, and is divided in cycle of rational interval T (selection of T will be finished the needs of oncotherapy and the computational speed of computer is determined according to treatment plan) to finish in whole therapeutic process.Can define the situation that whether reaches real-time dosage according to the size of T controls.When if radiotherapy machine and configuring computer system thereof also do not reach the level of real-time calculating, also can shine on certain angle and calculate by stopping at, the step-by-step system that is rotated irradiation after the acquisition correction again realizes, but the time of waiting on each angle should be that the patient can bear.The improvement that new dosage distributes is by the movement locus of relevant blade among the therapy apparatus anglec of rotation (selection at θ and α angle), rotary speed V, the 2DMLC and the change of movement velocity are realized.These data sets constitute the array of therapy apparatus motor control.Said process just is based on the conformal modulating radiotherapy (IGCIMRT) of the dynamic four-dimensional imaging of flat panel detector for the image-guided on basis.This radiotheraping method has the function of real shape, reverse optimization radiotherapy dosage field distribution in real time and improves the treatment precision.

The present invention is aspect image reconstruction, no matter be the patient on therapeutic bed from the movable pendulum position, still by realtime imaging to therapeutic process, and by the therapeutic community position of Flame Image Process acquisition and the data of tumor outer contour, realization to patient location from dynamic(al) correction and monitoring, thereby realize the deterioration of the dose distribution that causes from motion tracking and to the target area motion of motion target area is proofreaied and correct.The image reconstruction algorithm aspect, we adopt the true three-dimensional reconstruction fast algorithm based on string of a musical instrument section, and the algorithm of this algorithm and traditional tomographic image reconstructing is different, can realize the dynamic four-dimensional imaging to local region of interest.

When implementing IGCIMRT, though whole target area is in the process of implementing treatment, because the physiological movement of human body and patient's the motion of can't help, may move, but concerning most of entity tumors, we think between the inside tumor material by the definition of therapeutic community and tumor outer contour relative motion does not take place.And to the inner tumor that relative motion takes place of tumor material, for example lymphatic cancer and leukemia then adopt real shape to avoid radiotherapy (CART:Conform Avoiding Radiotherapy).The suitable shape here is for treatment target area (PTV); the zone that is whole multi-leaf collimator definition should be plan target area PTV; if in PTV, there is sensitive organization, need by be provided with MLC the position they are protected, this radiotherapy process is called real shape and avoids radiotherapy.

The three-dimensional space position of the position of the present invention's light source during, treatment target area and the size of imaging region also according to radiotherapy, develop and local tumor dosage inversion algorithm a kind of and that treatment tumor target area is complementary, this algorithm is applicable to the various devices of remote tumor radiotherapy, can save the time of Rapid Dose Calculation greatly, satisfy the needs of real-time dosage inverting.

The first embodiment of the present invention, ⁶⁰Co stereotactic radiotherapy machine is loaded onto the embodiment that flat panel detector can be used as above-mentioned thought afterwards.Because ⁶⁰Co γ-source is a monoenergetic, so it is just passable only to consider that on the energy fluence plane fluence distributes, and Ψ at this moment _Ij(E _{I, j}, Φ _{I, j})=Ψ _Ij(Φ _{I, j}).But, have only radiation source to rotate ⁶⁰The method that Co stereotactic radiotherapy machine could use this patent to provide.These are a plurality of to the production of Chinese company ⁶⁰Co or single rotation ⁶⁰Co stereotactic radiotherapy machine all is suitable for.At this moment need to increase by one and be specifically designed to the imaging source of patient being put the position, be reduced to the energy of image source gamma-radiation, can improve contrast to soft-tissue imaging, the formula that provides 1 of this patent, can realize reconstruction from the movable pendulum bit image to patient, by rebuilding the back, determine that the method for therapeutic community and outer contour is then identical to the cutting apart of image.Obtain after the position of the new therapeutic community of tumor, the process that patient is transplanted on therapeutic community automatically by control station be to patient from the movable pendulum position.

And the inversion algorithm to dosage in the patient body that this patent is described is applicable to single source or multi-source fully ⁶⁰The situation of Co stereotactic radiotherapy machine.

The second embodiment of the present invention is loaded onto the description that achieves the above object after the flat panel detector on the electron linear accelerator radiotherapy machine.Here also can adopt a mode in gamma-radiation source, location automatically, realization is to patient's automatic location, it is following better that the energy in gamma-radiation source is chosen in 150keV, and this source can be fixed on the exit of accelerator treatment head beam, the situation when the simulation electronic linear accelerator goes out to restraint fully.

The present invention is applicable to fully that also medical electronic linear accelerator radiation period Real Time Image System is used to monitor the motion of patient's internal organs, and realizes real-time gauged method.Also be fit to by data the four-dimensional dynamic imaging acquisition of patient target region, realization is to the control parameter of adjusting therapy apparatus from motion tracking and location, according to the variation of treatment position of motion target area, satisfy the requirement that changes the post dose field, by the good clinical flow scheme design, whole process is designed to smooth operating process degree.

Realize having adopted a series of advanced technologies in this system's corresponding function: 1. according to the probability of clinical implementation, light source, detector through collimation center on patient target area image data in different angular ranges, data acquisition can advance to revolve in the speed ratio numerical range at limited angle and limited bed to be realized, by the true three-dimensional imaging (VCT) of the true three-dimensional reconstruction algorithm of the string of a musical instrument to patient target area isotropic imaging resolution.2. by the size and the rational flat panel detector picture dot size of selection of restriction light source, guarantee to be better than the positioning accuracy of tomography radiotherapy with the space orientation that true said three-dimensional body imaging data is realized to patient target region; 3. based on the data of VCT imaging, calculating by image segmentation and barycenter, obtain the position coordinates (therapeutic community can be more than, is as the criterion with doctor's prescription) at patient center, by the motion of programme controlled therapeutic bed, automatically realize pendulum position to the patient, the situation of motion during is treating the target area, the motion of dynamic tracking target area realizes the modeling to the target area motion, and compensation guarantees the precision of radiotherapy real-time dynamicly owing to the deterioration to the dosage field distribution that the target area motion causes.For the situation of a plurality of therapeutic communities, then according to pre-determined order, realize respectively to lesions position from motion tracking, and distribute according to the dosage field of dynamic optimization and to adjust the parameter of therapy apparatus, satisfy the requirement of optimizing the post dose field distribution; 4. according to the shape of the characteristics of line and treatment target area, radiating light source is carried out modeling, calculates the kernel function of light source with covering moral Caro method, by with the dynamic rotary course of detector in image data be foundation, realize inverting to therapeutic dose; 5. by the 3-dimensional dose field distribution data that form in the dosage refutation process and the comparison between the field distribution of treatment plan 3-dimensional dose, realize that therapeutic process is to the therapy apparatus parameter control, realize in real time therapeutic process optimization, satisfy tracking and the incomplete compensation of dosage field the motion target area; 5. consider that the CT image that is used to formulate radiotherapy planning at present is still based on the faultage image of non-each homogeny in space, consider owing to reasons such as internal organs motion cause the deviation of radiotherapy planning need be corrected in next radiotherapy, the present invention has also finished being used to formulate the tomography CT diagram data collection of radiotherapy planning by cutting apart focus in three dimensions, join and show, by image process method the 3-dimensional dose field after integrating based on tomography CT image quasi-three-dimensional images and the comparison of true 3-dimensional dose field, comprising to the non-linear interpolation of faultage image in the bed thickness direction, thereby the three dimensions dosage field that obtains after the inverting of assurance process and the 3-dimensional dose field of treatment plan prescribes can compare and analyze on the basis of same spatial resolution, for the modification of next radiotherapy planning provides foundation with further optimizing.The method and system that the present invention realizes has been considered the multiformity of the radiotherapy machine of clinical use, is of universal significance in application facet.

Claims (5)

1. the automatic localized method of patient target region in the radiotherapy is characterized in that, may further comprise the steps:

(1) input image data;

(2) being provided with in therapy equipment can be around the light source and the flat panel detector of the synchronous rotation of patient;

(3) utilize light source and flat panel detector view data, write down corresponding data around patient target region rotation sweep acquisition patient target region;

(4) data are rebuild and handled, obtain the image of patient target region;

(5) view data of the view data of target area and input is compared, if two picture positions do not conform to, execution in step 6 is if two images conform to execution in step 7;

(6) adjust patient's position according to the result of two image comparisons, execution in step 2;

(7) continue treatment.

2. the automatic localized method of patient target region as claimed in claim 1 is characterized in that, the data in the described step (3) comprise: light source is to the distance A of detector, back projection data P, the starting point λ of light source scanning track ₁, the terminal point λ of light source scanning ₂, the scan angle theta of correspondence during light source scanning track starting point ₁, the scan angle theta of correspondence during the light source scanning terminal point ₂

3. the automatic localized method of patient target region as claimed in claim 2 is characterized in that, described step (4) is handled the algorithm for reconstructing that is based on string of a musical instrument section of use to data, promptly when light source from the position 1 when scanning position 2, the track of light source is an arc Corresponding string of a musical instrument section λ ₁λ ₂The image of the object part that is covered is rebuild with following formula:

$f(x_{\mathrm{\π}},{\mathrm{\λ}}_1,{\mathrm{\λ}}_2)=-\frac{1}{2\mathrm{\π}}\underset{{\mathrm{\λ}}_1}{\overset{{\mathrm{\λ}}_2}{\∫}}\mathrm{d\λ}\frac{A^2(u_d,v_d)}{|{\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}^2}H\left\{\frac{P^{\′}({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A^2({u_d}^{\′},{u_d}^{\′})}\right\}$

$-\frac{1}{2\mathrm{\π}}{\left[\frac{A(u_d,v_d)}{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}H\right\{\frac{P({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A({u_d}^{\′},{v_d}^{\′})}\left\}\right]}_{{\mathrm{\λ}}_1}^{{\mathrm{\λ}}_1}$

Wherein $H\left\{\frac{P^{\′}({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A^2({u_d}^{\′},{v_d}^{\′})}\right\}$ With $H\left\{\frac{P({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A({u_d}^{\′},{v_d}^{\′})}\right\}$ Be respectively right $\frac{P^{\′}({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A^2({u_d}^{\′},{v_d}^{\′})}$ With $\frac{P({u_d}^{\′},{v_d}^{\′},\mathrm{\λ})}{A({u_d}^{\′},{v_d}^{\′})}$ The Xi Er baud conversion, in the formula $\underset{{\mathrm{\λ}}_1}{\overset{{\mathrm{\λ}}_2}{\∫}}\mathrm{d\λ}\frac{A^2(u_d,v_d)}{|\stackrel{\→}{r}-{\stackrel{\→}{r}}_0\left(\mathrm{\λ}\right)|}$ Processing is projected in string of a musical instrument section x _{π 1}x _{π 2}On data carry out filtering, then at whole λ ₁λ ₂Integration in the scope, f (x _π, λ ₁, λ ₂) be the grey value profile of rebuilding in the subject image afterwards.

4. the automatic localized method of patient target region as claimed in claim 1 is characterized in that, in the described step (5), the view data of target area and the view data of input are compared by the barycenter of target area, determines two image centroids and obtains corresponding position difference.

5. the automatic localized method of patient target region as claimed in claim 1 is characterized in that, in the described step (6), pass through control system, according to the view data of target area and the view data difference relatively of input, move carrying patient's therapeutic bed, realize the adjustment of patient location.

Images