Method and apparatus for rapid localization of hand motor hotspot

Abstract

A method and apparatus for rapid localization of a hand motor hotspot. The method comprises: acquiring magnetic resonance imaging data and MEP signal data of experimental subjects, and preprocessing the magnetic resonance imaging data, so as to obtain electric field simulation results; then constructing a group hand motor hotspot map, and classifying the experimental subjects into positive samples and negative samples for training, so as to obtain an outlier classification model; by means of the outlier classification model, classifying subjects to be subjected to localization into an outlier group and a non-outlier group; when a subject to be subjected to localization is classified into the outlier group, performing correlation analysis on the electric field simulation result and MEP signal data corresponding to said subject, so as to obtain a target hand motor hotspot; and when a subject to be subjected to localization is classified into the non-outlier group, performing individual hand motor hotspot registration by means of the group hand motor hotspot map, so as to obtain a target hand motor hotspot. The method overcomes the defect in the prior art of clinical treatment efficiency being low due to a hand motor hotspot localization process being time-consuming and labor-intensive and being unable to ensure the accuracy of localization.

Description

A method and device for rapid positioning of hand movement hotspots

[0001] Cross-reference of related applications

[0002] This application claims priority to Chinese Patent Application No. 202411527128.3, filed on October 30, 2024, entitled “A Method and Device for Rapid Positioning of Hand Movement Hotspots”, which is incorporated herein by reference in its entirety. Technical Field

[0003] This application relates to the field of transcranial magnetic stimulation medical technology, and in particular to a method and device for rapid localization of hand movement hotspots. Background Technology

[0004] Transcranial magnetic stimulation (TMS) is a non-invasive neuromodulation technique used to study human neurophysiology and treat neurological diseases. The hand motor hotspot (hMHS) refers to the area of ​​the cerebral cortex where motor evoked potentials (MEPs) are most easily elicited during TMS. It is used to determine the resting motor threshold (MT) of an individual. The intensity of stimulation that evokes MEPs is often used as a reference value for determining individual-specific stimulation intensity. Therefore, identifying the hand motor hotspot is a common procedure in TMS therapy.

[0005] Current clinical methods for locating hand motor hotspots require multiple stimulations at various locations around the patient's motor cortex to collect motor epithelial signal (MEP) data. The collected MEP data is then fitted to a surface to pinpoint the location of the maximum MEP signal as the hand motor hotspot. However, this entire process is time-consuming, labor-intensive, and cannot guarantee accuracy, resulting in low treatment efficiency. Summary of the Invention

[0006] This application provides a method and device for rapid positioning of hand motion hotspots, which solves the technical problem that the existing hand motion hotspot positioning process is time-consuming and laborious, and cannot guarantee the accuracy of positioning, resulting in low clinical treatment efficiency.

[0007] This application provides a method for rapid location of hand movement hotspots, including the following steps:

[0008] Magnetic resonance imaging data and MEP signal data corresponding to different locations of stimulation of the cerebral cortex of the experimental subjects were acquired, and the magnetic resonance imaging data were preprocessed to obtain electric field simulation results.

[0009] constructing a group hand movement hotspot atlas according to the MEP signal data and the electric field simulation results, and dividing the experimental subjects into positive samples and negative samples through the group hand movement hotspot atlas;

[0010] training an outlier classification model according to the positive samples and the negative samples;

[0011] dividing the to-be-positioned subject into an outlier group and a non-outlier group through the outlier classification model;

[0012] when the to-be-positioned subject is in the outlier group, performing correlation analysis on the electric field simulation results and the MEP signal data corresponding to the to-be-positioned subject to obtain a target hand movement hotspot;

[0013] when the to-be-positioned subject is in the non-outlier group, performing individual hand movement hotspot registration through the group hand movement hotspot atlas to obtain a target hand movement hotspot.

[0014] In some embodiments, the constructing a group hand movement hotspot atlas according to the MEP signal data and the electric field simulation results comprises:

[0015] performing correlation analysis on the electric field simulation results and the MEP signal data to obtain individual hand movement hotspots of the experimental subjects;

[0016] constructing a corresponding sample probability map based on the individual hand movement hotspots by using a Gaussian window function;

[0017] accumulating each sample probability map to construct a group hand movement hotspot atlas.

[0018] In some embodiments, the performing correlation analysis on the electric field simulation results and the MEP signal data to obtain individual hand movement hotspots of the experimental subjects comprises:

[0019] determining vertices corresponding to the brain cortex stimulation positions of the experimental subjects, and determining electric field values of each vertex at different stimulation positions from the electric field simulation results;

[0020] for each vertex, determining a positive correlation value of the electric field value and the MEP signal data;

[0021] determining an individual hand movement hotspot probability map according to the positive correlation value, and determining a vertex with the highest probability in the individual hand movement hotspot probability map as an individual hand movement hotspot of the experimental subject.

[0022] In some embodiments, the dividing the experimental subjects into positive samples and negative samples through the group hand movement hotspot atlas comprises:

[0023] From the hotspot map of group hand movements, the vertex with the highest probability is identified as the hotspot of group hand movements;

[0024] Determine the cortical distance between the individual hand motion hotspot of each experimental subject and the group hand motion hotspot;

[0025] When the cortical distance is greater than the distance threshold, the experimental subject is determined to be a negative sample;

[0026] When the cortical distance is less than or equal to the distance threshold, the experimental subject is determined to be a positive sample.

[0027] In some embodiments, training an outlier classification model based on the positive samples and the negative samples includes:

[0028] From the hotspot map of group hand movements, the vertex with the highest probability is identified as the hotspot of group hand movements;

[0029] Centered on the hot spots of hand movements in the group, regions of interest are identified from the cerebral cortex of the training samples, wherein the training samples include the positive samples and the negative samples;

[0030] From the magnetic resonance imaging data corresponding to the training samples, the cerebral cortex structural information in the region of interest is extracted, wherein the cerebral cortex structural information includes cortical thickness, cortical curvature and cortical sulcus index;

[0031] The information about the cerebral cortex structure was used as training data to train an initial logistic regression model, resulting in an outlier classification model.

[0032] In some embodiments, the step of registering individual hand motion hotspots using the group hand motion hotspot map to obtain target hand motion hotspots includes:

[0033] From the group hand movement hotspot map, the vertex with the highest probability is determined as the group hand movement hotspot;

[0034] Determine the registration mapping relationship from group to individual;

[0035] Based on the registration mapping relationship, the group hand motion hotspots are mapped to the target hand motion hotspots of the subjects to be located.

[0036] This application also provides a device for rapid positioning of hand movement hotspots, the device comprising the following modules:

[0037] The preprocessing module is used to acquire magnetic resonance imaging data of experimental subjects and MEP signal data corresponding to different locations of the motor cortex of the brain, and to preprocess the magnetic resonance imaging data to obtain electric field simulation results.

[0038] The module is used to construct a group hand motion heat map based on the MEP signal data and the electric field simulation results, and to divide the experimental subjects into positive samples and negative samples through the group hand motion heat map.

[0039] The training module is used to train an outlier classification model based on the positive samples and the negative samples;

[0040] The segmentation module is used to divide the subject to be located into outlier groups and non-outlier groups using the outlier classification model.

[0041] The analysis module is used to perform correlation analysis between the electric field simulation results and MEP signal data of the subject to be located when the subject to be located is an outlier, so as to obtain the target hand movement hotspot.

[0042] The registration module is used to register individual hand motion hotspots using the group hand motion hotspot map when the subject to be located is not an outlier, thereby obtaining the target hand motion hotspot.

[0043] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the hand motion hotspot rapid positioning method as described above.

[0044] This application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the hand motion hotspot rapid localization method as described above.

[0045] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the hand motion hotspot rapid positioning method as described above.

[0046] The method and apparatus for rapid localization of hand movement hotspots provided in this application first collect magnetic resonance imaging data and MEP signal data of experimental subjects to construct a group hand movement hotspot map. The experimental subjects are then divided into negative and positive samples to train an outlier classification model. This model is used to classify the outliers of the subjects to be located. When the subject is an outlier, the target hand movement hotspot is determined using MEP signal data; when the subject is not an outlier, registration is performed directly using the group hand movement hotspot map to determine the target hand movement hotspot. Therefore, by combining the outlier classification model and the group hand movement hotspot map, the hand movement hotspot of the subjects to be located is not only rapidly located but also effectively ensured to maintain accuracy, thus improving clinical treatment efficiency. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced one by one below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 is a flowchart illustrating the method for rapid location of hand motion hotspots provided in this application.

[0049] Figure 2 is a visualization of the cerebral cortex search grid provided in this application.

[0050] Figure 3 is a distribution map of MEP signals at 25 stimulation points provided in this application.

[0051] Figure 4 is a visualization of the simulated electric field results for the 25 stimulation points provided in this application.

[0052] Figure 5 is a visualization of the cortical structure information provided in this application.

[0053] Figure 6 shows the positive correlation between the electric field value provided in this application and the MEP signal data.

[0054] Figure 7 is a visualization of the probability map of hand movement hotspots provided in this application.

[0055] Figure 8 is a schematic diagram of the region of interest in the cerebral cortex provided in this application.

[0056] Figure 9 is a structural schematic diagram of the hand motion hotspot rapid positioning device provided in this application.

[0057] Figure 10 is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0059] The rapid hand motion hotspot localization method of this application is described below with reference to Figures 1-8. Figure 1 is a flowchart of the rapid hand motion hotspot localization method provided by this application. As shown in Figure 1, the method includes the following steps 101 to 106.

[0060] Step 101: Obtain magnetic resonance imaging data and MEP signal data corresponding to different locations of the cerebral cortex of the experimental subjects, and preprocess the magnetic resonance imaging data to obtain electric field simulation results.

[0061] This embodiment of the application takes into account the differences among each subject to be located. Some groups exhibit similar hand movement hotspots, which can be summarized and identified as a group of hand movement hotspots. However, the hand movement hotspots of another group are more unique, belonging to outliers. Therefore, when locating hand movement hotspots, this embodiment of the application needs to construct a group hand movement hotspot map as prior knowledge. This not only allows for sample segmentation to train an outlier classification model but also enables rapid hand movement hotspot location for the subjects to be located.

[0062] Therefore, before locating hand movement hotspots in subjects, it is necessary to acquire magnetic resonance imaging (MRI) data and MEP signal data corresponding to different locations of stimulation in the cerebral cortex (motor area). Here, during the MEP signal data acquisition process, as shown in Figure 2, a neuromodulation robot can be used to firmly fix the TMS coil, traversing a predefined search grid corresponding to the subject's cerebral cortex in a random order, and applying single-pulse stimulation with a certain intensity, recording the corresponding MEP signals. The search grid is a 5×5 square matrix, with a spacing of 1 cm between each stimulation point, for a total of 25 stimulation points, with the center of the matrix falling on the central stimulation point. During the hand movement hotspot search, the TMS coil traverses the 25 stimulation points in a random order, stimulating each point with single-pulse transcranial magnetic stimulation (spTMS) at 5-second intervals, stimulating each point 3 times, and acquiring the corresponding MEP signals. From the three collected MEP signals, the maximum value of the MEP signal was selected as the MEP signal data induced by TMS stimulation at that point. The distribution of MEP signals at the 25 stimulation points of the experimental subjects can be seen in Figure 3.

[0063] Magnetic resonance imaging (MRI) data can be extracted from experimental subjects using relevant MRI instruments, which will not be elaborated here. After data acquisition, the MRI data is preprocessed to obtain electric field simulation results. This involves a series of preprocessing steps, including cortical reconstruction, electric field simulation, and calculation of cortical structural information. For example, the subject's brain is first extracted, brain tissue is segmented, and the subject's individual spatial T1w image is registered onto the MNI standard template. The cerebral cortex is also segmented and reconstructed to accurately delineate the cerebral sulci and gyri. Then, electric field simulation and cortical structural information are extracted at the corresponding stimulation points on the cerebral cortex. In this embodiment, the visualized simulated electric field results for 25 stimulation points are shown in Figure 4, while the visualized results of the extracted cortical structural information are shown in Figure 5. The cortical structural information specifically includes cortical thickness, cortical curvature, and cortical sulci and gyri index.

[0064] Step 102: Based on the MEP signal data and electric field simulation results, construct a group hand motion heat map, and divide the experimental subjects into positive and negative samples using the group hand motion heat map.

[0065] After obtaining the MEP signal data and electric field simulation results of the experimental subjects, the next step is to construct a heat map of hand movements in the group based on the MEP signal data and electric field simulation results.

[0066] First, it is necessary to determine the individual hand motion hotspots for each experimental subject. In this embodiment, the electric field simulation results of the experimental subjects are correlated with the MEP signal data to obtain the individual hand motion hotspots. The specific process of the correlation analysis is described below.

[0067] First, the vertices corresponding to different locations in the cerebral cortex (motor area) of the experimental subjects were identified (e.g., the 25 stimulation points described above), and the electric field value of each vertex at different stimulation locations was determined from the electric field simulation results.

[0068] Then, for each vertex, the positive correlation between the electric field value and the MEP signal data is determined. As shown in Figure 6, the positive correlation between the electric field value (E Field) and the MEP signal data can be calculated based on this correlation. This positive correlation value can be mapped to the corresponding hand movement hotspot probability. The higher the positive correlation value, the stronger the positive correlation between the electric field value and the MEP signal, and thus the higher the hand movement hotspot probability. Collecting the hand movement hotspot probabilities for each vertex forms a hand movement hotspot probability map. As shown in Figure 7, this visualization shows the hand movement hotspot probability map for some vertices. In the hand movement hotspot probability map, the vertex corresponding to the highest probability is identified as the individual hand movement hotspot.

[0069] After determining the individual hand motion hotspots based on the electric field simulation results and MEP signal data, each experimental subject can identify their corresponding individual hand motion hotspots. Then, based on the individual hand motion hotspots, a Gaussian window function is used to construct the corresponding sample probability map. Here, the corresponding sample probability map g(x) is constructed based on the true value of the individual hand motion hotspot (i.e., the coordinates of the individual hand motion hotspot on the cerebral cortex (motor area)). For each individual hand motion hotspot's true value, a Gaussian window function is used to construct the corresponding sample probability model, expressed as the following formula (1):

[0070] In the above formula (1), μ represents the true value of any individual's hand movement hotspot, x represents any vertex on the cerebral cortex (motor area), and σ is a hyperparameter representing the effective stimulation range of TMS, which is generally determined based on the interval between 25 stimulation points. By performing the calculation process as in formula (1) on any vertex on the cerebral cortex (motor area), the probability corresponding to the vertex can be obtained. Thus, multiple different probabilities can be obtained, thereby forming a sample probability map.

[0071] Finally, the sample probability maps of each experimental subject are summed to construct a group hand movement hotspot map. Here, the sample probability maps corresponding to the hand movement hotspots of each individual are summed to obtain the group hand movement hotspot map p(x) corresponding to the population. The summation process is to add the Gaussian window functions corresponding to each hand movement hotspot, as shown in the following formula (2):

[0072] In the above formula (2), g i (x) represents the sample probability map corresponding to the individual hand movement hotspot of the i-th experimental subject. The meanings of the other parameters are the same as those in formula (1) and can be referred to each other. They will not be elaborated here.

[0073] Next, the experimental subjects were divided into positive and negative samples using a group hand movement hotspot map. The division process was as follows: First, the vertex with the highest probability was identified from the group hand movement hotspot map as the group hand movement hotspot. Then, the cortical distance between each experimental subject's individual hand movement hotspot and the group hand movement hotspot was determined. This distance was generally calculated by comparing the ground truth value of the individual hand movement hotspot (coordinates on the cerebral cortex (motor area)) with the ground truth value of the group hand movement hotspot.

[0074] When the cortical distance is greater than the distance threshold, it indicates that the individual hand movement hotspot of the experimental subject is far from the group's hand movement hotspot, belonging to the outgroup, and thus the out-of-group experimental subject is identified as a negative sample. When the cortical distance is less than or equal to the distance threshold, it indicates that the individual hand movement hotspot of the experimental subject is close to the group's hand movement hotspot, belonging to the non-out-of-group, and thus the non-out-of-group experimental subject is identified as a positive sample.

[0075] Step 103: Train the outlier classification model based on positive and negative samples.

[0076] After dividing the experimental subjects into positive and negative samples using the group hand movement heatmap in step 102, an outlier classification model can be trained based on the positive and negative samples.

[0077] When training an outlier classification model to distinguish between outliers and non-outliers, negative and positive samples are mixed to obtain training samples for the model. First, training data needs to be obtained from these samples. Here, the vertex with the highest probability is identified from the group hand movement heatmap as the group hand movement heatmap. Because the probability of each vertex in the group hand movement heatmap represents the likelihood of being identified as a group hand movement heatmap, the vertex with the highest probability can be used as the group hand movement heatmap to represent the hand movement heatmap of this population.

[0078] Next, using the group's hand movement hotspots as the center, the Region of Interest (ROI) was determined from the cerebral cortex (motor area) of the training samples (including negative and positive samples). Here, as shown in Figure 8, a 20mm region (the yellow area in Figure 8) can be defined in the cerebral cortex (motor area) centered on the group's hand movement hotspots as the ROI. Then, the cerebral cortex structural information within the ROI was extracted from the magnetic resonance imaging data of the training samples (i.e., the experimental subjects corresponding to the negative and positive samples). The cerebral cortex structural information includes cortical thickness, cortical curvature, and cortical sulcus index.

[0079] Finally, the structural information of the cerebral cortex (i.e., cortical thickness, cortical curvature, and cortical sulcus index) is used as training data to train the initial logistic regression model, resulting in an outlier classification model. During training, information features can be extracted from the structural information of the cerebral cortex, and the logistic regression model can learn the differences in information features between negative and positive samples. This enables the model to subsequently identify the information features of the subject being located, i.e., to identify negative or positive samples, thereby distinguishing whether the subject's hand movement hotspots are outliers. The logistic regression model can employ classification models from machine learning, such as random forests, decision trees, etc., which are not limited to the embodiments described in this application.

[0080] Step 104: For the subjects to be located, the subjects to be located are divided into outlier groups and non-outlier groups using an outlier classification model.

[0081] In some embodiments, after training, the outlier classification model can be used to distinguish between subjects to be located (i.e., new subjects). For subjects to be located, magnetic resonance imaging data can be acquired first, and then information about the cerebral cortex structure can be extracted from it. Then, information features can be extracted from it, and the outlier classification model can then identify the information features and predict whether the subject to be located is an outlier or not, thereby classifying the subject to be located into an outlier or not.

[0082] Step 105: When the subject to be located is an outlier, perform correlation analysis between the electric field simulation results corresponding to the subject to be located and the MEP signal data to obtain the target hand motion hotspot.

[0083] When the subject to be located was identified as an outlier, it indicated that the subject's hand movement hotspot was far from the group's hand movement hotspot, making it impossible to determine the target hand movement hotspot using the group's hand movement hotspot map. Therefore, correlation analysis was used to locate the target hand movement hotspot. First, magnetic resonance imaging (MRI) data of the subject to be located and MEP signal data corresponding to different locations of stimulation in the cerebral cortex were acquired. The MRI data was preprocessed to obtain electric field simulation results. Then, the vertices corresponding to the stimulation locations in the subject's cerebral cortex were determined, and the electric field values ​​of each vertex at different stimulation locations were determined from the electric field simulation results.

[0084] Next, for each vertex, the positive correlation value between the electric field value and the MEP signal data is determined. Based on the positive correlation value, an individual hand motion hotspot probability map is determined, and the vertex with the highest probability in the individual hand motion hotspot probability map is identified as the target hand motion hotspot for the subject to be located. This correlation analysis process is similar to step 102, and the specific implementation details will not be elaborated here.

[0085] Step 106: When the subject to be located is not an outlier, the individual hand motion hotspot is registered using the group hand motion hotspot map to obtain the target hand motion hotspot.

[0086] When the subject to be located is determined to be not an outlier, it means that the subject's hand movement hotspot is close to the group's hand movement hotspot. Therefore, the target hand movement hotspot can be determined by using the group's hand movement hotspot map. Thus, individual hand movement hotspot registration is performed using the group's hand movement hotspot map to obtain the target hand movement hotspot.

[0087] Specifically, the process involves identifying the vertices with the highest probability from the group hand motion hotspot map as the group's hand motion hotspots, and then determining the registration mapping relationship from the group to the individual. This can be done by first using the open-source registration tool Advanced Normalization Tools (ANTs) to register the T1w image corresponding to the subject's MRI data onto a standard MNI152 image, and then using the resulting mapping relationship as the registration mapping relationship. Finally, based on the registration mapping relationship, the group's hand motion hotspots are mapped onto the individual to obtain the target hand motion hotspots for the subject.

[0088] The method and apparatus for rapid localization of hand movement hotspots provided in this application first collect magnetic resonance imaging data and MEP signal data of experimental subjects to construct a group hand movement hotspot map. The experimental subjects are then divided into negative and positive samples to train an outlier classification model. This model is used to classify the outliers of the subjects to be located. When the subject is an outlier, the target hand movement hotspot is determined using MEP signal data; when the subject is not an outlier, registration is performed directly using the group hand movement hotspot map to determine the target hand movement hotspot. Therefore, by combining the outlier classification model and the group hand movement hotspot map, the hand movement hotspot of the subjects to be located is not only rapidly located but also effectively ensured to maintain accuracy, thus improving clinical treatment efficiency.

[0089] The rapid positioning device for hand motion hotspots provided in this application is described below. The rapid positioning device for hand motion hotspots described below can be referred to in correspondence with the rapid positioning method for hand motion hotspots described above.

[0090] Referring to Figure 9, which is a structural schematic diagram of the hand motion hotspot rapid localization device provided in this application, the device includes a preprocessing module 901, a construction module 902, a training module 903, a segmentation module 904, an analysis module 905, and a registration module 906. The preprocessing module 901 is used to acquire magnetic resonance imaging data of experimental subjects and MEP signal data corresponding to different locations of the stimulated motor cortex, and preprocesses the magnetic resonance imaging data to obtain electric field simulation results. The construction module 902 is used to construct a group hand motion hotspot map based on the MEP signal data and the electric field simulation results, and to segment the experimental subjects into positive and negative samples using the group hand motion hotspot map. The training module 903 is used to train outliers based on the positive and negative samples. The system includes a classification model; a segmentation module 904, used to classify the subject to be located into an outlier group and a non-outlier group using the outlier classification model; an analysis module 905, used to perform correlation analysis between the electric field simulation results and MEP signal data corresponding to the subject to be located when the subject to be located is in the outlier group, to obtain the target hand motion hotspot; and a registration module 906, used to register the individual hand motion hotspot using the group hand motion hotspot map when the subject to be located is in the non-outlier group, to obtain the target hand motion hotspot.

[0091] It should be noted that the beneficial effects of the hand motion hotspot rapid positioning device here correspond to those of the hand motion hotspot rapid positioning method mentioned above, so the beneficial effects of the hand motion hotspot rapid positioning device will not be repeated here.

[0092] Figure 10 is a schematic diagram of the physical structure of an electronic device provided in this application. As shown in Figure 10, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940. The processor 910, the communication interface 920, and the memory 930 communicate with each other through the communication bus 940. The processor 910 can call logic instructions in the memory 930 to execute a method for rapid localization of hand movement hotspots. This method includes: acquiring magnetic resonance imaging data and MEP signal data corresponding to different locations of the cerebral cortex from experimental subjects, and preprocessing the magnetic resonance imaging data to obtain electric field simulation results; constructing a group hand movement hotspot map based on the MEP signal data and the electric field simulation results, and classifying experimental subjects into positive and negative samples using the group hand movement hotspot map; training an outlier classification model based on the positive and negative samples; classifying the subject to be located into an outlier group and a non-outlier group using the outlier classification model; when the subject to be located is an outlier, performing correlation analysis between the electric field simulation results and the MEP signal data to obtain the target hand movement hotspot; when the subject to be located is a non-outlier, performing individual hand movement hotspot registration using the group hand movement hotspot map to obtain the target hand movement hotspot.

[0093] Furthermore, the logical instructions in the aforementioned memory 930 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0094] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the rapid hand motion hotspot localization method provided by the above methods. This method includes: acquiring magnetic resonance imaging data and MEP signal data corresponding to different locations of stimulation of the cerebral cortex of experimental subjects, and preprocessing the magnetic resonance imaging data to obtain electric field simulation results; constructing a group hand motion hotspot map based on the MEP signal data and the electric field simulation results, and then... The experimental subjects were divided into positive and negative samples using the group hand motion hotspot map. An outlier classification model was trained based on the positive and negative samples. For the subject to be located, the outlier classification model was used to divide the subject to be located into an outlier group and a non-outlier group. When the subject to be located was in the outlier group, the correlation analysis between the electric field simulation results and the MEP signal data corresponding to the subject to be located was performed to obtain the target hand motion hotspot. When the subject to be located was in the non-outlier group, the individual hand motion hotspot was registered using the group hand motion hotspot map to obtain the target hand motion hotspot.

[0095] Furthermore, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, this computer program implements a method for rapid localization of hand motion hotspots provided by the methods described above. This method includes: acquiring magnetic resonance imaging data and MEP signal data corresponding to different locations of stimulation in the cerebral cortex of experimental subjects; preprocessing the magnetic resonance imaging data to obtain electric field simulation results; constructing a group hand motion hotspot map based on the MEP signal data and the electric field simulation results; and using the group hand motion hotspot map to... Experimental subjects were divided into positive and negative samples; an outlier classification model was trained based on the positive and negative samples; for the subject to be located, the outlier classification model was used to divide the subject to be located into an outlier group and a non-outlier group; when the subject to be located was in the outlier group, the correlation analysis between the electric field simulation results corresponding to the subject to be located and the MEP signal data was performed to obtain the target hand motion hotspot; when the subject to be located was in the non-outlier group, the individual hand motion hotspot was registered using the group hand motion hotspot map to obtain the target hand motion hotspot.

[0096] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0097] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for rapid localization of hand movement hotspots, the method comprising: Magnetic resonance imaging data and MEP signal data corresponding to different locations of stimulation of the cerebral cortex of the experimental subjects were acquired, and the magnetic resonance imaging data were preprocessed to obtain electric field simulation results. Based on the MEP signal data and the electric field simulation results, a group hand motion heat map was constructed, and the experimental subjects were divided into positive and negative samples by the group hand motion heat map. An outlier classification model was trained based on the positive and negative samples. For the subjects to be located, the outlier classification model is used to divide the subjects to be located into outlier groups and non-outlier groups; When the subject to be located is an outlier, the electric field simulation results corresponding to the subject to be located are correlated with the MEP signal data to obtain the target hand movement hotspot; When the subject to be located is not an outlier, the individual hand motion hotspot is registered using the group hand motion hotspot map to obtain the target hand motion hotspot.

2. The method for rapid location of hand movement hotspots according to claim 1, wherein, The step of constructing a heat map of group hand movements based on the MEP signal data and the electric field simulation results includes: Correlation analysis was performed between the electric field simulation results and the MEP signal data to obtain the individual hand movement hotspots of the experimental subjects; Based on the individual hand movement hotspots, a Gaussian window function is used to construct the corresponding sample probability map; By summing the probability maps of each sample, a heat map of hand movements in the group is constructed.

3. The method for rapid location of hand movement hotspots according to claim 2, wherein, The correlation analysis between the electric field simulation results and the MEP signal data is performed to obtain the individual hand movement hotspots of the experimental subjects, including: The vertex corresponding to the stimulation location of the cerebral cortex of the experimental subjects was determined, and the electric field value of each vertex at different stimulation locations was determined from the electric field simulation results. For each vertex, determine the positive correlation between the electric field value and the MEP signal data; Based on the positive correlation value, an individual hand movement hotspot probability map is determined, and the vertex with the highest probability in the individual hand movement hotspot probability map is determined as the individual hand movement hotspot of the experimental subject.

4. The method for rapid location of hand movement hotspots according to claim 1, wherein, The experimental subjects were divided into positive and negative samples based on the group hand movement heatmap, including: From the hotspot map of group hand movements, the vertex with the highest probability is identified as the hotspot of group hand movements; Determine the cortical distance between the individual hand motion hotspot of each experimental subject and the group hand motion hotspot; When the cortical distance is greater than the distance threshold, the experimental subject is determined to be a negative sample; When the cortical distance is less than or equal to the distance threshold, the experimental subject is determined to be a positive sample.

5. The method for rapid location of hand movement hotspots according to claim 1, wherein, The process of training an outlier classification model based on the positive and negative samples includes: From the hotspot map of group hand movements, the vertex with the highest probability is identified as the hotspot of group hand movements; Centered on the hot spots of hand movements in the group, regions of interest are identified from the cerebral cortex of the training samples, wherein the training samples include the positive samples and the negative samples; From the magnetic resonance imaging data corresponding to the training samples, the cerebral cortex structural information in the region of interest is extracted, wherein the cerebral cortex structural information includes cortical thickness, cortical curvature and cortical sulcus index; The information about the cerebral cortex structure was used as training data to train an initial logistic regression model, resulting in an outlier classification model.

6. The method for rapid location of hand movement hotspots according to claim 1, wherein, The step of registering individual hand motion hotspots using the group hand motion hotspot map to obtain target hand motion hotspots includes: From the group hand movement hotspot map, the vertex with the highest probability is determined as the group hand movement hotspot; Determine the registration mapping relationship from group to individual; Based on the registration mapping relationship, the group hand motion hotspots are mapped to the target hand motion hotspots of the subjects to be located.

7. A device for rapid positioning of hand movement hotspots, the device comprising: The preprocessing module is used to acquire magnetic resonance imaging data of experimental subjects and MEP signal data corresponding to different locations of the motor cortex of the brain, and to preprocess the magnetic resonance imaging data to obtain electric field simulation results. The module is used to construct a group hand motion heat map based on the MEP signal data and the electric field simulation results, and to divide the experimental subjects into positive samples and negative samples through the group hand motion heat map. The training module is used to train an outlier classification model based on the positive samples and the negative samples; The segmentation module is used to divide the subject to be located into outlier groups and non-outlier groups using the outlier classification model. The analysis module is used to perform correlation analysis between the electric field simulation results and MEP signal data of the subject to be located when the subject to be located is an outlier, so as to obtain the target hand movement hotspot. The registration module is used to register individual hand motion hotspots using the group hand motion hotspot map when the subject to be located is not an outlier, in order to obtain the target hand motion hotspot.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein, When the processor executes the computer program, it implements the method for rapid location of hand movement hotspots as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein, When the computer program is executed by the processor, it implements the method for rapid location of hand motion hotspots as described in any one of claims 1 to 6.

10. A computer program product comprising a computer program, wherein, When the computer program is executed by the processor, it implements the method for rapid location of hand motion hotspots as described in any one of claims 1 to 6.

Images