Target spot optimization method and device and storage medium

Abstract

The invention relates to a target spot optimization method and device and a storage medium, and the method comprises the steps: firstly obtaining the head modeling data, brain function partition, intervention process information and intervention curative effect information of a modeling object, then carrying out the calculation of a cerebral cortex induced electric field in an intervention process based on the head modeling data corresponding to the modeling object, and carrying out the calculation of a cerebral cortex induced electric field. Obtaining accumulated cerebral cortex induced electric field distribution in the intervention process, obtaining an induced electric field function connection matrix according to the accumulated cerebral cortex induced electric field distribution and the brain function partition, and finally establishing a target optimization model by utilizing the data of the induced electric field function connection matrix in combination with the intervention curative effect information. By more accurately determining the stimulation pose and parameters, accurate feedback of individual differences of application objects is realized, so that the overall curative effect of intervention is improved.

Description

Technical Field

[0001] This application relates to the field of medical imaging technology, and in particular to target optimization methods, devices and storage media. Background Technology

[0002] Transcranial magnetic stimulation (TMS) is a non-invasive neuromodulation technique that generates electrical currents in localized areas of the cerebral cortex to temporarily activate or inhibit neurons in those areas. During TMS intervention, the TMS coil needs to be precisely positioned at a specific location on the patient's head to ensure that the current effectively acts on the target cerebral cortex region.

[0003] In related technologies, the placement of the TMS coil is mainly determined by the patient's skull bony landmarks and the doctor's experience. However, the intervention effect varies greatly among different patients by relying solely on skull bony landmarks and doctor's experience, and the accuracy and consistency of each treatment cannot be guaranteed.

[0004] Therefore, existing technologies lack a more precise coil positioning method to improve the overall efficacy of interventions. Summary of the Invention

[0005] To address or partially address the problems existing in related technologies, this application provides a target optimization method for achieving accurate feedback on individual differences. By acquiring brain modeling data and brain functional partitions for each individual, the pose and parameters of stimuli can be precisely determined.

[0006] The first aspect of this application provides a target optimization method, the method comprising:

[0007] Obtain the modeling data corresponding to the modeling object; the modeling object includes n individual objects, where n≥2; the modeling data includes head modeling data, brain functional areas, intervention process information and intervention efficacy information corresponding to each individual object;

[0008] Based on the head modeling data of the modeling object, the induced electric field of the cerebral cortex is calculated during the intervention process to obtain the cumulative induced electric field distribution of the cerebral cortex during the intervention process.

[0009] Based on the cumulative distribution of induced electric fields in the cerebral cortex and the brain functional regions, an induced electric field functional connectivity matrix is ​​obtained; using the induced electric field functional connectivity matrix data combined with the intervention efficacy information, a target optimization model is established.

[0010] Optionally, obtain the data to be modeled corresponding to the modeling object, including:

[0011] Acquire MRI images of the modeling object, including T1 images and rest-fMRI images;

[0012] The T1 image is input into the feature extraction model to obtain the head model data of the modeling object;

[0013] The rest-fMRI image is input into the feature extraction model to obtain the brain functional regions of the modeled object.

[0014] Optionally, the intervention process information includes the current pose of the TMS coil, the time point of TMS pulse generation during the TMS intervention process, and the TMS output intensity.

[0015] Optionally, obtaining the induced electric field functional connectivity matrix based on the accumulated induced electric field distribution in the cerebral cortex and the brain functional regions includes:

[0016] The cumulative induced electric field distribution of the cerebral cortex of multiple individuals is binarized and segmented according to a preset threshold to obtain the target area of ​​action;

[0017] Extract the rest-fMRI signal of the target area and calculate the average value to obtain the average signal of the target area.

[0018] The induced electric field functional connection matrix is ​​constructed based on the average signal in the target area.

[0019] Optionally, constructing the induced electric field functional connectivity matrix based on the average signal of the target area includes:

[0020] Calculate the correlation coefficient between the target area and other brain functional regions;

[0021] A functional connectivity matrix is ​​constructed based on the correlation coefficient, wherein each row and each column of the induced electric field functional connectivity matrix corresponds to a different brain functional region, and the element values ​​in the induced electric field functional connectivity matrix are used to characterize the functional connectivity strength between the corresponding brain functional regions.

[0022] Optionally, the step of establishing a target optimization model by combining the induced electric field functional connection matrix data with the intervention efficacy information includes:

[0023] The induced electric field functional connection matrix is ​​divided according to preset rules to obtain multiple functional subnets;

[0024] For each of the aforementioned functional subnetworks, the functional connectivity strength between brain regions within the functional subnetwork is calculated;

[0025] A target prediction model is established based on the functional connectivity strength of each functional subnet.

[0026] Optionally, calculating the functional connectivity strength between brain regions within the functional subnetwork includes:

[0027] Calculate the Pearson correlation coefficients between brain regions within the aforementioned functional subnetworks;

[0028] The functional connectivity strength between brain regions within the functional subnetwork is calculated based on the Pearson correlation coefficient.

[0029] Optionally, the method further includes:

[0030] Based on the target optimization model, the target coil pose and / or target output intensity during the intervention process are adjusted to optimize the intervention effect.

[0031] A second aspect of this application provides a target optimization apparatus, comprising:

[0032] The data acquisition module is used to obtain the data to be modeled corresponding to the modeling object; the modeling object includes n individual objects, where n≥2; the data to be modeled includes head modeling data, brain functional areas, intervention process information and intervention efficacy information for each individual.

[0033] The calculation module is used to calculate the cortical induced electric field of the intervention process based on the head modeling data corresponding to the modeling object, to obtain the cumulative cortical induced electric field distribution of the intervention process, and to obtain the induced electric field functional connectivity matrix according to the cumulative cortical induced electric field distribution and the brain functional partition.

[0034] The model building module is used to establish a target optimization model by combining the induced electric field function connection matrix data with the intervention efficacy information.

[0035] A third aspect of this application provides a non-transitory machine-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method described above.

[0036] Therefore, the target optimization method provided in this application first obtains head modeling data, brain functional partitions, intervention process information, and intervention efficacy information for multiple individuals. Then, based on the head modeling data of multiple individuals, the induced electric field of the cerebral cortex is calculated during the intervention process to obtain the cumulative induced electric field distribution of the cerebral cortex during the intervention process for multiple individuals. Next, based on the cumulative induced electric field distribution of the cerebral cortex and the brain functional partitions of multiple individuals, an induced electric field functional connectivity matrix is ​​obtained. Finally, the induced electric field functional connectivity matrix data is combined with the intervention efficacy information to establish a target optimization model. By more accurately determining the pose and parameters of the stimulus, accurate feedback on the differences among multiple individuals can be achieved, thereby improving the overall efficacy of the intervention.

[0037] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0038] The above and other objects, features and advantages of this application will become more apparent from the following description of exemplary embodiments of this application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of this application.

[0039] Figure 1 This is a schematic flowchart illustrating the target optimization method in an embodiment of this application;

[0040] Figure 2 This is a schematic diagram of the model structure shown in the embodiments of this application;

[0041] Figure 3 This is a structural diagram of the TMS inductive electric field functional connection matrix shown in the embodiments of this application;

[0042] Figure 4 This is a schematic diagram of the target optimization device shown in the embodiments of this application;

[0043] Figure 5 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation

[0044] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0045] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0046] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0047] Transcranial magnetic stimulation (TMS) is a non-invasive neuromodulation technique. During intervention, an electromagnetic coil is placed above the scalp, generating a rapidly changing magnetic field. This magnetic field penetrates the scalp and skull, inducing electrical currents in the cerebral cortex, thereby modulating the activity of local neurons. TMS technology is widely used in researching brain function and treating various neurological and psychiatric disorders, such as depression, anxiety, migraines, and chronic pain.

[0048] During TMS intervention, the brain regions requiring stimulation are first identified, including the target location. The specific brain regions to be stimulated are determined based on the intervention objectives and the patient's individual condition. For example, for patients with depression, the left prefrontal cortex is commonly used as the target. Then, based on the identified target, the TMS coil is placed at a specific location on the patient's scalp, and the intensity, frequency, and duration of stimulation are set according to the intervention protocol.

[0049] Understandably, TMS coils can be used during intervention. These coils can be devices used to generate magnetic fields and stimulate the cerebral cortex, or they can be devices used to generate low currents for continuous stimulation, thereby modulating brain activity. For example, transcranial electrical stimulation (tES) delivers weak currents to the brain through electrodes on the scalp to modulate neuronal activity in the cerebral cortex. Unlike transcranial magnetic stimulation, tES directly stimulates the brain with electrical current, without the need for magnetic field conversion, and can be configured according to actual needs; further details will not be elaborated here.

[0050] Current technologies primarily rely on skull bony landmarks and physician experience to determine the placement of TMS coils for stimulation of specific cortical regions. Furthermore, some existing technologies are beginning to utilize TMS navigation systems and brain imaging techniques, combining the patient's brain MRI or CT data to provide more precise coil positioning. These technologies have, to some extent, improved the accuracy and efficacy of TMS, allowing intervention programs to better adapt to individual patient anatomical differences and the complexity of brain structures.

[0051] However, methods relying on skull bony landmarks and physician experience show significant differences in effectiveness among different patients and cannot guarantee precise consistency in each intervention. Secondly, while existing navigation systems offer more precise coil positioning, they still cannot completely eliminate the impact of individual patient differences on intervention outcomes. In practical applications, the aforementioned technical solutions still encounter situations where some patients respond well, while others respond slowly or even unresponsively.

[0052] Please see Figure 1 , Figure 1 This is a flowchart illustrating the target optimization method in an embodiment of this application.

[0053] This application provides a target optimization method, the method comprising:

[0054] S100. Obtain the data to be modeled corresponding to the modeling object; the modeling object includes n individual objects, where n≥2; the data to be modeled includes head modeling data, brain functional areas, intervention process information and intervention efficacy information corresponding to each individual object.

[0055] Obtain MRI brain images of the modeling object, including T1 images and rest-fMRI images.

[0056] Based on T1 images and rest-fMRI images, individual head modeling data and corresponding brain functional regions of the modeling objects were obtained.

[0057] See Figure 2 , Figure 2 This is a schematic diagram of the model structure shown in the embodiments of this application.

[0058] In the figure, a) is the T1 image input, b) is the rest-fMRI image input, c) is the individualized head model constructed based on the T1 image, d) is the brain functional division based on individual rest-fMRI data, e) is the cumulative cerebral cortical electric field intensity distribution calculated based on TMS navigation recording and individual head model, f) is the TMS induced electric field functional connectivity matrix calculated based on the cumulative electric field distribution and brain functional division, g) the cerebral cortical induced electric field of multiple individual head models during the simulation process, and the TMS target point position (x, y, z) and pose (yaw, pitch, roll) are determined.

[0059] In this embodiment, the modeling object includes n individual objects, where n ≥ 2. In specific implementation, the number of individual objects n needs to be greater than the number of parameters in the model. The data to be modeled includes head modeling data, brain functional areas, intervention process information, and intervention efficacy information corresponding to each individual object, which are used to subsequently construct a target optimization model.

[0060] In this embodiment, the T1 image is an imaging modality in magnetic resonance imaging (MRI) used to describe the longitudinal relaxation time of tissues. Relaxation time refers to the time required for protons in tissue to return to equilibrium after being excited by a radiofrequency pulse. This technique is particularly useful for generating high-contrast images of human tissues. In MRI, the T1 value of tissue affects the brightness and contrast of the image. Tissues with shorter T1 values ​​appear brighter in the image, while tissues with longer T1 values ​​appear darker. This contrast can help doctors diagnose different lesions and tissue types.

[0061] Optional image acquisition also includes T2 images, DTI (Diffusion Tensor Imaging) images, and fMRI (functional Magnetic Resonance Imaging) images.

[0062] T2 images are used to describe the lateral relaxation time of tissues and are used to examine the internal structures and tissue conditions of the human body. T2 images are mainly used to observe the water content and spin velocity of tissues, thereby identifying lesions and abnormalities in different tissues, and providing better contrast between gray and white matter, enhancing the resolution of different brain tissues.

[0063] Understandably, the recon-all function in the FreeSurfer software package is used to reconstruct the morphology of the cerebral cortex of the tested subject and perform brain tissue segmentation. Then, the mri2mesh function in the SimNIBS software package is used to reconstruct the morphology of the scalp skin, generating an individualized head model.

[0064] Rest-state functional magnetic resonance imaging (fMRI) is a technique used to study brain activity. Unlike task-based fMRI, rest-state fMRI records and acquires functional network structures and functions in a resting state while the subject is at rest or not performing a specific task.

[0065] Functional connectivity was measured by calculating the temporal correlation of BOLD signals between different brain regions based on the degree of coordinated activity between different brain regions obtained from rest-fMRI images. A set of brain regions demonstrating a consistent activity model of the brain at rest was registered with T1 structural images and normalized to template space.

[0066] In practical applications, T1-weighted images and rest-fMRI scans are performed on multiple individuals. T1-weighted images are used to acquire high-resolution brain anatomical images, while rest-fMRI records the brain functional activity of multiple individuals at rest. The acquired data includes T1-weighted images (resolution no higher than 1mm*1mm*1mm) and at least 20 minutes of rest-fMRI data.

[0067] Temporal and head movement corrections were performed on the rest-fMRI data, and the fMRI images were aligned with high-resolution T1 images. BOLD signals from the cerebral cortex were extracted, and regression analysis was performed to remove confounding factors such as head movement and heart rate. The processed images were then registered to a standard brain template to obtain individual brain images in the template space. Brain functional partitioning techniques were then applied to these individual brain images to obtain functional partitions of the individual brain.

[0068] In this embodiment of the application, the intervention process information can refer to all relevant data recorded during the TMS intervention process, which is used to analyze and evaluate the accuracy and effect of the intervention. This information may include the TMS coil position, which is the position of the coil relative to the heads of multiple individuals each time TMS is applied; the TMS coil posture; the TMS pulse time; and the TMS pulse output intensity.

[0069] Understandably, TMS intervention efficacy information refers to data assessing the effects and efficacy of TMS intervention on multiple individuals. This includes pre-intervention assessment values. Before the TMS intervention begins, baseline assessments of relevant functions are conducted on multiple individuals, using various scales or tests such as the Hamilton Depression Rating Scale (HRSD) and the Beck Depression Rating Scale (BDI). After the TMS intervention, the same assessment is performed on each individual. By comparing the assessment values ​​before and after the intervention, the intervention effect can be determined. Based on this, the degree of improvement in the condition or function of multiple individuals can be calculated by observing changes in assessment values ​​before and after the intervention.

[0070] S101. Based on the head model data of the modeling object, calculate the induced electric field of the cerebral cortex during the TMS intervention process to obtain the cumulative induced electric field distribution of the cerebral cortex during the intervention process.

[0071] In this embodiment, a pre-created individualized head model and TMS pulse data, TMS coil position and orientation are acquired, and a corresponding simulation engine, such as SimNIBS, is selected to calculate the electric field distribution. SimNIBS, as an open-source software package, can be used to simulate and analyze the electric field distribution in the brain during non-invasive brain stimulation. It can combine relevant data from structural and functional images for analysis to obtain a more comprehensive evaluation of the stimulation effect.

[0072] Understandably, by using a simulation engine to calculate the electric field of a single TMS pulse, the electric field distribution at each grid node at a specific time point can be obtained through Maxwell's equations. At each node of the 3D mesh model, the electric field strength and direction at that point are calculated, and the electric field strengths of each grid node are accumulated to obtain the cumulative electric field distribution throughout the intervention process. A 3D distribution map is then generated, displaying the cumulative induced electric field strength in various regions of the cerebral cortex. Electromagnetic field simulation software such as Ansys and Comsol can be used for this purpose.

[0073] In this embodiment, based on the head model data of the modeled object, and the intensity and pose information of the TMS coil during TMS pulse release, pulse-by-pulse cortical induction calculations can be performed to further optimize the TMS treatment plan and improve the treatment effect. By accurately calculating the distribution of the induced electric field in the cerebral cortex, the range and intensity of the TMS pulse in the brain can be better understood, thereby adjusting the position and orientation of the TMS coil to achieve the best stimulation effect.

[0074] Furthermore, a three-dimensional distribution map of the accumulated induced electric field can visually demonstrate the stimulation intensity received by different brain regions, helping to assess the coverage and potential side effects of TMS treatment for each individual. For example, if the induced electric field intensity in a certain brain region is too high, it may cause discomfort or side effects. The position or output intensity of the TMS coil can be adjusted accordingly to avoid overstimulation.

[0075] Understandably, pulse-by-pulse computation can provide extremely high time resolution. For example, in fields such as medical imaging, physical experiments, and communication systems, pulse-by-pulse computation can ensure precise synchronization of data acquisition and processing, thereby improving the overall system performance and reliability.

[0076] S102. Based on the cumulative induced electric field distribution in the cerebral cortex and the brain functional regions, the TMS induced electric field functional connectivity matrix is ​​obtained.

[0077] In this embodiment, the cumulative induced electric field distribution can reflect the induced electric field intensity at different locations in the brains of multiple individuals, and further analyze the functional connections between different brain regions of multiple individuals based on the obtained brain functional partitions, which facilitates the understanding of the influence of TMS on brain functional networks.

[0078] First, based on the cumulative induced electric field intensity distribution, a preset electric field intensity quantile is selected as the threshold. That is, the region of interest (ROI) is defined as the area of ​​cumulative induced electric field intensity with an intensity greater than the threshold quantile. The cumulative induced electric field intensity can be binarized and segmented into two classes. Points above the threshold are marked as 1, and the rest are marked as 0. The electric field intensity distribution is transformed into a binary image, and only the region with the highest electric field intensity is retained as the main area of ​​effect of the TMS.

[0079] Understandably, regions with electric field strength exceeding a preset threshold are marked as Regions of Interest (ROIs). These regions are where TMS primarily operates, reflecting the intensity and extent of the electric field's influence in the cerebral cortex.

[0080] For a marked Region of Interest (ROI), the average rest-state fMRI signal within that region is calculated. Rest-state fMRI signals indicate neural activity in different brain regions during rest. The average fMRI signal of the ROI is obtained by averaging the fMRI information of all voxels within the ROI.

[0081] The calculated average signal was used as a seed signal, and correlation analysis was performed with the average signals from other functional areas of the brain. By calculating the correlation coefficient between the average signal and the seed signal for each functional area, a TMS induced electric field functional connectivity matrix was constructed.

[0082] In practical applications, the range of the cumulative induced electric field strength can be from 0 to 100, with a preset threshold electric field strength of 98%, meaning a range of 90. Figure 3 This is a structural diagram of the TMS induced electric field functional connection matrix shown in an embodiment of this application; in the diagram, 301 represents the region where the electric field intensity exceeds a preset threshold, and 302 represents the region where the electric field intensity does not exceed the preset threshold. The cumulative induced electric field intensity after determining the preset threshold is binarized, and regions where the cumulative induced electric field intensity is greater than the 98th percentile are marked as 1, while the remaining regions are marked as 0, thus obtaining the main region of effect (ROI) of the TMS.

[0083] For example, the cumulative induced electric field intensity distribution is as follows:

[0084]

[0085] If the 98% intensity quantile is 90, then the binarization result is:

[0086]

[0087] Further, the average signal of the ROI is calculated. In the binarization result formula (2), all regions (ROIs) marked as 1 are extracted, and the average signal value of the rest-fMRI data within the region is calculated. Assuming the rest-fMRI data is:

[0088]

[0089] The average signal of the ROI, corresponding to the binarized labeling of the cumulative induced electric field intensity distribution, is:

[0090]

[0091] The average signal of the ROI is used as the seed signal, and its correlation with the average signal within each brain functional region is calculated. For each brain functional region, the average of all rest-fMRI signals within the region is calculated. For example, the rest-fMRI data for brain functional region 1 are as follows:

[0092]

[0093] The average signal in brain functional area 1 was 12.5.

[0094] Similarly, the rest-fMRI data for brain functional region 2 are as follows:

[0095]

[0096] The average signal in brain functional region 2 was 32.5.

[0097] Calculate the Pearson correlation coefficient between the average signal of the ROI and the average signal of each functional zone to construct the functional link matrix. The correlation coefficient is calculated as follows:

[0098]

[0099] In equation (7), x is the average signal of the ROI, and y is the average signal of each functional zone.

[0100] By filling the matrix elements with the above correlation values, the TMS induced electric field functional connection matrix can be obtained.

[0101] Based on the TMS induced electric field functional linkage matrix obtained from S102 and the intervention effect information of multiple individuals, a target optimization model is constructed to predict the intervention effect of TMS.

[0102] In this embodiment, the TMS induced electric field functional connectivity matrix is ​​a symmetric matrix that can represent the connection strength between different brain functional regions. The functional connectivity matrix is ​​expanded into a one-dimensional vector, and the elements of the upper or lower triangular part are extracted. Since the matrix is ​​symmetric, only the data of its upper or lower triangular part can be used to avoid redundancy problems, thereby reducing the input dimension of the model.

[0103] Specifically, in practical applications, let the TMS induced electric field functional link matrix be M, with a dimension of n×n. Extract the upper triangular part of matrix M and flatten it into a one-dimensional vector with a length of n(n-1) / 2, so that the TMS induced electric field functional link matrix of each individual is transformed into a one-dimensional vector.

[0104] For example, the TMS induced electric field functional connection matrix M is:

[0105]

[0106] Extract the upper triangular portion to form a vector x = [m 12 ,m 13 ,m 23 ].

[0107] In this embodiment of the application, a ridge regression model can be constructed. The ridge regression model is a linear regression model with an L2 regularization term, which can be used to prevent overfitting and is suitable for processing multidimensional input data.

[0108] Using the functional connectivity vector x of the multiple individuals as input features and the HRSD improvement score y as the target value, the ridge regression algorithm is used to fit the model, and the regression coefficient ω and intercept term b are calculated to form the regression equation:

[0109] y=ω T x+b....................(9)

[0110] For example, when selecting 50 individuals for depression testing, the vector obtained by expanding the TMS-induced electric field function connectivity matrix for each individual and the difference between their HRSD scores are used as model inputs. The dataset is divided into training and testing sets. The ridge regression model is trained using the training set data, the model parameters are optimized, and an appropriate regularization strength is selected accordingly. For example, 80% of the dataset can be used as the training dataset, and 20% as the testing set.

[0111] Fit the model using the ridge regression algorithm on the training set:

[0112]

[0113] In equation (8), x i It is the functional connection vector of the i-th detected object, y iω represents the corresponding HRSD improvement score, ω is the regression coefficient, and λ is the regularization parameter.

[0114] In this embodiment of the application, a target optimization model is established by combining the TMS induced electric field functional connection matrix data with the intervention efficacy information, including:

[0115] The induced electric field functional connection matrix is ​​divided according to preset rules to obtain multiple functional subnets;

[0116] For each of the aforementioned functional subnetworks, the functional connectivity strength between brain regions within the functional subnetwork is calculated;

[0117] A target prediction model is established based on the functional connectivity strength of each functional subnet.

[0118] Based on the above embodiments, calculating the functional connectivity strength between brain regions within the functional subnetwork includes:

[0119] Calculate the Pearson correlation coefficients between brain regions within the aforementioned functional subnetworks;

[0120] The functional connectivity strength between brain regions within the functional subnetwork is calculated based on the Pearson correlation coefficient.

[0121] In this embodiment of the application, the functional subnet can characterize a set of specific brain regions affected during transcranial magnetic stimulation, including the directly stimulated regions and other functionally related brain regions.

[0122] Functional connectivity between brain regions can be assessed by measuring the correlation of neural activity between brain regions, where the Pearson correlation coefficient can be selected for calculation. In this application, the average signal of each brain region within the functional subnetwork is extracted from rest-fMRI data, and the signal of each brain region represents the neural activity of that brain region during stimulation.

[0123] For example, given brain regions A, B, and C, at each time point t, the fMRI signal of each brain region is extracted, and the correlation coefficient γ between brain regions A and B is calculated. AB The correlation coefficient γ between brain region A and brain region C AC The correlation coefficient γ between brain regions B and C BC By filling the Pearson correlation coefficient values ​​between all brain regions into the matrix, the functional connectivity matrix is ​​obtained as follows:

[0124]

[0125] In Equation (9), 1 represents the complete correlation of each brain region itself, and the elements outside the diagonal represent the strength of functional connectivity between different brain regions.

[0126] The foregoing embodiments describe the process of target optimization in practical applications, including the method for constructing the target optimization model, which includes:

[0127] In model construction, group intervention data was used as the training set, and statistical and cluster analysis-based methods were employed to supplement the target optimization model. First, the collected group intervention data were sorted according to intervention effects and divided into three groups with approximately equal data sizes, representing good, moderate, and poor intervention effects, respectively. Next, statistical tests were performed on the experimental data from the good and poor intervention effect groups to identify TMS induced electric field functional connectivity matrix elements related to intervention effects; these elements were defined as efficacy-related TMS electric field functional connectivity features. Then, the mean values ​​of efficacy-related TMS electric field functional connectivity features were calculated for the good and poor intervention effect groups to determine the category center feature points. During target planning, for each specific pose, the corresponding TMS induced electric field functional connectivity matrix at that pose can be obtained. By extracting efficacy-related TMS electric field functional connectivity features from the matrix and calculating the distances between these features and the two category center feature points (good and poor efficacy), the therapeutic effect of the target can be predicted, thereby achieving target optimization.

[0128] In this embodiment of the application, the method further includes:

[0129] Based on the target optimization model, the target coil pose and / or target output intensity during the intervention process are adjusted to optimize the intervention effect.

[0130] In one application scenario, target optimization is performed on candidate target regions of an application object. A grid search method is used to calculate and predict the improvement rate based on the current position and attitude combination of the TMS coil of the detection object. In this embodiment, the improvement rate is a number between 0 and 1, with higher values ​​indicating better efficacy. For example, the target position has three values: 10mm, 15mm, and 20mm (actually three coordinates: x, y, z; only one dimension is shown here), and the target attitude has three values: -5 degrees, 0 degrees, and -5 degrees (actually three angles: pitch, yaw, and roll; only one is shown here). The improvement rate results predicted by the target optimization model are shown in Table 1 below. After obtaining the target optimization model prediction results for various combinations, the combination with the highest expected improvement is selected as the preferred treatment plan. In this example, the combination of position 20mm and attitude -5 degrees is selected as the preferred treatment plan.

[0131]

[0132] Table 1

[0133] Corresponding to the aforementioned application function implementation method embodiments, this application also provides a target optimization device, an electronic device, and corresponding embodiments.

[0134] Figure 4 This is a schematic diagram of the target optimization device shown in the embodiments of this application. See also... Figure 4 Target optimization device, comprising:

[0135] Data acquisition module 41 is used to obtain the data to be modeled corresponding to the modeling object; the modeling object includes n individual objects, where n≥2; the data to be modeled includes head modeling data, brain functional areas, intervention process information and intervention efficacy information corresponding to each individual.

[0136] The calculation module 42 is used to calculate the induced electric field of the cerebral cortex during the intervention process based on the head modeling data corresponding to the modeling object, to obtain the cumulative induced electric field distribution of the cerebral cortex during the intervention process, and to obtain the induced electric field functional connectivity matrix based on the cumulative induced electric field distribution of the cerebral cortex and the brain functional partitions.

[0137] The model building module 43 is used to establish a target optimization model by combining the induced electric field function connection matrix data with the intervention efficacy information.

[0138] The data acquisition module is used for:

[0139] Acquire MRI images of the modeling object, including T1 images and rest-fMRI images;

[0140] The T1 image is input into the feature extraction model to obtain the head model data of the modeling object;

[0141] The rest-fMRI image is input into the feature extraction model to obtain the brain functional regions of the modeled object.

[0142] Optionally, the intervention process information includes the current pose of the TMS coil, the time point of TMS pulse generation during the TMS intervention process, and the TMS output intensity.

[0143] The calculation module is used for:

[0144] The cumulative induced electric field distribution in the cerebral cortex is binarized and segmented according to a preset threshold to obtain the target area of ​​action;

[0145] Extract the rest-fMRI signal of the target area and calculate the average value to obtain the average signal of the target area.

[0146] The induced electric field functional connection matrix is ​​constructed based on the average signal in the target area.

[0147] Optionally, constructing the induced electric field functional connectivity matrix based on the average signal of the target area includes:

[0148] Calculate the correlation coefficient between the target area and other brain functional regions;

[0149] A functional connectivity matrix is ​​constructed based on the correlation coefficient, wherein each row and each column of the induced electric field functional connectivity matrix corresponds to a different brain functional region, and the element values ​​in the induced electric field functional connectivity matrix are used to characterize the functional connectivity strength between the corresponding brain functional regions.

[0150] The model building module is used for:

[0151] The induced electric field functional connection matrix is ​​divided according to preset rules to obtain multiple functional subnets;

[0152] For each of the aforementioned functional subnetworks, the functional connectivity strength between brain regions within the functional subnetwork is calculated;

[0153] A target prediction model is established based on the functional connectivity strength of each functional subnet.

[0154] Optionally, calculating the functional connectivity strength between brain regions within the functional subnetwork includes:

[0155] Calculate the Pearson correlation coefficients between brain regions within the aforementioned functional subnetworks;

[0156] The functional connectivity strength between brain regions within the functional subnetwork is calculated based on the Pearson correlation coefficient.

[0157] Optionally, the method further includes:

[0158] Based on the target prediction model, the target coil pose and / or target output intensity during the intervention process are adjusted to optimize the intervention effect.

[0159] Therefore, the target optimization device provided in this application first obtains head modeling data, brain functional partitions, intervention process information, and intervention efficacy information of the modeling object. Then, based on the head modeling data of the modeling object, it calculates the cortical induced electric field of the intervention process to obtain the cumulative cortical induced electric field distribution of the intervention process. Then, based on the cumulative cortical induced electric field distribution and the brain functional partitions, it obtains the induced electric field functional connectivity matrix. Finally, it uses the induced electric field functional connectivity matrix data combined with the intervention efficacy information to establish a target optimization model. By more accurately determining the pose and parameters of the stimulus, it achieves accurate feedback on multiple individual differences, thereby improving the overall efficacy of the intervention.

[0160] Figure 5 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application.

[0161] See Figure 5The electronic device 500 includes a memory 510 and a processor 520.

[0162] The processor 520 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0163] Memory 510 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by the processor 520 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 510 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, memory 510 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital multifunction optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, an ultra-high-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0164] The memory 510 stores executable code, which, when processed by the processor 520, can cause the processor 520 to execute part or all of the methods described above.

[0165] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0166] Alternatively, this application may be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of an electronic device (or server, etc.), causes the processor to perform part or all of the steps of the methods described above according to this application.

[0167] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A target optimization method, characterized in that, The method includes: Obtain the modeling data corresponding to the modeling object; the modeling object includes n individual objects, where n≥2; the modeling data includes head modeling data, brain functional areas, intervention process information and intervention efficacy information corresponding to each individual object; Based on the head modeling data of the modeling object, the induced electric field of the cerebral cortex is calculated during the intervention process to obtain the cumulative induced electric field distribution of the cerebral cortex during the intervention process. Based on the cumulative distribution of induced electric fields in the cerebral cortex and the brain functional regions, an induced electric field functional connectivity matrix is ​​obtained; using the induced electric field functional connectivity matrix data combined with the intervention efficacy information, a target optimization model is established.

2. The method according to claim 1, characterized in that, Obtain the data to be modeled corresponding to the object being modeled, including: Obtain the MRI image corresponding to the modeling object, the MRI image including T1 image and rest-fMRI image; The T1 image is input into the feature extraction model to obtain the head model data of the modeling object; The rest-fMRI image is input into the feature extraction model to obtain the brain functional regions of the modeled object.

3. The method according to claim 1, characterized in that, The intervention process information includes the current pose of the TMS coil, the time point of TMS pulse generation during the TMS intervention process, and the TMS output intensity.

4. The method according to claim 1, characterized in that, The step of obtaining the induced electric field functional connectivity matrix based on the accumulated induced electric field distribution in the cerebral cortex and the brain functional regions includes: The cumulative induced electric field distribution of the cerebral cortex is binarized and segmented according to a preset threshold to obtain the target action area in the brain functional partition. Extract the rest-fMRI signal of the target area and calculate the average value to obtain the average signal of the target area. The induced electric field functional connection matrix is ​​constructed based on the average signal in the target area.

5. The method according to claim 4, characterized in that, The construction of the induced electric field functional connectivity matrix based on the average signal of the target area includes: Calculate the correlation coefficient between the target area and other brain functional regions; A functional connectivity matrix is ​​constructed based on the correlation coefficient, wherein each row and each column of the induced electric field functional connectivity matrix corresponds to a different brain functional region, and the element values ​​in the induced electric field functional connectivity matrix are used to characterize the functional connectivity strength between the corresponding brain functional regions.

6. The method according to claim 1, characterized in that, The step of establishing a target optimization model by combining the induced electric field functional connection matrix data with the intervention efficacy information includes: The induced electric field functional connection matrix is ​​divided according to preset rules to obtain multiple functional subnets; For each of the aforementioned functional subnets, calculate the functional connectivity strength between brain functional regions within the functional subnet; A target optimization model is established based on the functional connection strength of each functional subnet.

7. The method according to claim 6, characterized in that, The calculation of the functional connectivity strength between brain regions within the functional subnetwork includes: Calculate the Pearson correlation coefficients between brain functional regions within the aforementioned functional subnetworks; The functional connectivity strength between brain functional regions within the functional subnetwork is calculated based on the Pearson correlation coefficient.

8. The method according to any one of claims 1 to 7, characterized in that, The method further includes: Based on the target optimization model, the target coil pose and / or target output intensity during the intervention process are adjusted to optimize the intervention effect.

9. A target optimization device, characterized in that, include: The data acquisition module is used to obtain the data to be modeled corresponding to the modeling object; The modeling object includes n individual objects, where n ≥ 2; The data to be modeled includes head modeling data, brain functional areas, intervention process information, and intervention efficacy information for each individual. The calculation module is used to calculate the cortical induced electric field of the intervention process based on the head modeling data corresponding to the modeling object, to obtain the cumulative cortical induced electric field distribution of the intervention process, and to obtain the induced electric field functional connectivity matrix according to the cumulative cortical induced electric field distribution and the brain functional partition. The model building module is used to establish a target optimization model by combining the induced electric field function connection matrix data with the intervention efficacy information.

10. A computer-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method as described in any one of claims 1-8.

Images