BNCT treatment planning optimization method

Abstract

The present application provides a method for optimizing a BNCT treatment plan, comprising: obtaining a candidate set of BNCT treatment plans; determining multiple quality indicators and corresponding standard values ​​of the BNCT treatment plan, and determining an indicator scoring function based on the quality indicators and the corresponding standard values; obtaining indicator weights of the multiple quality indicators, and obtaining a plan scoring function based on the indicator weights and the indicator scoring function; obtaining a scoring regression model based on the plan scoring function, and iteratively calculating the maximum score in the candidate set based on the scoring regression model and the plan scoring function to obtain the corresponding optimal BNCT treatment plan. By setting a plan scoring function, the present application can select the highest-scoring BNCT treatment plan with fewer computing resources, thereby achieving accurate and efficient BNCT treatment.

Description

Technical Field

[0001] The present application belongs to the field of radiological technology and relates to a BNCT treatment plan optimization method. Background Art

[0002] Boron Neutron Capture Therapy (BNCT) is a precision radiotherapy that combines targeted drugs with neutron irradiation, selectively killing tumor cells while sparing normal cells. Compared to photon radiotherapy, BNCT effectively reduces radiation exposure to normal cells, minimizing the side effects of radiotherapy. Photon radiotherapy typically requires multiple doses, resulting in a long treatment cycle. However, BNCT requires only one or two doses to achieve the required tumor-killing dose, making treatment time very patient-friendly and reducing the burden of treatment. Compared to other radiotherapy techniques, BNCT offers significant advantages in treatment time, precision, and side effect control, and holds broad application prospects.

[0003] Treatment planning (TP) plays a crucial role in BNCT. A well-designed BNCT treatment plan ensures the radiation dose to the target tumor while minimizing the radiation dose to normal tissues, ensuring successful treatment. Therefore, how to efficiently optimize BNCT treatment plans is a technical challenge urgently needed by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide a BNCT treatment plan optimization method to solve the technical problem of how to efficiently optimize the BNCT treatment plan.

[0005] In a first aspect, the present application provides a method for optimizing a BNCT treatment plan, the method comprising:

[0006] Obtaining a candidate set of BNCT treatment plans;

[0007] Determining a plurality of quality indicators of the BNCT treatment plan and corresponding standard values ​​of the indicators, and determining an indicator scoring function based on the quality indicators and the corresponding standard values ​​of the indicators;

[0008] Obtaining indicator weights of a plurality of the quality indicators to obtain a plan scoring function based on the indicator weights and the indicator scoring function;

[0009] A scoring regression model is obtained based on the plan scoring function, and a maximum score in the candidate set is iteratively calculated based on the scoring regression model and the plan scoring function to obtain a corresponding optimal BNCT treatment plan.

[0010] In one embodiment of the present application, determining an indicator scoring function based on the quality indicator and the corresponding indicator standard value includes:

[0011] Obtaining an indicator gap between a current value of the quality indicator and a standard value of the indicator;

[0012] Perform reward evaluation or penalty evaluation on the indicator gap based on the evaluation slope;

[0013] The indicator gap is optimized based on the reward evaluation and the penalty evaluation to obtain the indicator scoring function.

[0014] In one embodiment of the present application, performing reward evaluation or penalty evaluation on the indicator gap based on the evaluation slope includes:

[0015] When the indicator gap meets the preset gap target, the indicator gap is rewarded and evaluated using the reward evaluation slope;

[0016] When the indicator gap does not meet the preset gap target, a penalty evaluation slope is used to perform a penalty evaluation on the indicator gap; wherein the reward evaluation slope is smaller than the penalty evaluation slope.

[0017] In one embodiment of the present application, multiple quality indicators for determining a BNCT treatment plan include:

[0018] Determine the region of interest for the BNCT treatment plan, which includes the tumor target volume, organs at risk, and planning area;

[0019] Determine a plurality of quality indicators of the region of interest based on the radiation standard of the region of interest

[0020] In one embodiment of the present application, obtaining a scoring regression model based on the plan scoring function includes:

[0021] Select the mean function and covariance function;

[0022] specifying a sparse prior distribution for hyperparameters of the covariance function;

[0023] The planned scoring function is fitted based on the mean function, the covariance function, and the sparse prior distribution of the hyperparameters to obtain the scoring regression model.

[0024] In one embodiment of the present application, iteratively calculating the maximum score value in the candidate set based on the score regression model and the planned score function includes:

[0025] Randomly selecting multiple BNCT treatment plans from the candidate set as known sample points;

[0026] Calculate the scores of the known sample points using the planned scoring function, and use the known sample points and corresponding scores as a known sample set;

[0027] Iteratively updating the posterior distribution of the scoring regression model using the known sample set to obtain a predicted distribution of unscored plans in the candidate set based on the posterior distribution;

[0028] The known sample set is iteratively updated based on the predicted distribution of the unscored plans until a maximum number of iterations is reached, and the BNCT treatment plan with the largest score in the known sample set is determined as the optimal plan.

[0029] In one embodiment of the present application, iteratively updating the known sample set based on the predicted distribution of the unrated plan includes:

[0030] Selecting an optimal plan to be scored based on the predicted distribution of the unscored plans, and calculating a score of the optimal plan to be scored using the scoring function;

[0031] The optimal plan to be scored and the corresponding score are added to the known sample set to iteratively update the known sample set.

[0032] In one embodiment of the present application, selecting the optimal plan to be scored based on the predicted distribution of the unscored plans includes:

[0033] evaluating the potential value of the unrated plan based on the predicted distribution of the unrated plan by an acquisition function;

[0034] The unrated plan with the highest potential value is selected as the optimal plan to be rated.

[0035] In one embodiment of the present application, obtaining a candidate set of BNCT treatment plans includes:

[0036] Determine multiple irradiation fields;

[0037] Randomly determining a plurality of groups of corresponding numbers of irradiation weights, wherein the sum of the irradiation weights in each group is 1; the irradiation weights correspond to the irradiation time of the irradiation field;

[0038] A plurality of irradiation fields and a plurality of sets of irradiation weights are randomly and exhaustively combined to form a plurality of candidate BNCT treatment plans.

[0039] In one embodiment of the present application, determining a plurality of irradiation fields includes:

[0040] Determine the geometric center point of the tumor target area;

[0041] Calculating the distance between the contour point of the tumor target area and the geometric center point to determine the closest irradiation angle;

[0042] Taking the closest irradiation angle as the center, a rotation is performed within the horizontal plane of the tumor target area to determine multiple irradiation fields with different irradiation angles.

[0043] As described above, the BNCT treatment plan optimization method described in this application has the following beneficial effects:

[0044] This application obtains an indicator scoring function by setting multiple quality indicators and indicator standard values ​​of the BNCT treatment plan, and further obtains a plan scoring function based on different weights, thereby scoring multiple BNCT treatment plans, which helps to select the optimal BNCT treatment plan based on the patient's actual situation to achieve the best treatment effect. In addition, the plan scoring function helps to quantitatively evaluate the quality of the BNCT treatment plan, avoiding the impact of plan quality on differences in designer experience or time costs. In addition, this application iteratively calculates the maximum score based on the plan scoring function, and can set the number of iterations according to actual conditions. When the number of iterations is reached, the BNCT treatment plan with the maximum score is selected as the optimal plan. This significantly reduces computing resources, can more efficiently select the optimal plan based on actual conditions, and reduces optimization time. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A flow chart of the BNCT treatment plan optimization method described in an embodiment of the present application is shown.

[0046] Figure 2 A schematic diagram of the process of determining the indicator scoring function described in an embodiment of the present application is shown.

[0047] Figure 3 A schematic diagram of the process of obtaining a scoring regression model described in an embodiment of the present application is shown.

[0048] Figure 4 A schematic diagram of the process of iteratively calculating the maximum score in a candidate set according to an embodiment of the present application is shown.

[0049] Figure 5 A logical diagram of the BNCT treatment plan optimization method described in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0050] The following describes the embodiments of the present application through specific examples. Those skilled in the art can easily understand the other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0051] It should be noted that in the following description, reference is made to the accompanying drawings, which describe several embodiments of the present application. It should be understood that other embodiments may also be used, and that mechanical composition, structural, electrical, and operational changes may be made without departing from the spirit and scope of the present application. The following detailed description should not be considered restrictive, and the scope of the embodiments of the present application is limited only by the claims of the published patents. The terms used herein are only for describing specific embodiments and are not intended to limit the present application. Spatially related terms, such as "upper", "lower", "left", "right", "below", "below", "lower", "above", "upper", etc., may be used in the text to facilitate the description of the relationship between an element or feature shown in the figure and another element or feature.

[0052] Furthermore, as used herein, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context indicates otherwise. It should be further understood that the terms "comprise," "include," and "include" indicate the presence of the stated features, operations, elements, components, items, categories, and / or groups, but do not preclude the presence, occurrence, or addition of one or more other features, operations, elements, components, items, categories, and / or groups. The terms "or" and "and / or" used herein are to be interpreted as inclusive, or mean any one or any combination.

[0053] Treatment plan quality assessment (TPQA) technology is crucial in radiotherapy, ensuring the accuracy and safety of treatment plans. Accurate quality assessment can help planners optimize BNCT treatment plans and select the optimal plan to achieve the best treatment outcome. Generally speaking, ensuring the radiation dose to the tumor target while minimizing the radiation dose to normal tissues is the primary principle for evaluating BNCT treatment plan quality. To achieve this, plan designers often employ Monte Carlo simulations to continuously adjust BNCT treatment plan parameters, such as the duration of different irradiation fields, in order to achieve optimal treatment outcomes. This is inherently time-consuming and labor-intensive, requiring significant computational resources and time. Furthermore, due to the lack of specific evaluation metrics, BNCT treatment plan quality is often impacted by differences in plan designer experience and time costs. This makes it difficult to design an appropriate treatment plan based on the patient's specific condition and achieve the best treatment outcome through quantitative evaluation of the plan.

[0054] In order to at least solve the above-mentioned technical problems, an embodiment of the present application provides a BNCT treatment plan optimization method. Through a plan scoring function, the quality of the BNCT treatment plan can be quantitatively evaluated with less computing resources and computing time, avoiding the impact of plan quality due to differences in designer experience or time costs, and helping to select the optimal BNCT treatment plan based on the patient's actual condition to achieve the best treatment effect.

[0055] Figure 1 FIG. 1 shows a flow chart of the BNCT treatment plan optimization method described in an embodiment of the present application, as shown in FIG. Figure 1 As shown, the BNCT treatment plan optimization method provided in the embodiment of the present application includes steps S1 to S4.

[0056] S1. Obtain a candidate set of BNCT treatment plans.

[0057] In some embodiments, obtaining a candidate set of BNCT treatment plans includes: determining multiple irradiation fields; randomly determining multiple groups of corresponding numbers of irradiation weights, where the sum of the irradiation weights in each group is 1; the irradiation weights correspond to the irradiation time of the irradiation fields; and randomly and exhaustively combining the multiple irradiation fields with the multiple groups of irradiation weights to form multiple candidate BNCT treatment plans.

[0058] In some embodiments, determining multiple irradiation fields includes: determining the geometric center point of the tumor target area; calculating the distance between the contour point of the tumor target area and the geometric center point to determine the nearest irradiation angle; and rotating one circle in the horizontal plane of the tumor target area with the nearest irradiation angle as the center to determine multiple irradiation fields with different irradiation angles.

[0059] Specifically, after the tumor target volume is determined, the positions of the tumor target area contour points can be outlined using the DICOM-RT Struct file. The arithmetic average of these positions is then taken as the geometric center point A of the tumor target volume. The distance between the skin contour points outlined in the DICOM-RT Struct file and point A is then calculated. The point with the smallest distance is used as the starting point, and the vector pointing to point A is the irradiation angle of the skin surface closest to the geometric center point A. The irradiation angle is rotated 360° in the horizontal plane with point A as the center, and any angle is selected and recorded as the irradiation angle of an irradiation field, thereby allowing the selection of irradiation fields with multiple different irradiation angles.

[0060] Furthermore, since the irradiation field usually needs to be selected on the side of the skin closer to the tumor and needs to avoid the treatment bed, the final multiple irradiation fields can be obtained by removing these unreasonable angles.

[0061] Furthermore, after determining the multiple irradiation fields, multiple sets of corresponding irradiation weights are randomly determined, each irradiation weight corresponding to the irradiation time of the irradiation field, and the sum of the irradiation weights of each set is 1. In this manner, the multiple irradiation fields and the multiple sets of irradiation weights are randomly and exhaustively combined to form multiple candidate BNCT treatment plans. That is, in this embodiment of the present application, each candidate BNCT treatment plan actually differs only in the irradiation time of the irradiation field. That is, the candidate set of BNCT treatment plans includes all possible combinations of irradiation times for the multiple irradiation fields.

[0062] For example, consider two irradiation fields with a fixed total irradiation time of 60 minutes. Multiple sets of two irradiation weights are randomly determined in the range [0, 1], with the sum of the two weights being 1. For example, the first set of weights is 0.9 and 0.1; the second set is 0.8 and 0.2, and so on. A total of 10 sets of irradiation weights are randomly determined. These 10 sets of weights are then exhaustively combined with the two irradiation fields to form a total of 20 candidate BNCT treatment plans, which form the candidate set of BNCT treatment plans.

[0063] It should be noted that, when randomly determining multiple groups of corresponding numbers of irradiation weights, the irradiation weight can be set to 0, which means that the irradiation time of a certain irradiation field can be 0.

[0064] S2. Determine a plurality of quality indicators of the BNCT treatment plan and corresponding standard values ​​of the indicators, and determine an indicator scoring function based on the quality indicators and the corresponding standard values ​​of the indicators.

[0065] In some embodiments, the designer of a BNCT treatment plan can design multiple quality indicators and corresponding standard values ​​for evaluating the treatment plan based on treatment needs. For example, the designer can first determine the region of interest (ROI) for the BNCT treatment plan within the patient's body, including the tumor target, organs at risk, or other planned areas. The designer can then determine the radiation standards for the ROI based on the treatment focus and the safety criteria of the plan. This can be used to determine multiple quality indicators for the ROI, such as the minimum radiation indicator for the tumor target, the average radiation indicator for the tumor target, the maximum radiation indicator for the organs at risk, and so on, and set standard values ​​for these quality indicators accordingly. The radiation standard refers to the standard that radiation must meet within the ROI. Taking the minimum radiation indicator for the tumor target as an example, during BNCT treatment, to ensure that all cancer cells are irradiated and treatment efficacy is guaranteed, the minimum radiation for the tumor target should be no less than the prescribed dose.

[0066] Furthermore, designers can set quality indicators in a more detailed manner, such as further dividing the tumor target area and setting the minimum value indicator for the first sub-area of ​​the tumor target area, the minimum value indicator for the second sub-area, etc. Similarly, designers can also subdivide the organs at risk into each organ at risk and set corresponding standard values ​​for each organ at risk. In other words, when the quality indicators are designed in more detail, designers can evaluate the BNCT treatment plan itself in more multi-dimensional ways. In addition, when it is necessary to reduce computing resources or time, designers can also design only key quality indicators.

[0067] It should be understood that the design of any quality indicator is related to the BNCT treatment plan. Generally speaking, for the treatment effectiveness and safety of BNCT treatment plans, minimum indicators for the tumor target and maximum indicators for organs at risk are necessary. The former is to ensure that the tumor target can receive sufficient radiation dose, while the latter is to ensure that normal tissues are not affected by radiation.

[0068] Figure 2 A schematic diagram of the process of determining the index scoring function according to the embodiment of the present application is shown in FIG. Figure 2 As shown, determining the indicator scoring function based on the quality indicator and the corresponding indicator standard value includes steps S21 to S23.

[0069] S21. Obtain an indicator gap between the current value of the quality indicator and the standard value of the indicator.

[0070] S22. Perform reward evaluation or penalty evaluation on the indicator gap based on the evaluation slope.

[0071] S23: Optimizing the indicator gap based on the reward evaluation and the penalty evaluation to obtain the indicator scoring function.

[0072] As previously mentioned, the quality indicator is essentially a quantitative evaluation of the quality of the BNCT treatment plan. When the quality of the BNCT treatment plan is higher, the current value of the quality indicator should be closer to the standard value, and vice versa. In other words, the difference between the current value of the quality indicator and the standard value reflects the evaluation result of the BNCT treatment plan under that quality indicator. To this end, the indicator gap between the current value of the quality indicator and the standard value can be obtained, and the indicator gap can be rewarded or penalized based on the evaluation slope, thereby optimizing the indicator gap to obtain the indicator scoring function.

[0073] When the indicator gap meets the preset gap target, a reward evaluation slope is used to reward the indicator gap; when the indicator gap does not meet the preset gap target, a penalty evaluation slope is used to penalize the indicator gap; wherein the reward evaluation slope is smaller than the penalty evaluation slope. Thus, during the quality indicator optimization process, the indicator scoring function can be used to output an optimal solution that is closer to the standard value of the quality indicator based on reward evaluation or penalty evaluation.

[0074] In some embodiments, the indicator scoring function can be expressed as formula (1):

[0075]

[0076] in, is the score of the i-th quality indicator, is the current value of the i-th quality indicator, is the standard value of the i-th quality indicator, To evaluate the slope.

[0077] S3. Obtain indicator weights of the plurality of quality indicators to obtain a plan scoring function based on the indicator weights and the indicator scoring function.

[0078] As mentioned above, multiple quality indicators can evaluate the treatment of BNCT treatment plans from multiple dimensions. However, during the evaluation, the focus of the BNCT treatment plan may be different due to individual differences and actual treatment conditions. For example, the importance of different organs at risk may vary. Therefore, this application obtains the indicator weights of multiple quality indicators to achieve that when evaluating the entire treatment plan, the scores of important quality indicators occupy a more important part. In some embodiments, the plan scoring function can be expressed by formula (2):

[0079]

[0080] in, Treatment planning for BNCT Rating, is the independent variable vector, represents the irradiation time of the u-th irradiation field, and its domain is ; is the weight of the i-th indicator; is the importance of the i-th indicator.

[0081] In some embodiments, importance can be divided into levels. Designers assess the importance levels of different indicators based on the patient's specific conditions, such as tumor type, size, depth, and location, and calculate corresponding indicator weights based on the importance levels.

[0082] S4. Obtaining a scoring regression model based on the plan scoring function, and iteratively calculating the maximum score in the candidate set based on the scoring regression model and the plan scoring function to obtain a corresponding optimal BNCT treatment plan.

[0083] As can be seen from Equation (2), this application quantitatively scores different candidate BNCT treatment plans using a plan scoring function, thereby selecting the plan with the highest score as the optimal treatment plan. In some embodiments, obtaining the plan with the highest score can be used as an optimization problem to be solved, and the maximum value of the optimization problem can be solved using a Bayesian optimization algorithm to obtain the optimal BNCT treatment plan. Figure 3 A schematic diagram of the process of obtaining a scoring regression model according to an embodiment of the present application is shown in FIG. Figure 3 As shown, obtaining a scoring regression model based on the planned scoring function includes steps S41 to S43.

[0084] S41. Select the mean function and covariance function.

[0085] In some embodiments, the mean function is set to zero mean, and the covariance function is selected as the RBF kernel (Radial Basis Function Kernel) to describe the input point and The similarity can be expressed by formula (3):

[0086]

[0087] in, is the kernel variance; is the length scale.

[0088] S42. Specify a sparse prior distribution for the hyperparameters of the covariance function.

[0089] In the process of constructing a scoring regression model for the objective function, the present application quantifies the uncertainty of the prediction by specifying a prior distribution for the hyperparameters of the covariance function. For dimensions with input (irradiation fields with irradiation times other than 0), the length scale is large, indicating the relevance of the input of this dimension in the planned scoring function. For dimensions without input (irradiation fields with irradiation times of 0), the length scale tends to be concentrated near zero, indicating that this dimension is irrelevant to the planned scoring function. In other words, the length scale can be used to ensure that the scoring regression model effectively captures the structure of the data based on different dimensions (irradiation times of different irradiation fields) and significantly reduces computational complexity. In some embodiments, specifying a sparse prior distribution for the hyperparameters of the covariance function can be expressed by Equation (4):

[0090]

[0091] Among them, the sparse prior distribution of kernel variance is gamma distribution; the sparse prior distribution of length scale is semi-Cauchy distribution, Determines the smoothness of the covariance function in the u-th dimension. The larger it is, the lower the sensitivity of the dimension.

[0092] S43. Fit the planned scoring function based on the mean function, the covariance function, and the sparse prior distribution of the hyperparameters to obtain the scoring regression model.

[0093] In the process of obtaining the scoring regression model, a noise error can be set to further improve the anti-interference ability of the scoring regression model. In some embodiments, the scoring regression model can be expressed by formula (5):

[0094]

[0095] in, is the noise term. In some embodiments, the noise term is set to 10 -6 However, it should be noted that this application is not limited to this.

[0096] Furthermore, from formula (5), we can see that the scoring regression model actually represents the given input In the case of , the prior distribution of the plan scoring function is a multivariate normal distribution with a mean of , the variance is This is a distributional assumption about the plan scoring function itself. That is, in the absence of any samples, the prior distribution is objectively the same for every dimension and does not favor any particular dimension. Assuming that the scores of all BNCT treatment plans in the candidate set follow the prior distribution of the scoring regression model, we can update the model's prior distribution and obtain the posterior distribution by giving some samples (i.e., inputting sample observations). In this case, unknown samples should also follow the model's posterior distribution. Figure 4 FIG. 4 shows a flow chart of iteratively calculating the maximum score value in the candidate set according to an embodiment of the present application, as shown in FIG. Figure 4 As shown, iteratively calculating the maximum score value in the candidate set based on the score regression model and the planned score function includes steps S441 to S444.

[0097] S441. Randomly select multiple BNCT treatment plans from the candidate set as known sample points.

[0098] S442: Calculate the scores of the known sample points using the planned scoring function, and use the known sample points and corresponding scores as a known sample set.

[0099] In some embodiments, assuming that the scores of all BNCT treatment plans in the candidate set obey the prior distribution of the score regression model, a plurality of BNCT treatment plans are randomly selected from the candidate set as known sample points to form a known sample point set. , and calculate the score of known sample points (The known sample points and the corresponding scores constitute the known sample set.) Then the scores of the known sample points obey the prior distribution of the score regression model, that is, , then its probability density function is shown by formula (6):

[0100]

[0101] in, is the mean vector, that is , , which is the variance; I is the unit matrix.

[0102] S443. Iteratively update the posterior distribution of the scoring regression model using the known sample set to obtain a predicted distribution of unscored plans in the candidate set based on the posterior distribution.

[0103] In some embodiments, the known sample point set Y and the score corresponding to each known sample point in the score regression model are As observation data, it is input into the No-U-Turn Sampler (NUTS) together with the sparse prior distribution of the hyperparameters of the covariance function for sampling to update the distribution of the model, that is, to obtain the posterior distribution of the model.

[0104] Then the unrated plans in the candidate set (unknown sample point set The scores of the unknown sample points in the score regression model should also obey the posterior distribution of the score regression model, and the conditional distribution X|Y also obeys the posterior distribution of the score regression model, which can be shown by formula (7):

[0105]

[0106] The conditional mean is , the conditional covariance is Combining Equations (6) and (7), we can obtain the predicted mean and predicted variance of X, that is, the predicted distribution of the unrated plans in the candidate set.

[0107] S444. Iteratively update the known sample set based on the predicted distribution of the unscored plan until a maximum number of iterations is reached, and use the BNCT plan with the maximum score in the known sample set as the optimal BNCT treatment plan.

[0108] In fact, obtaining the optimal plan among the candidate BNCT treatment plans is to find the candidate plan with the largest score. After obtaining the scores of the known sample point set, the scores of the unknown sample points, that is, the unscored plans in the candidate set, can be further calculated and compared with the scores of the known sample points to update the maximum score in the known sample set, and further update the posterior distribution of the scoring regression model, and iterate repeatedly until the maximum score is found. In some embodiments, iteratively updating the known sample set based on the predicted distribution of the unscored plans includes: selecting the optimal plan to be scored based on the predicted distribution of the unscored plans, so as to calculate the score of the optimal plan to be scored by the calculation scoring function; adding the optimal plan to be scored and the corresponding score to the known sample set to iteratively update the known sample set.

[0109] To further reduce computing resources, this application does not need to calculate scores for all unrated plans in the candidate set. Instead, the potential value of the unrated plans is evaluated based on their predicted distribution, and the optimal plan to be scored is selected and its score is calculated. That is, this embodiment of the application uses an acquisition function to evaluate the potential value of the unrated plans based on their predicted distribution, and selects the unrated plan with the highest potential value as the optimal plan to be scored.

[0110] In some embodiments, the acquisition function is used to select and calculate the improvement degree in the unknown sample point set X, and the unknown sample point with the largest improvement degree is selected. To calculate its score, and Add to the known sample point set Y, and compare with the known maximum score in the known sample set Compare and update the maximum score and the posterior distribution of the score regression model. After that, use the updated posterior distribution of the score regression model to re-predict the predicted distribution in the unknown sample point set X, and use the acquisition function again to find , repeat the iteration until the maximum number of iterations is reached and stop, outputting the maximum score at this time and the corresponding BNCT treatment plan is taken as the optimal plan.

[0111] In some embodiments, an EI acquisition function is selected to select the unknown sample point with the greatest degree of improvement from the unknown sample point set X (unrated plans) as the optimal plan to be scored. This acquisition function provides a quantitative evaluation of the degree of improvement by calculating the integral over the predicted distribution of the unknown sample point set X, balancing the exploration and development of unknown sample points and facilitating the efficient optimization process. The EI acquisition function can be expressed as Equation (8):

[0112]

[0113] in, For the unknown sample point set X The predicted mean of For the unknown sample point set X The prediction variance of Expressed as transforming potentially better observations into a standard Gaussian distribution, is the probability density function of the standard normal distribution, is the cumulative distribution function of the standard normal distribution.

[0114] It should be noted that in the process of calculating the score of the BNCT treatment plan through the plan scoring function, it is necessary to use the Monte Carlo method to simulate the dose distribution of multiple irradiation fields to determine whether the treatment plan can meet the quality indicators.

[0115] Figure 5 FIG. 1 shows a logic diagram of the BNCT treatment plan optimization method described in an embodiment of the present application. Figure 5 As shown in the figure, a candidate set of BNCT treatment plans is obtained, and multiple plans are selected from them as known sample points. Their scores are calculated to form the known sample point set. Next, the prior distribution of the scoring regression model is obtained, and the posterior distribution of the model is updated based on the known sample points to predict the predicted distribution of the unscored plans in the candidate set. After obtaining the predicted distribution, the degree of improvement of the unscored plans is calculated, and the unscored plan with the greatest degree of improvement is selected as the plan to be scored. Its score is calculated, thereby updating the known sample point set. This process is repeated until the maximum number of iterations is reached. The maximum score in the known sample point set is output, and the corresponding plan is selected as the optimal plan.

[0116] It should be noted that, in addition to the number of iterations, a scoring threshold can also be designed as a stopping condition, that is, the iteration is stopped when the scoring threshold is reached, and the maximum score at that time is output.

[0117] Therefore, this application obtains an indicator scoring function by setting multiple quality indicators and indicator standard values, and further obtains a plan scoring function based on different weights, which helps to select the optimal BNCT treatment plan based on the patient's actual situation to achieve the best treatment effect. And the plan scoring function helps to quantitatively evaluate the quality of the BNCT treatment plan, avoiding the impact of plan quality on designer experience differences or time costs. In addition, this application iteratively calculates the maximum score based on the plan scoring function, and can set the number of iterations according to actual conditions. When the number of iterations is reached, the BNCT treatment plan with the maximum score is selected as the optimal plan. This significantly reduces computing resources, can more efficiently select the optimal plan based on actual conditions, and reduces optimization time.

[0118] The protection scope of the BNCT treatment plan optimization method of the embodiment of the present application is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, subtracting, or replacing steps in the prior art based on the principles of the present application are included in the protection scope of the present application.

[0119] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0120] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical concepts disclosed in this application shall be covered by the claims of this application.

Claims

1. A BNCT treatment plan optimization method, characterized in that: The method comprises: Obtaining a candidate set of BNCT treatment plans; the candidate set of BNCT treatment plans includes all possible irradiation time combinations of multiple irradiation fields; Determining a plurality of quality indicators of the BNCT treatment plan and corresponding standard values ​​of the indicators, and determining an indicator scoring function based on the quality indicators and the corresponding standard values ​​of the indicators; Obtaining indicator weights of a plurality of the quality indicators to obtain a plan scoring function based on the indicator weights and the indicator scoring function; A scoring regression model is obtained based on the plan scoring function, and a maximum score in the candidate set is iteratively calculated based on the scoring regression model and the plan scoring function to obtain a corresponding optimal BNCT treatment plan.

2. The BNCT treatment plan optimization method according to claim 1, characterized in that: Determining an indicator scoring function based on the quality indicator and the corresponding indicator standard value includes: Obtaining an indicator gap between a current value of the quality indicator and a standard value of the indicator; Perform reward evaluation or penalty evaluation on the indicator gap based on the evaluation slope; The indicator gap is optimized based on the reward evaluation and the penalty evaluation to obtain the indicator scoring function.

3. The BNCT treatment plan optimization method according to claim 2, characterized in that: The reward evaluation or penalty evaluation based on the evaluation slope of the indicator gap includes: When the indicator gap meets the preset gap target, the indicator gap is rewarded and evaluated using the reward evaluation slope; When the indicator gap does not meet the preset gap target, a penalty evaluation slope is used to perform a penalty evaluation on the indicator gap; wherein the reward evaluation slope is smaller than the penalty evaluation slope.

4. The BNCT treatment plan optimization method according to claim 1, characterized in that: Multiple quality indicators were used to determine the quality of BNCT treatment plans, including: Determine the region of interest for the BNCT treatment plan, which includes the tumor target volume, organs at risk, and planning area; A plurality of quality indicators of the region of interest are determined based on the radiation standard of the region of interest.

5. The BNCT treatment plan optimization method according to claim 1, characterized in that: Acquiring a scoring regression model based on the planned scoring function includes: Select the mean function and covariance function; specifying a sparse prior distribution for hyperparameters of the covariance function; The planned scoring function is fitted based on the mean function, the covariance function, and the sparse prior distribution of the hyperparameters to obtain the scoring regression model.

6. The BNCT treatment plan optimization method according to claim 1, characterized in that: Iteratively calculating the maximum score value in the candidate set based on the score regression model and the planned score function includes: Randomly selecting multiple BNCT treatment plans from the candidate set as known sample points; Calculate the scores of the known sample points using the planned scoring function, and use the known sample points and corresponding scores as a known sample set; Iteratively updating the posterior distribution of the scoring regression model using the known sample set to obtain a predicted distribution of unscored plans in the candidate set based on the posterior distribution; The known sample set is iteratively updated based on the predicted distribution of the unscored plans until a maximum number of iterations is reached, and the BNCT treatment plan with the largest score in the known sample set is determined as the optimal plan.

7. The BNCT treatment plan optimization method according to claim 6, characterized in that: Iteratively updating the known sample set based on the predicted distribution of the unrated plan includes: Selecting an optimal plan to be scored based on the predicted distribution of the unscored plans, and calculating a score of the optimal plan to be scored using the scoring function; The optimal plan to be scored and the corresponding score are added to the known sample set to iteratively update the known sample set.

8. The BNCT treatment plan optimization method according to claim 7, characterized in that: Selecting the optimal plan to be scored based on the predicted distribution of the unscored plans includes: evaluating the potential value of the unrated plan based on the predicted distribution of the unrated plan by an acquisition function; The unrated plan with the highest potential value is selected as the optimal plan to be rated.

9. The BNCT treatment plan optimization method according to claim 1, characterized in that: The candidate set for obtaining BNCT treatment plans includes: Determine multiple irradiation fields; Randomly determining a plurality of groups of corresponding numbers of irradiation weights, wherein the sum of the irradiation weights in each group is 1; the irradiation weights correspond to the irradiation time of the irradiation field; A plurality of irradiation fields and a plurality of sets of irradiation weights are randomly and exhaustively combined to form a plurality of candidate BNCT treatment plans.

10. The BNCT treatment plan optimization method according to claim 9, characterized in that: Determining multiple radiation fields includes: Determine the geometric center point of the tumor target area; Calculating the distance between the contour point of the tumor target area and the geometric center point to determine the closest irradiation angle; Taking the closest irradiation angle as the center, a rotation is performed within the horizontal plane of the tumor target area to determine multiple irradiation fields with different irradiation angles.

Images