Proton arc radiotherapy plan planning method and electronic device

Abstract

The application discloses a kind of proton arc radiotherapy plan planning method and electronic equipment, belong to the technical field of proton radiotherapy plan.The proton arc radiotherapy plan planning method provided in the application comprises: obtaining the joint optimization target model with the objective of minimizing total energy switching time of full arc and maximizing total biological LET benefit of target area of full arc, and solving the joint optimization target model to obtain optimal candidate energy layer in each gantry angle;Based on the optimal candidate energy layer in each gantry angle, an initial radiotherapy plan is established, and then based on the radiotherapy plan optimization objective, the optimal beam spot weight vector is obtained, thereby obtaining the optimal radiotherapy plan.In the optimal candidate energy layer selection, the energy switching time of full arc and the target area LET biological effect are considered jointly, which can shorten the treatment time under the premise of guaranteeing the clinical dose requirement, and effectively improve the LET in the target area, enhance the killing effect on tumor cells.

Description

Technical Field

[0001] This invention belongs to the field of proton radiotherapy planning technology, and more specifically, relates to a proton arc radiotherapy planning method and electronic equipment. Background Technology

[0002] Proton arc therapy (PAT) significantly improves target dose conformity and uniformity by continuously delivering the irradiation beam at multiple gantry angles, and is considered to be the next-generation core technology of proton radiotherapy.

[0003] However, the existing PAT technology faces two major bottlenecks: (1) The existing PAT technology usually constructs a radiotherapy plan by randomly selecting energy layers, which requires frequent switching of energy layers, resulting in low delivery efficiency, excessively long total treatment time, increased risk of intra-fractional motion error for patients, and limited clinical promotion; (2) The existing PAT technology usually only considers physical dose or delivery efficiency when selecting energy layers. The selected energy layer combination cannot effectively optimize the LET (i.e., energy transfer line density) distribution in the target area, so that the biological effect dose of tumor cells cannot reach the optimal level and the killing effect is poor. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention provides a proton arc radiotherapy planning method and electronic device, the purpose of which is to shorten the treatment time and effectively increase the LET in the target area while ensuring the clinical dose requirements, thereby enhancing the killing effect on tumor cells.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a proton arc radiotherapy planning method for a cyclotron proton therapy system; wherein the proton arc radiotherapy planning method includes: A joint optimization objective model was developed to minimize the total time of full-arc energy switching and maximize the total biological benefit of LET in the full-arc target area. The joint optimization objective model was solved to obtain the optimal candidate energy layer for each rack angle. The total time of full-arc energy switching is the sum of the switching times between corresponding candidate energy layers in all two adjacent rack angles in the system. The total biological benefit of LET in the full-arc target area is the depth of the target area irradiated by the corresponding candidate energy layer at each rack angle in the system. The sum of the energy transfer linear density values ​​corresponding to the time; Based on the optimal candidate energy layer at each gantry angle, an initial radiotherapy plan is established, including a beam spot weight vector, beam information, and the corresponding dose influence matrix and energy transfer linear density contribution matrix. The beam information includes the optimal candidate energy layer for each gantry angle and the beam spot position for each optimal candidate energy layer. The beam spot weight vector has the dimension of the total number of known beam spots. The dose influence matrix and the energy transfer linear density contribution matrix have the same dimension, with both having the number of rows equal to the total number of voxels divided within the same target region and the number of columns equal to the total number of known beam spots. Based on the initial radiotherapy plan and the radiotherapy plan optimization objective, the optimal beam weight vector is obtained, and thus the optimal radiotherapy plan is obtained. The radiotherapy planning optimization objectives include: minimizing the deviation between the actual dose distribution and the prescribed dose distribution in the target area, and the deviation between the actual energy transfer line density distribution and the dose-weighted target energy transfer line density distribution; the actual dose distribution is the product of the dose influence matrix and the beam spot weight vector; the actual energy transfer line density distribution is the product of the energy transfer line density contribution matrix and the beam spot weight vector; and the dose-weighted target energy transfer line density distribution is the pixel-by-pixel multiplication result of the actual dose distribution and the target energy transfer line density distribution.

[0006] More preferably, the objective of the aforementioned joint optimization objective model further includes minimizing the total energy change rate; wherein, the total energy change rate is used to characterize the difference in amplitude between the corresponding candidate energy layers in each of two adjacent gantry angles and the maximum single energy change value allowed by the cyclotron proton therapy system. The sum of the ratios; The constraints of the above joint optimization objective model include: the difference in amplitude of the corresponding candidate energy layer in any two adjacent rack angles is less than or equal to And the candidate energy layer corresponding to each rack angle belongs to the corresponding candidate energy layer set.

[0007] More preferably, the candidate energy layer set for any rack angle a is obtained by the following method: clustering each energy layer in rack angle a, obtaining the energy layers whose distance from the cluster center is less than a preset distance, and forming the candidate energy layer set for rack angle a.

[0008] More preferably, the candidate energy layer set for any rack angle a is obtained by the following method: extracting the feature vectors of each energy layer in rack angle a, clustering them in the latent space, obtaining the energy layers whose distance from the cluster center is less than a preset distance, and forming the candidate energy layer set for rack angle a.

[0009] More preferably, the objective function of the joint optimization objective model is:

[0010] Among them, the The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle The amplitude of each energy layer; For the first The index of the candidate energy layer corresponding to each rack angle; This represents the total number of frame angles. , and All are adaptive dynamic weighting factors; the first The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle One energy layer; For the first The rack angle and the first The switching time between candidate energy layers in each rack angle; For the first The depth of the candidate energy layer irradiating the target area at each rack angle. The corresponding linear energy density value at that time.

[0011] More preferably, the switching time between corresponding candidate energy layers in any two adjacent rack angles in the system is obtained by substituting the absolute value of the difference in amplitude between the corresponding candidate energy layers in the two adjacent rack angles into the energy layer switching time model; Among them, the energy layer switching time model is based on... The relationship between the measured energy layer switching time and the actual energy layer switching time is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; For the first The first rack angle The amplitude of each energy layer; ; ; ; For the first The number of energy layers in each rack angle; For the first The number of energy layers in each rack angle; The depth of the candidate energy layer irradiating the target area at any rack angle 'a' in the system. The corresponding energy transfer linear density value is obtained by combining the amplitude of the candidate energy layer at the gantry angle α with the target depth. Substituting this into the LET bootstrapping model yields the following result; Among them, the LET-guided model uses amplitude and target depth The actual measured depth at which the corresponding energy layer at the corresponding gantry angle is irradiated into the target area. The relationship between the energy transfer linear density values ​​is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; ; ; For the first The number of energy layers in each rack angle; The first in the preset target depth set Target depth; ; This represents the number of target depths in the preset target depth set.

[0012] More preferably, the above-mentioned radiotherapy planning optimization objectives also include: minimizing the total dose on organs at risk, and the deviation between the dose distribution on the target area and the prescribed dose distribution on the target area.

[0013] More preferably, the dose influence matrix and energy transfer linear density contribution matrix corresponding to the beam information in the initial radiotherapy plan are determined as follows: Based on the beam information and the parameters of the cyclotron proton therapy system, the dose contribution of each beam spot to each voxel in the target area is calculated, and the corresponding dose influence matrix is ​​obtained. Based on the beam information and the parameters of the cyclotron proton therapy system, the energy transfer linear density contribution of each beam spot to each voxel in the target area is calculated, and the corresponding energy transfer linear density contribution matrix is ​​obtained.

[0014] In a second aspect, the present invention provides an electronic device comprising: a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the proton arc radiotherapy planning method provided in the first aspect of the present invention.

[0015] Thirdly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed by a processor, controls the device containing the storage medium to perform the proton arc radiotherapy planning method provided in the first aspect of the present invention.

[0016] Fourthly, the present invention also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the proton arc radiotherapy planning method provided in the first aspect of the present invention.

[0017] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: 1. This invention provides a proton arc radiotherapy planning method. It obtains a joint optimization objective model aimed at minimizing the total time of full-arc energy switching and maximizing the total LET biological benefit of the full-arc target area. The joint optimization objective model is solved to obtain the optimal candidate energy layer for each gantry angle. Based on the optimal candidate energy layer for each gantry angle, an initial radiotherapy plan is established. Then, based on the radiotherapy plan optimization objective, the optimal beam weight vector is solved, thus obtaining the optimal radiotherapy plan. This invention considers that LET directly determines the relative biological effect, and high LET regions have stronger killing power against hypoxic and radioresistant tumor cells. In the selection of the optimal candidate energy layer, the full-arc energy switching time and the target area LET biological effect are jointly considered, resulting in a better initial radiotherapy plan. The optimized radiotherapy plan obtained on this basis can significantly improve the average LET in the clinical target area while shortening the treatment time, achieving synergistic optimization of treatment efficiency and biological efficacy. Meanwhile, since the total LET biological benefit of the full-arc target area is the depth of the target area irradiated by the corresponding candidate energy layer at each gantry angle... The sum of the corresponding linear energy density values ​​is closely related to the target depth. This invention can preferentially select the optimal energy layers that can provide higher LET values ​​to deep or core target areas, thereby enhancing the killing effect on radioresistant tumor cells (such as hypoxic cells) without increasing or even decreasing the dose to normal tissues, thus improving the local tumor control rate. Based on this, this invention can shorten the treatment time and effectively increase the LET within the target area while ensuring clinical dose requirements, thereby enhancing the killing effect on tumor cells.

[0018] 2. Furthermore, in the proton arc radiotherapy planning method provided by the present invention, the objective of the aforementioned joint optimization target model further includes minimizing the total energy change rate; wherein, the total energy change rate is used to characterize the difference in amplitude between the corresponding candidate energy layers in each of two adjacent gantry angles and the maximum single-shot energy change value allowed by the cyclotron proton therapy system. The sum of the ratios. This mechanism, by suppressing large jumps in energy layers between adjacent angles, improves the smoothness of the energy sequence and the stability of system delivery while ensuring the global optimum, avoiding unnecessary hardware stress, and ensuring that the optimization results are feasible and efficient in clinical practice. Attached Figure Description

[0019] Figure 1 This is a flowchart of a proton arc radiotherapy planning method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the optimal energy layer distribution using a brain cancer case as an example, provided in an embodiment of the present invention. Figure 3 The optimal energy transfer linear density volume histogram and the corresponding dose volume histogram for brain cancer cases provided in the embodiments of the present invention; Figure 4 Dose slice diagram of a brain cancer case provided in this embodiment of the invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0021] Example 1 This embodiment provides a proton arc radiotherapy planning method for a cyclotron proton therapy system; wherein, as Figure 1 As shown, the proton arc radiotherapy planning method includes: A joint optimization objective model was developed to minimize the total time of full-arc energy switching and maximize the total biological benefit of LET (Light Transfer Linear Density) in the full-arc target area. The model was then solved to obtain the optimal candidate energy layer for each rack angle. The total time of full-arc energy switching is the sum of the switching times between corresponding candidate energy layers in all adjacent rack angles within the system. The total biological benefit of LET (Light Transfer Linear Density) in the full-arc target area is the depth of the corresponding candidate energy layer irradiated into the target area at each rack angle within the system. The sum of the energy transfer linear density values ​​corresponding to the time; Based on the optimal candidate energy layer at each gantry angle, an initial radiotherapy plan is established, including a beam spot weight vector, beam information, and the corresponding dose influence matrix and energy transfer linear density contribution matrix. The beam information includes the optimal candidate energy layer for each gantry angle and the beam spot position for each optimal candidate energy layer. The beam spot weight vector has the dimension of the total number of known beam spots. The dose influence matrix and the energy transfer linear density contribution matrix have the same dimension, with both having the number of rows equal to the total number of voxels divided within the same target region and the number of columns equal to the total number of known beam spots. Based on the initial radiotherapy plan and the radiotherapy plan optimization objective, the optimal beam weight vector is obtained, and thus the optimal radiotherapy plan is obtained. The radiotherapy planning optimization objectives include: minimizing the deviation between the actual dose distribution and the prescribed dose distribution in the target area, and the deviation between the actual energy transfer line density distribution and the dose-weighted target energy transfer line density distribution; the actual dose distribution is the product of the dose influence matrix and the beam spot weight vector; the actual energy transfer line density distribution is the product of the energy transfer line density contribution matrix and the beam spot weight vector; and the dose-weighted target energy transfer line density distribution is the pixel-by-pixel multiplication result of the actual dose distribution and the target energy transfer line density distribution.

[0022] This embodiment constructs a dual-criteria optimization framework comprising a machine delivery time model and a depth-dependent linear energy density guidance model. It performs joint solutions at two levels: energy layer selection and beam spot weight optimization, achieving synergistic optimization of treatment efficiency and biological effects. This facilitates the clinical application of bio-guided proton arc radiotherapy. Specifically, in selecting the optimal candidate energy layer, this embodiment jointly considers machine delivery time and the target area LET biological effect. Compared to methods that only optimize delivery efficiency or independently optimize LET, this embodiment can significantly shorten treatment time while significantly improving the average LET within the clinical target area, achieving synergistic optimization of treatment efficiency and biological efficacy.

[0023] Preferably, in an optional implementation, the objective of the aforementioned joint optimization objective model further includes minimizing the total energy change rate; wherein the total energy change rate is used to characterize the difference in amplitude between the corresponding candidate energy layers in each of two adjacent gantry angles and the maximum single energy change value allowed by the cyclotron proton therapy system. The sum of the ratios. Typically determined by the accelerator physics constraints in a cyclotron proton therapy system, these constraints ensure that the selected energy layer sequence is physically feasible.

[0024] The constraints of the above joint optimization objective model include: the difference in amplitude of the corresponding candidate energy layer in any two adjacent rack angles is less than or equal to And the candidate energy layer corresponding to each rack angle belongs to the corresponding candidate energy layer set.

[0025] It should be noted that the candidate energy layer set can be the set of all energy layers in the corresponding rack angle.

[0026] To further simplify the candidate range and improve computational efficiency, preferably, in one optional implementation, the candidate energy layer set for any rack angle a is obtained by clustering each energy layer in rack angle a, and obtaining the energy layers whose distance from the cluster center is less than a preset distance, thus forming the candidate energy layer set for rack angle a.

[0027] It should be noted that clustering can employ any existing clustering algorithm, such as K-means or KNN, and is not limited here. Preferably, in one optional implementation, the candidate energy layer set for any rack angle 'a' is obtained by: extracting the feature vectors of each energy layer in rack angle 'a', clustering them in the latent space, and obtaining energy layers whose distance from the cluster center is less than a preset distance, thus forming the candidate energy layer set for rack angle 'a'. This method can further improve the accuracy of clustering.

[0028] In one alternative implementation, the objective function of the joint optimization objective model is:

[0029] Among them, the The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle The amplitude of each energy layer; For the first The index of the candidate energy layer corresponding to each rack angle; This represents the total number of frame angles. , and All are adaptive dynamic weighting factors; the first The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle One energy layer; For the first The rack angle and the first The switching time between candidate energy layers in each rack angle; For the first The depth of the candidate energy layer irradiating the target area at each rack angle. The corresponding linear energy density value at that time.

[0030] The corresponding constraints include: ; ; ;in, For the first A set of candidate energy layers for each rack angle; For the first A set of candidate energy layers for each rack angle.

[0031] It should be noted that the switching time between corresponding candidate energy layers in any two adjacent rack angles within the system can be obtained through actual measurement or based on an energy layer switching time model. Preferably, in one optional implementation, By Substitute into the energy layer switching time model The calculations were performed; among them, the energy layer switching time model was obtained. The energy layer switching time model is obtained by fitting the relationship between the absolute value of the difference in amplitude between corresponding candidate energy layers in two adjacent rack angles and the corresponding measured energy layer switching time. Specifically, the energy layer switching time model is obtained by fitting the relationship between the absolute value of the difference in amplitude between corresponding candidate energy layers in two adjacent rack angles and the measured energy layer switching time. The relationship between the measured energy layer switching time and the actual energy layer switching time is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; For the first The first rack angle The amplitude of each energy layer; ; ; ; For the first The number of energy layers in each rack angle; For the first The number of energy layers in each rack angle.

[0032] In this embodiment, a mathematical model is established between the energy layer switching time and the energy change amplitude by fitting the absolute value of the amplitude difference between corresponding candidate energy layers in two adjacent gantry angles in the beam transport log of the cyclotron proton therapy system with the corresponding measured energy layer switching time. This model serves as the energy layer switching time model. The fitting method can be the least squares method, spline difference method, etc., and is not limited here. During fitting, the combination of amplitude differences of all candidate energy layers in two adjacent gantry angles is obtained, i.e. ; ; ; ; For the first The number of energy layers in each rack angle; For the first The number of energy layers in each rack angle. Preferably, to avoid overfitting and oscillation problems that may be caused by simple polynomial fitting, this invention uses cubic spline interpolation to smoothly fit the discrete measurement points, thereby obtaining the energy layer switching time model.

[0033] It should be noted that the depth of the candidate energy layer irradiating the target area at any rack angle 'a' in the system is... The corresponding linear energy density value can be obtained through actual measurement or based on the LET-guided model. Preferably, in one optional implementation, By amplitude and target depth The results are obtained by substituting these values ​​into the LET guiding model; the LET guiding model is obtained by fitting the relationship between the amplitude of different energy layers and different target depths at different rack angles and the energy transfer linear density values ​​of the corresponding candidate energy layers at that rack angle irradiating the corresponding target depths. Specifically, the LET guiding model is obtained by fitting the amplitude... and target depth The actual measured depth at which the corresponding energy layer at the corresponding gantry angle is irradiated into the target area. The relationship between the energy transfer linear density values ​​is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; ; ; For the first The number of energy layers in each rack angle; The first in the preset target depth set Target depth; ; This represents the number of target depths in the preset target depth set. The fitting method can include least squares, spline interpolation, etc., and is not limited here. During fitting, the amplitudes of all candidate energy layers at different gantry angles are obtained. ; During fitting, the relationship curve representing the model can be an S-shaped growth curve, a polynomial curve, etc., and there is no limitation here. Preferably, an S-shaped growth curve is used in this embodiment.

[0034] By introducing a depth-dependent LET-guided model, this embodiment can preferentially select energy layers that can provide higher LET values ​​for deep or core target areas, thereby enhancing the killing effect on radioresistant tumor cells (such as hypoxic cells) without increasing or even decreasing the dose to normal tissues, and is expected to improve the local tumor control rate.

[0035] In one optional implementation, the dose influence matrix and energy transfer linear density contribution matrix corresponding to the beam information in the initial radiotherapy plan are determined as follows: Based on the beam information and the parameters of the cyclotron proton therapy system, the dose contribution of each beam spot to each voxel in the target area is calculated, and the corresponding dose influence matrix is ​​obtained. Based on the beam information and the parameters of the cyclotron proton therapy system, the energy transfer linear density contribution of each beam spot to each voxel in the target area is calculated, and the corresponding energy transfer linear density contribution matrix is ​​obtained.

[0036] It should be noted that there are various methods for solving the joint optimization objective model, such as dynamic programming algorithms, gradient-based optimization methods (e.g., gradient descent, Newton's method), alternating direction multiplier method, simulated annealing algorithm, etc., which are not limited here. Preferably, in one optional implementation, a dynamic programming algorithm is used to solve the joint optimization objective model. In this implementation, the energy layer selection problem is modeled as a two-dimensional dynamic programming problem, and the state transition cost is defined as follows: = -

[0037] The dynamic programming recurrence formula is: ; Boundary conditions are = Finally, the optimal energy layer sequence is obtained by backtracking. .

[0038] Figure 2 The optimal energy layer distribution is illustrated using a brain cancer case as an example.

[0039] In one alternative implementation, the functional expression for the above-mentioned radiotherapy planning optimization objective is:

[0040] in, This is the dose effect matrix; Here is the beam spot weight vector; Prescription dose distribution in the target area; Contribution matrix to energy transfer linear density; Here, is the regularization parameter, representing the weighting coefficient of the energy transfer linear density target term; This represents an element-wise multiplication operation; The target energy transfer linear density distribution.

[0041] Preferably, in an optional implementation, the above-mentioned radiotherapy planning optimization objective further includes: minimizing the total dose on organs at risk and the deviation between the dose distribution on the target area and the prescribed dose distribution on the target area. In this case, the functional expression for the above-mentioned radiotherapy planning optimization objective is:

[0042] in, Total dose to organs at risk; This refers to the deviation between the dose distribution on the target area and the prescribed dose distribution on the target area.

[0043] It should be noted that there are various methods for obtaining the optimal beam spot weight vector based on the initial radiotherapy plan and the radiotherapy plan optimization objective, such as gradient descent, proximal gradient descent, Newton's method, quasi-Newton's method, simulated annealing algorithm, etc., which are not limited here. Preferably, in one optional implementation, the proximal gradient descent method is used to solve for the beam spot weight vector, defining... :

[0044] in, and All are auxiliary variables; For dose calculation, the region of interest is defined. Let y be the LET contribution distribution; Let y be the target LET contribution distribution; Optimize the region of interest for LET.

[0045] The updated formula is: The near-end projection operator is:

[0046] in, The minimum hop count constraint threshold, Step size; Iterate to Or, upon reaching the maximum number of iterations, the optimal beam spot weight vector will be output. ; This is a preset threshold.

[0047] like Figure 3 The optimal energy transfer linear density volume histogram and the corresponding dose volume histogram for a brain cancer case are shown.

[0048] Figure 4 A dose-slice diagram of a brain cancer case is shown.

[0049] In summary, this embodiment simultaneously optimizes the delivery efficiency and biological effects of proton arc radiotherapy, constructing a complete dual-criteria joint-modulation treatment planning framework. This achieves shorter treatment times and a more optimized target area energy transfer line density distribution, providing a feasible and efficient bio-optimized proton arc radiotherapy planning method. In this embodiment, based on the beam transport time characteristics of the cyclotron proton therapy system, an energy layer switching time penalty term was constructed, effectively optimizing the arc delivery path, significantly improving delivery efficiency, and shortening the single treatment cycle. Furthermore, a depth-dependent energy transfer line density guidance model (i.e., LET-guided model) was introduced, assigning higher biological effect weights to deep target areas during the energy layer selection stage, fundamentally improving the target area LET distribution. This embodiment generates a highly efficient and biologically enhanced proton arc radiotherapy plan, improving radiotherapy delivery efficiency, providing superior dose coverage and a highly modulated energy transfer line density distribution. It is expected to clear obstacles for the clinical application of proton arc radiotherapy technology and promote the clinical application and development of bio-guided proton rotational intensity-modulated radiotherapy.

[0050] Example 2 This embodiment provides an electronic device, including: a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention.

[0051] The relevant technical solutions are the same as the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0052] Example 3 This embodiment also provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when run by a processor, controls the device where the storage medium is located to execute the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention.

[0053] The relevant technical solutions are the same as the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0054] Example 4 This embodiment also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention.

[0055] The relevant technical solutions are the same as the proton arc radiotherapy planning method provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0056] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for planning proton arc radiotherapy, characterized in that, For use in cyclotron proton therapy systems; The proton arc radiotherapy planning method includes: A joint optimization objective model is obtained, aiming to minimize the total time of full-arc energy switching and maximize the total biological benefit of LET in the full-arc target area. This model is then solved to obtain the optimal candidate energy layer for each rack angle. The total time of full-arc energy switching is the sum of the switching times between corresponding candidate energy layers in all adjacent rack angles of the system. The total biological benefit of LET in the full-arc target area is the depth of the target area irradiated by the corresponding candidate energy layer at each rack angle in the system. The sum of the energy transfer linear density values ​​corresponding to the time; Based on the optimal candidate energy layer at each gantry angle, an initial radiotherapy plan is established, including a speckle weight vector, beam information, and the corresponding dose influence matrix and energy transfer linear density contribution matrix. The beam information includes the optimal candidate energy layer for each gantry angle and the speckle position for each optimal candidate energy layer. The dimension of the speckle weight vector is the total number of known specks. The dose influence matrix and the energy transfer linear density contribution matrix have the same dimension, with the number of rows being the total number of voxels divided within the same target region and the number of columns being the total number of known specks. Based on the initial radiotherapy plan and the radiotherapy plan optimization objective, the optimal beam spot weight vector is obtained, and thus the optimal radiotherapy plan is obtained. The radiotherapy planning optimization objectives include: minimizing the deviation between the actual dose distribution and the prescribed dose distribution in the target area, and the deviation between the actual energy transfer line density distribution and the dose-weighted target energy transfer line density distribution; the actual dose distribution is the product of the dose influence matrix and the beam spot weight vector; the actual energy transfer line density distribution is the product of the energy transfer line density contribution matrix and the beam spot weight vector; and the dose-weighted target energy transfer line density distribution is the pixel-by-pixel multiplication result of the actual dose distribution and the target energy transfer line density distribution.

2. The proton arc radiotherapy planning method according to claim 1, characterized in that, The objective of the joint optimization objective model also includes minimizing the total energy change rate; wherein, the total energy change rate is used to characterize the difference in amplitude between the corresponding candidate energy layers in each of the two adjacent gantry angles in the system and the maximum single energy change value allowed by the cyclotron proton therapy system. The sum of the ratios; The constraints of the joint optimization objective model include: the difference in amplitude of the corresponding candidate energy layer in any two adjacent rack angles is less than or equal to And the candidate energy layer corresponding to each rack angle belongs to the corresponding candidate energy layer set.

3. The proton arc radiotherapy planning method according to claim 2, characterized in that, The candidate energy layer set for any rack angle a is obtained by clustering each energy layer in rack angle a, and obtaining the energy layers whose distance from the cluster center is less than a preset distance, thus forming the candidate energy layer set for rack angle a.

4. The proton arc radiotherapy planning method according to claim 3, characterized in that, The candidate energy layer set for any rack angle a is obtained by the following method: extracting the feature vectors of each energy layer in rack angle a, clustering them in the latent space, obtaining the energy layers whose distance from the cluster center is less than a preset distance, and forming the candidate energy layer set for rack angle a.

5. The proton arc radiotherapy planning method according to any one of claims 2-4, characterized in that, The objective function of the joint optimization objective model is: Among them, the The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle The amplitude of each energy layer; For the first The index of the candidate energy layer corresponding to each rack angle; This represents the total number of frame angles. , and All are adaptive dynamic weighting factors; the first The amplitude of the corresponding candidate energy layer at each rack angle For the first The first rack angle One energy layer; For the first The rack angle and the first The switching time between candidate energy layers in each rack angle; For the first The depth of the candidate energy layer irradiating the target area at each rack angle. The corresponding energy transfer linear density value at that time.

6. The proton arc radiotherapy planning method according to any one of claims 1-4, characterized in that, The switching time between corresponding candidate energy layers in any two adjacent rack angles in the system is obtained by substituting the absolute value of the difference in amplitude between the corresponding candidate energy layers in the two adjacent rack angles into the energy layer switching time model; The energy layer switching time model is based on... The relationship between the measured energy layer switching time and the actual energy layer switching time is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; For the first The first rack angle The amplitude of each energy layer; ; ; ; For the first The number of energy layers in each rack angle; For the first The number of energy layers in each rack angle; In the system, the depth to which the candidate energy layer at any rack angle 'a' illuminates the target area is [depth]. The corresponding energy transfer linear density value is obtained by combining the amplitude of the candidate energy layer at the gantry angle α with the target depth. Substituting this into the LET bootstrapping model yields the following result; The LET guidance model uses amplitude... and target depth The actual measured depth at which the corresponding energy layer at the corresponding gantry angle is irradiated into the target area. The relationship between the energy transfer linear density values ​​is obtained by fitting the data. For the first The first rack angle The amplitude of each energy layer; ; ; For the first The number of energy layers in each rack angle; The first in the preset target depth set Target depth; ; This represents the number of target depths in the preset target depth set.

7. The proton arc radiotherapy planning method according to any one of claims 1-4, characterized in that, The radiotherapy planning optimization objectives also include minimizing the total dose to organs at risk and the deviation between the dose distribution on the target area and the prescribed dose distribution on the target area.

8. The proton arc radiotherapy planning method according to any one of claims 1-4, characterized in that, The dose influence matrix and energy transfer linear density contribution matrix corresponding to the beam information in the initial radiotherapy plan are determined as follows: Based on the beam information and the parameters of the cyclotron proton therapy system, the dose contribution of each beam spot to each voxel in the target area is calculated, and the corresponding dose influence matrix is ​​obtained. Based on the beam information and the parameters of the cyclotron proton therapy system, the energy transfer linear density contribution of each beam spot to each voxel in the target area is calculated, and the corresponding energy transfer linear density contribution matrix is ​​obtained.

9. An electronic device, characterized in that, include: A memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the proton arc radiotherapy planning method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed by a processor, it controls the device containing the storage medium to perform the proton arc radiotherapy planning method according to any one of claims 1-8.

Images