MRI-based methods and devices for predicting the efficacy of treatment for depression

Abstract

This application provides a method and related apparatus for predicting the treatment efficacy of depression based on MRI, relating to the medical field. In this method, an electronic device acquires initial depression data and treatment efficacy data from multiple test subjects, where the initial depression data for each test subject represents pre-treatment depressive characteristics. Then, a model to be trained is generated based on the initial depression data and treatment efficacy data from the multiple test subjects to obtain a treatment efficacy prediction model that meets preset conditions. The initial depression data from the multiple test subjects serves as input data for training the model, and the treatment efficacy data from the multiple test subjects serves as supervision data for training the model. The treatment efficacy prediction model is used to predict the treatment efficacy of depression; thus, the treatment efficacy prediction model trained in the above manner can predict treatment effects relatively objectively.

Description

Technical Field

[0001] This application relates to the medical field, and more specifically, to an MRI-based method and device for predicting the efficacy of treatment for depression. Background Technology

[0002] Subthreshold depression (StD), also known as subclinical depression, mild depression, or subsyndromic depression, refers to clinically significant depressive symptoms that do not meet the diagnostic criteria for major depressive disorder. Subthreshold depression has become a growing mental health problem. Increasing evidence suggests that subthreshold depression predicts major depressive disorder in adulthood, with a 67% risk of developing into a full syndrome. Furthermore, adolescent subthreshold depression carries a risk of developing other illnesses and is associated with functional impairment, low quality of life, and high mortality, increasing economic costs. Therefore, early intervention for subthreshold depression in adolescence and early adulthood is crucial.

[0003] Compared to antidepressants and psychotherapy, light therapy is a recognized, effective, safe, non-pharmacological, and low-cost intervention for both seasonal affective disorder (SAD) and non-seasonal depression. Numerous studies have shown that light therapy can effectively alleviate depressive symptoms in patients with seasonal affective disorder, non-seasonal depression, bipolar disorder, geriatric depression, and adolescent depression. Furthermore, some studies have demonstrated that light therapy has a certain interventional effect on subthreshold depression in college students.

[0004] However, in practice, it has been found that the current assessment of the effectiveness of phototherapy in improving depression relies heavily on doctors' professional knowledge and subjective judgment, and lacks objective neuroimaging indicators. Summary of the Invention

[0005] To overcome at least one deficiency in the prior art, this application provides an MRI-based method and related device for predicting the efficacy of treatment for depression, specifically including:

[0006] In a first aspect, this application provides an MRI-based method for predicting the efficacy of treatment for depression, the method comprising:

[0007] Acquire initial depression data and efficacy test data from multiple test subjects, wherein the initial depression data for each test subject represents pre-treatment depression characteristics data;

[0008] The model to be trained is trained based on the first depression data of the multiple test subjects and the measured efficacy data to obtain an efficacy prediction model that meets preset conditions. The first depression data of the multiple test subjects is the input data when the model to be trained is trained, and the measured efficacy data of the multiple test subjects is the supervision data when the model to be trained is trained. The efficacy prediction model is used to predict the efficacy of depression treatment.

[0009] In conjunction with the optional implementation of the first aspect, the model to be trained includes multiple parameters to be fitted. The model is trained based on the first depression data of the multiple test subjects and the measured efficacy data to obtain an efficacy prediction model that meets preset conditions, including:

[0010] The first depression data and second depression data of the multiple test subjects are divided into multiple subsets, wherein each subset includes the first depression data and efficacy test data of a portion of the test subjects;

[0011] The model to be trained is trained according to each of the said subsets of data to obtain a set of initial fitting results for the multiple parameters to be fitted for each of the said subsets of data.

[0012] Based on the multiple sets of initial fitting results, the target fitting results of the multiple parameters to be fitted are obtained;

[0013] The therapeutic effect prediction model is obtained based on the target fitting results of the multiple parameters to be fitted.

[0014] In conjunction with the optional implementation of the first aspect, based on multiple sets of initial fitting results, a target fitting result for the multiple parameters to be fitted is obtained, including:

[0015] Based on the multiple sets of initial fitting results, the average fitting result of each parameter to be fitted is obtained;

[0016] The average fitting result of each of the parameters to be fitted is taken as the target fitting result of each parameter to be fitted.

[0017] In conjunction with the optional implementation method of the first aspect, primary depression data and efficacy test data of multiple test subjects are obtained, including:

[0018] First and second images of the brains of the multiple test subjects are acquired, wherein the first image of the brain of each test subject is a medical image of the test subject before treatment, and the second image of the brain of each test subject is a medical image of the test subject after treatment.

[0019] Based on the first and second images of the brains of the multiple test subjects, the first depression data and the actual therapeutic effect data of the multiple test subjects were obtained.

[0020] In conjunction with the optional implementation of the first aspect, based on the first and second images of the brains of the plurality of test subjects, the first depression data and efficacy measurement data of the plurality of test subjects are obtained, including:

[0021] Based on the first and second images of the brains of the multiple test subjects, the characteristic brain regions that changed before and after brain treatment were identified.

[0022] The characteristic brain regions in the first image of each test subject's brain were analyzed to obtain the first depression data for each test subject;

[0023] The characteristic brain regions in the second image of each test subject's brain are analyzed to obtain the second depression data of each test subject, wherein the second depression data of each test subject represents the depression characteristic data after treatment;

[0024] Based on the first depression data and the second depression data of each test subject, the measured efficacy data of each test subject were obtained.

[0025] In conjunction with an optional implementation of the first aspect, based on first and second images of the brains of the plurality of test subjects, characteristic brain regions that changed before and after brain treatment are identified, including:

[0026] Based on the first image of each test subject's brain, brain biometric indicators were obtained for each test subject before treatment.

[0027] Based on the second image of each test subject's brain, the brain biometric indicators of each test subject after treatment were obtained;

[0028] The brain biometrics of each test subject before and after treatment were compared to identify the characteristic brain regions.

[0029] In conjunction with the optional implementation of the first aspect, characteristic brain regions in the first image of each test subject's brain are analyzed to obtain the test subject's first depression data, including:

[0030] For each test subject, based on the test subject's first image, the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala in the characteristic brain regions of the test subject before treatment are obtained.

[0031] The pre-treatment static functional connectivity index and dynamic functional connectivity index of the amygdala were used as the first depression data of the test subjects.

[0032] The characteristic brain regions in the second brain images of each test subject were analyzed to obtain the second depression data for each test subject, including:

[0033] For each test subject, based on the second image of the test subject, the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala after treatment of the characteristic brain regions of the test subject are obtained.

[0034] The post-treatment static functional connectivity index and dynamic functional connectivity index of the amygdala were used as the second depression data for the test subjects.

[0035] Secondly, this application also provides an MRI-based device for predicting the efficacy of treatment for depression, the device comprising:

[0036] The data processing module acquires the initial depression data and efficacy test data of multiple test subjects, wherein the initial depression data of each test subject represents the depression characteristic data before treatment;

[0037] The model training module trains the model to be trained based on the first depression data of the multiple test subjects and the measured efficacy data to obtain an efficacy prediction model that meets preset conditions. The first depression data of the multiple test subjects is the input data when the model to be trained is trained, and the measured efficacy data of the multiple test subjects is the supervision data when the model to be trained is trained. The efficacy prediction model is used to predict the efficacy of depression treatment.

[0038] In conjunction with the optional implementation of the second aspect, the model to be trained includes multiple parameters to be fitted, and the model training module is further specifically used for:

[0039] The first depression data and second depression data of the multiple test subjects are divided into multiple subsets, wherein each subset includes the first depression data and efficacy test data of a portion of the test subjects;

[0040] The model to be trained is trained according to each of the said subsets of data to obtain a set of initial fitting results for the multiple parameters to be fitted for each of the said subsets of data.

[0041] Based on the multiple sets of initial fitting results, the target fitting results of the multiple parameters to be fitted are obtained;

[0042] The therapeutic effect prediction model is obtained based on the target fitting results of the multiple parameters to be fitted.

[0043] In conjunction with the optional implementation of the second aspect, the model training module is further specifically used for:

[0044] Based on the multiple sets of initial fitting results, the average fitting result of each parameter to be fitted is obtained;

[0045] The average fitting result of each of the parameters to be fitted is taken as the target fitting result of each parameter to be fitted.

[0046] In conjunction with the optional implementation of the second aspect, the data processing module is also specifically used for:

[0047] First and second images of the brains of the multiple test subjects are acquired, wherein the first image of the brain of each test subject is a medical image of the test subject before treatment, and the second image of the brain of each test subject is a medical image of the test subject after treatment.

[0048] Based on the first and second images of the brains of the multiple test subjects, the first depression data and the actual therapeutic effect data of the multiple test subjects were obtained.

[0049] In conjunction with the optional implementation of the second aspect, the data processing module is also specifically used for:

[0050] Based on the first and second images of the brains of the multiple test subjects, the characteristic brain regions that changed before and after brain treatment were identified.

[0051] The characteristic brain regions in the first image of each test subject's brain were analyzed to obtain the first depression data for each test subject;

[0052] The characteristic brain regions in the second image of each test subject's brain are analyzed to obtain the second depression data of each test subject, wherein the second depression data of each test subject represents the depression characteristic data after treatment;

[0053] Based on the first depression data and the second depression data of each test subject, the measured efficacy data of each test subject were obtained.

[0054] In conjunction with the optional implementation of the second aspect, the data processing module is also specifically used for:

[0055] Based on the first image of each test subject's brain, brain biometric indicators were obtained for each test subject before treatment.

[0056] Based on the second image of each test subject's brain, the brain biometric indicators of each test subject after treatment were obtained;

[0057] The brain biometrics of each test subject before and after treatment were compared to identify the characteristic brain regions.

[0058] In conjunction with the optional implementation of the second aspect, the data processing module is also specifically used for:

[0059] For each test subject, based on the test subject's first image, the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala in the characteristic brain regions of the test subject before treatment are obtained.

[0060] The pre-treatment static functional connectivity index and dynamic functional connectivity index of the amygdala were used as the first depression data of the test subjects.

[0061] The characteristic brain regions in the second brain images of each test subject were analyzed to obtain the second depression data for each test subject, including:

[0062] For each test subject, based on the second image of the test subject, the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala after treatment of the characteristic brain regions of the test subject are obtained.

[0063] The post-treatment static functional connectivity index and dynamic functional connectivity index of the amygdala were used as the second depression data for the test subjects.

[0064] Thirdly, this application also provides a storage medium storing a computer program, which, when executed by a processor, implements the MRI-based method for predicting the efficacy of depressive therapy.

[0065] Fourthly, this application also provides an electronic device, which includes a processor and a memory, the memory storing a computer program, which, when executed by the processor, implements the MRI-based method for predicting the efficacy of depression treatment.

[0066] Compared with the prior art, this application has the following beneficial effects:

[0067] This application provides a method and related apparatus for predicting the treatment efficacy of depression based on MRI. In this method, an electronic device acquires initial depression data and treatment efficacy data from multiple test subjects, where the initial depression data for each test subject represents pre-treatment depressive characteristics. Then, a model to be trained is generated based on the initial depression data and treatment efficacy data from the multiple test subjects to obtain a treatment efficacy prediction model that meets preset conditions. The initial depression data from the multiple test subjects serves as input data for training the model, and the treatment efficacy data from the multiple test subjects serves as supervision data for training the model. The treatment efficacy prediction model is used to predict the treatment efficacy of depression. Thus, the treatment efficacy prediction model trained in the above manner can predict the treatment effect relatively objectively. Attached Figure Description

[0068] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 This is one of the procedures for the MRI-based method for predicting the efficacy of depressive therapy provided in the embodiments of this application;

[0070] Figure 2 The second flowchart of the MRI-based method for predicting the efficacy of treatment for depression provided in the embodiments of this application;

[0071] Figure 3 The third flowchart of the MRI-based method for predicting the efficacy of treatment for depression provided in the embodiments of this application;

[0072] Figure 4 A schematic diagram of the structure of the MRI-based depression treatment prediction device provided in the embodiments of this application;

[0073] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0074] Icons: 11-Data processing module; 12-Model training module; 21-Memory; 22-Processor; 23-Communication unit; 24-System bus. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments 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, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0076] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0077] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0078] In the description of this application, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. In addition, the terms "first," "second," and "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0079] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0080] Furthermore, terms such as "horizontal," "vertical," and "sag" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0081] In the description of this application, it should also be noted that, unless otherwise expressly specified and limited, the terms "set up," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0082] Based on the above statement, as described in the background section, current assessments of the effectiveness of phototherapy in improving depression heavily rely on physicians' professional knowledge and subjective judgment, lacking objective neuroimaging indicators. Specifically, the effectiveness of phototherapy in improving depression is mainly assessed through clinical scales, such as the Hamilton Depression Rating Scale score. Although these clinical measures have some efficacy, they still lack sufficient specificity to reduce the uncertainty of prognosis, heavily relying on physicians' professional knowledge and subjective judgment, and lacking objective neuroimaging indicators.

[0083] Based on the discovery of the aforementioned technical problems, the inventors, through creative labor, proposed the following technical solutions to solve or improve these problems. It should be noted that the deficiencies in the solutions of the prior art are the result of the inventors' practical experience and careful research. Therefore, the discovery process of the aforementioned problems and the solutions proposed in the embodiments of this application below should be considered contributions made by the inventors to this application during the inventive process, and should not be construed as technical content known to those skilled in the art.

[0084] In view of the above problems, this embodiment provides a method for predicting the treatment efficacy of depression based on MRI. In this method, an electronic device acquires initial depression data and treatment efficacy data from multiple test subjects, where the initial depression data for each test subject represents pre-treatment depressive characteristics. Then, a model to be trained is generated based on the initial depression data and treatment efficacy data from multiple test subjects to obtain a treatment efficacy prediction model that meets preset conditions. The initial depression data from multiple test subjects serves as input data for training the model, and the treatment efficacy data from multiple test subjects serves as supervision data for training the model. The treatment efficacy prediction model is used to predict the treatment efficacy of depression. Thus, the treatment efficacy prediction model trained in the above manner can predict the treatment effect relatively objectively.

[0085] It is worth noting that, for the solutions involved in this application, all data related to user privacy was explicitly disclosed and authorized by the user before being collected or generated during the execution of the embodiments of this application. Furthermore, the storage device location for the personal information involved in the embodiments of this application complies with the legal requirements of the country / region where the aforementioned data-related activities occurred, including but not limited to: authorization, generation, use, and storage. It can be understood that the storage device location for the personal information used in implementing this solution complies with the legal requirements of the country / region where the aforementioned data-related activities occurred, including but not limited to: authorization, generation, use, and storage.

[0086] Furthermore, the electronic device implementing the above method can be, but is not limited to, a mobile terminal, a computer (e.g., a tablet computer, a laptop computer, a desktop computer), a server, etc. The server can be a single server or a group of servers. The server group can be centralized or distributed (e.g., the servers can be a distributed system). In some embodiments, the server can be local or remote relative to the user terminal. In some embodiments, the server can be implemented on a cloud platform; by way of example only, the cloud platform can include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, inter-cloud, multi-cloud, etc., or any combination thereof. In some embodiments, the server can be implemented on an electronic device having one or more components.

[0087] To make the solution provided in this embodiment clearer, it is assumed below that the electronic device is a computer and combined with... Figure 1 Each step of the method is described in detail. However, it should be understood that the operations in the flowchart may not be implemented in sequence, and steps without logical contextual relationships may be reversed in order or implemented simultaneously. Furthermore, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowchart, or remove one or more operations from the flowchart. Figure 1 As shown, the method includes:

[0088] S1: Obtain initial depression data and efficacy test data from multiple test subjects.

[0089] In this embodiment, the initial depression data for each participant represents their pre-treatment depressive characteristics. Multiple participants were treated with light therapy, and pre- and post-treatment depression data were obtained. The specific experimental procedure is as follows:

[0090] For example, multiple participants were randomly assigned to a first group and a second group. The first group received an 8-week intervention with 5000 lux of light, while the second group received an 8-week intervention with <5 lux of light. The first group was instructed to expose themselves to a custom-designed lightbox (5000 Lux) for 30 minutes each morning (before noon). The lightbox contained small white LEDs with a color temperature of 5000 Kelvin and a 100% UV filter. The device was placed on a table 50 cm away, and participants were allowed to fully expose their faces to the light without looking directly at it. They were permitted to read but not use any electronic devices. The second group received the same intervention device but was set to emit a very low light (<5 Lux). Both switches for the first and second groups were on, indicating that the lights were on and powered. Both groups used the lightbox at home with standardized verbal and written instructions.

[0091] The aforementioned first and second depression data were obtained by analyzing brain medical images of the test subjects before and after treatment with the aforementioned treatment method. These brain medical images can be MRI (Magnetic Resonance Imaging).

[0092] Understandably, computer-generated randomization codes were used at recruitment sites to conceal the assignments. Both groups of participants underwent assessments for depressive symptoms and MRI scans before and 8 weeks after the light intervention. Researchers responsible for head scans, depressive symptom assessments, and data analysis were unaware of the randomization assignments. Non-blinded investigators assigned participants, distributed lightboxes, and provided instructions for their use. Therefore, as... Figure 2 As shown, Figure 1 An optional implementation of step S1 shown may include:

[0093] S1-1, acquire first and second images of the brains of multiple test subjects.

[0094] The first image of each test subject's brain is a medical image taken before treatment, and the second image is a medical image taken after treatment. For example, the first and second images of the test subject's brain can be obtained using MRI equipment, and may include structural magnetic resonance imaging (MRI) and resting-state functional magnetic resonance imaging (fMRI).

[0095] Structural magnetic resonance imaging (MRI) is primarily used to acquire information about the brain's anatomical structure. It generates detailed images of the brain by measuring the density of different tissues within it, such as gray matter, white matter, and cerebrospinal fluid. These images can be used to study the brain's structural features, such as the shape, size, and location of brain regions, as well as the connections between different areas within the brain.

[0096] Resting-state functional magnetic resonance imaging (fMRI) focuses on the functional activity of the brain, especially spontaneous activity in the absence of a specific task or stimulus. Even at rest, different areas of the brain still communicate with each other, forming what is known as "resting-state networks." By measuring changes in blood oxygen level-dependent (BOLD) signals, the activity patterns and functional connectivity of these networks can be revealed. Therefore, resting-state fMRI brain imaging is independent of specific external tasks and can be used to study the basic functional organization of the brain, as well as changes in these functional networks in disease states.

[0097] In practice, it was found that when using the original images measured before and after treatment as the first and second images, interference information may exist in the original images due to equipment limitations or the tester's own limitations. Therefore, the aforementioned first and second images can be preprocessed versions of the original images measured before and after treatment. To address this, this embodiment employs methods such as instability correction, temporal layer correction, head motion correction, spatial normalization, delinear drift removal, and low-frequency filtering to preprocess the original images. Each of these correction methods is described in detail below:

[0098] (1) Instability correction: During the brain scan process, the scanning instrument may affect the collected brain image data due to magnetic field instability in the early stage of the scan. Therefore, in order to ensure the stability of longitudinal magnetization of the scanning instrument, the data of the first N time points in the time series signal can be removed to avoid the influence of magnetic field instability on the data. Preferably, N can be 10. Users can also set the value of N according to the actual situation, which is not limited here.

[0099] (2) Time-layer correction: During the brain scan of the test subject, the brain is divided into N layers at each scanning time point. The scanning process is to scan layer 1, layer 3... layer N, layer 2, layer 4... layer N-1 sequentially, that is, scan the odd-numbered layers first and then scan the even-numbered layers. However, theoretically, the blood sample levels between adjacent layers are relatively close. In order to ensure the consistency of blood sample level signals between adjacent layers, for the N layers of brain imaging data scanned at each time point, the scanning time of the middle layer (layer N) is used as the reference, and the scanning time of the corresponding brain imaging data of the other layers is shifted so that each layer is aligned to the same time origin, so as to avoid the large fluctuation of the obtained blood oxygen level signal affecting the accuracy of the data. For example, layer 2 is aligned to the space between layers 1 and 3, and layer 4 is aligned to the space between layers 3 and 5. Of course, this step can be omitted for sequential scanning MRI equipment.

[0100] (3) Head movement correction: During brain scans, body movement and brain swaying may cause deviations between the scanned brain images, affecting subsequent data processing. Therefore, alignment calibration can obtain head movement records for each subject during resting-state functional magnetic resonance imaging (fMRI) scans. These head movement records can be used as screening criteria to remove brain images with significant deviations, ensuring the accuracy of data processing. In this embodiment, the brain image screening criteria are: the maximum displacement of the subject on any plane does not exceed 2 mm, the rotation angle does not exceed 2 degrees, and the frame shift does not exceed 0.2 mm. Images that do not meet these criteria will be removed. Of course, users can set the above screening criteria according to actual application scenarios, which is not limited here.

[0101] (4) Spatial standardization: Since the gray matter region of the brain is closely related to the behavior of the test subject, the DARTEL segmentation method can be used to segment the gray matter, white matter (gray matter is responsible for processing information, while white matter is responsible for transmitting information), and cerebrospinal fluid in the structural images. Then, the segmented structural images are registered to the functional images through a 6-DOF linear transformation, so that the gray matter regions in the structural images match those in the functional images, avoiding data errors. To obtain a standard image processing template, after completing the registration between the structural and functional images, the brain image data can be processed to make the voxel size 3×3×3mm, and then the functional images can be registered to a specific template. By registering the registered functional images to a specific template, such as a specific template in the standard Montreal Neurological Institute (MNI) space, the problem of differences in brain structure between different subjects can be overcome.

[0102] It should be understood that in neuroimaging research, researchers use standardized brain templates to compare brain structures across different individuals. Specific templates within the Montreal Neurological Institute (MNI) space serve as such a standardized reference framework, helping researchers map brain data from different individuals into a common space, thus overcoming the challenges of individual differences in brain structure. The MNI template, also known as the MNI standard brain or MNI152 template, is an average brain template derived through statistical analysis of brain scans from a large number of normal subjects. This template represents the structure of an "average" brain, including tissues such as gray matter, white matter, and cerebrospinal fluid. Researchers can use this template as a reference to register individual brain images into the MNI space, allowing for comparison and analysis of brain data from different individuals within a common, standardized framework.

[0103] (5) Delinear drift and low-frequency filtering: In order to adapt to the temperature rise and other changes that occur during the operation of the scanning instrument, all voxels in the functional image data can be delinear drifted to remove the data errors caused by the time alignment correction and spatial standardization. At the same time, the functional image signal after low-frequency filtering may reflect spontaneous neural activity. Generally, the low-frequency filtering frequency band is 0.01-0.1Hz. Signal filtering in this frequency band can avoid the influence of physiological noise such as heartbeat and breathing.

[0104] For the functional imaging data that has undergone de-line drift and low-frequency filtering, the influence of whole-brain time series signals, brain gray matter signals, brain white matter signals, cerebrospinal fluid signals, and head motion parameter models on the time series signals corresponding to the functional imaging data can be reduced by using a multivariate linear model.

[0105] Thus, after preprocessing the original images of the test subjects before and after treatment using the aforementioned methods such as instability correction, temporal layer correction, head motion correction, spatial standardization, delinear drift removal, and low-frequency filtering, the first and second images of the test subjects are obtained.

[0106] Based on the first and second images obtained from the above embodiments, see below. Figure 2 , Figure 1 Step S1 shown also includes:

[0107] S1-2: Based on the first and second images of the brains of multiple test subjects, the first depression data and efficacy test data of multiple test subjects were obtained.

[0108] Studies have found that after light therapy, changes occur in the brain's biometric indicators, with the most significant changes observed in the lesion areas leading to depressive symptoms. This embodiment refers to this area as the characteristic brain region. Therefore, steps S1-2 above also include:

[0109] S1-2-1: Based on the first and second images of the brains of multiple test subjects, identify the characteristic brain regions that changed before and after brain treatment.

[0110] Optionally, the computer can obtain the brain biometrics of each test subject before treatment based on the first image of each test subject's brain; obtain the brain biometrics of each test subject after treatment based on the second image of each test subject's brain; and compare the brain biometrics of each test subject before treatment with the brain biometrics of each test subject after treatment to determine the characteristic brain regions.

[0111] In this embodiment, the brain biometrics before and after treatment can be both static and dynamic functional connectivity indicators of the amygdala before and after treatment of the entire brain. It should be understood that the amygdala is an important structure in the brain, located within the limbic system, and is a key area for emotional and memory processing, particularly related to fear and pleasure. Amygdala dysfunction is considered to be associated with various neuropsychiatric disorders, such as anxiety, depression, and post-traumatic stress disorder.

[0112] Static functional connectivity (sFC) of the amygdala refers to the level of synchronized activity between the amygdala and other brain regions at rest. This connectivity is measured using functional magnetic resonance imaging (fMRI) while individuals are not performing any specific task. sFC can reflect how the amygdala interacts with other parts of the brain at rest, which may be related to its role in emotion regulation and memory.

[0113] Dynamic functional connectivity (dFC) of the amygdala refers to the changes in connectivity between the amygdala and other brain regions over time when performing a specific task or in different states. Unlike static functional connectivity, dFC considers changes in connectivity over time, which can provide more information about how the amygdala dynamically adjusts its functional connectivity in different contexts.

[0114] These two data indicators can characterize the functional properties of the amygdala in different states, and can be used to help understand the role of the amygdala in emotional and cognitive processes, as well as the relationship between its functional abnormalities and neuropsychiatric disorders. Furthermore, by studying the static and dynamic functional connectivity of the nucleoamygdala, we can gain a deeper understanding of how the brain processes emotions and memories, and may provide new targets for the diagnosis and treatment of these diseases. Therefore, in this embodiment, these two data points are used as depressive characteristic data reflecting depressive symptoms. For ease of description, the static and dynamic functional connectivity of the nucleoamygdala will be represented by abbreviations (sFC, dFC) in subsequent embodiments.

[0115] To make the concepts of static functional connectivity index (sFC) and dynamic functional connectivity index (dFC) of the amygdala easier to understand, the following examples illustrate the specific methods for obtaining the amygdala's sFC and dFC values:

[0116] The first and second images of the aforementioned test subjects are T1-weighted images. The computer obtains the amygdala subregion of each test subject's cerebral hemisphere from the T1-weighted images. In a specific implementation, the computer uses FreeSurfer software to process the T1-weighted images of each test subject. The amygdala segmentation module is used to quantify the volume of the nine amygdala subregions (lateral nucleus, basal nucleus, central nucleus, medial nucleus, cortical nucleus, parabasal nucleus, paralaminar nucleus, cortical amyloid transition zone, and anterior amygdala) and the entire amygdala in each test subject's cerebral hemisphere. Those skilled in the art can use some or all of the above nine amygdala subregions when implementing this scheme. This embodiment selects four amygdala subregions—the basolateral amygdala, the central medial amygdala, the superficial amygdala, and the paralaminar amygdala—as the research objects.

[0117] It should be noted that the T1-weighted images mentioned above are a type of medical imaging. In MRI scans, T1-weighted images are obtained using a specific pulse sequence that highlights the contrast between tissues, particularly based on the tissue's T1 relaxation time—the rate at which protons in the tissue return from a high-energy state to a low-energy state. Therefore, T1-weighted images are commonly used to assess anatomical structures because they provide good tissue contrast, making the boundaries between different types of tissue (such as gray matter, white matter, and cerebrospinal fluid) clearer. In brain MRI, T1-weighted images are particularly helpful in observing structural details of the brain, including structures such as the cerebral cortex, white matter fiber tracts, ventricles, basal ganglia, and amygdala.

[0118] Finally, the computer used seed-point-based sFC and dFC analysis to analyze the sFC and dFC connectivity patterns of the bilateral amygdala and its subregions (identified using a probability map defined by the amygdala cell structure) for each test subject.

[0119] (1) The sFC index was obtained by extracting the time series average of the amygdala seed point region of interest and performing Pearson correlation analysis on it with the time series of each voxel in the whole brain. The correlation coefficient (r value) is the sFC between each seed point and other voxels in the whole brain. To improve normality, Fisher's r-to-z transformation (also known as Fisher's z-transform, a statistical transformation used to convert the correlation coefficient (r) into an approximately normally distributed variable. This transformation is particularly useful because the distribution of a single correlation coefficient is not normal, especially with a small sample size. By applying Fisher's r-to-z transformation, hypothesis testing, confidence interval calculation, or meta-analysis can be performed more easily) was used to transform the single correlation matrix into a z-score matrix. Finally, each subject obtained an sFC brain map of 8 amygdala subregions in both brains (i.e., using 4 amygdala subregions in each brain). Finally, all sFC brain maps were further smoothed using a smoothing kernel with a full width at half maximum (FWHM) of 6 mm. The sFC index includes:

[0120] Correlation indicators are used to assess the level of static synchronization activity between the amygdala and other brain regions.

[0121] Phase synchronization indexes assess functional connectivity by analyzing the phase relationships of time series from different regions.

[0122] The Granger causality index is used to assess whether one time series can help predict another time series, thereby revealing potential causal relationships.

[0123] (2) dFC indices: The dFC variability of the amygdala was calculated using the Temporal Dynamic Analysis (TDA) toolkit in DPABI software. Hamming sliding windows were used for time series of whole-brain oxygenation level-related signals, resulting in 151 dFC sliding windows. For a given seed point, the Pearson correlation coefficient between the seed point's time series and the whole-brain voxel time series within any time window was calculated. To improve data normality, all 151 correlation coefficient matrices were subjected to Fisher r-to-z transformation, resulting in 151 z-value matrices. The standard deviation of the 151 z-values ​​for each voxel was calculated, representing the dFC value of that voxel. All dFC brain maps were smoothed using a smoothing kernel with a full width at half maximum (FWHM) of 6 mm. dFC indices include:

[0124] Time variability metrics are used to analyze changes in functional connectivity over time and to understand the dynamic characteristics of amygdala connection patterns.

[0125] State switching indicators are used to assess changes in amygdala functional connectivity patterns in different brain states (such as rest and task performance).

[0126] Modularity indicators are used to analyze how the amygdala forms and detaches from other brain regions as functional modules at different time points.

[0127] Thus, by comparing the sFC and dFC before and after treatment, the region with the most significant changes is identified as the characteristic brain region. Based on the method described above for identifying the characteristic brain region, step S1-2 further includes:

[0128] S1-2-2, analyze the characteristic brain regions in the first image of each test subject's brain to obtain the first depression data for each test subject.

[0129] In this embodiment, for each test subject, the computer can obtain the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala in the characteristic brain regions of the test subject before treatment based on the test subject's first image, and use the static functional connectivity index of the amygdala and the dynamic functional connectivity index of the amygdala before treatment as the test subject's first depression data.

[0130] S1-2-3, analyze the characteristic brain regions in the second image of each test subject's brain to obtain the second depression data for each test subject.

[0131] In this context, the second depression data for each test subject represents the depressive characteristics data after treatment. Similarly to the first depression data, for each test subject, the computer obtains the static functional connectivity index and dynamic functional connectivity index of the amygdala after treatment of the characteristic brain regions based on the test subject's second image; and uses the static functional connectivity index and dynamic functional connectivity index of the amygdala after treatment as the test subject's second depression data.

[0132] S1-2-4: Based on the first and second depression data of each test subject, the actual therapeutic effect data of each test subject is obtained.

[0133] Regarding the aforementioned efficacy measurement data, it should be understood that the severity of depression is typically quantified using the Hamilton Depression Rating Scale (HDR). This scale includes multiple parameters or items covering different aspects of depressive symptoms. Therefore, for each test subject, this embodiment maps the subject's first depression data to a first score using the HDR; maps the subject's second depression data to a second score using the HDR; and finally, the difference between the first and second scores is used as the test subject's efficacy measurement data. Of course, in some other embodiments, the second score can also be directly used as the test subject's efficacy measurement data.

[0134] Continue to see Figure 1 Based on the pre- and post-treatment depression data of the test subjects and the measured efficacy data, the MRI-based depression efficacy prediction method provided in this embodiment further includes:

[0135] S2, based on the first depression data of multiple test subjects and the actual efficacy test data, train the model to be trained to obtain an efficacy prediction model that meets the preset conditions.

[0136] In this model, the initial depression data from multiple test subjects serves as the input data for training the model, while the measured efficacy data from multiple test subjects serves as the supervision data for training the model. The efficacy prediction model is used to predict the efficacy of treatment for depression. Figure 3 As shown, Figure 1 An optional implementation of step S2 shown includes:

[0137] S2-1 divides the initial depression data and efficacy test data of multiple test subjects into multiple subsets.

[0138] Each subset of the dataset includes initial depression data and efficacy test data from a portion of the test participants. For example, a leave-one-out approach based on an unbiased strategy can be used to divide the test participants into subsets. That is, for each of the multiple test participants, the initial depression data and efficacy test data of one participant are retained for each training iteration, while the initial depression data and efficacy test data of the remaining participants are used as the training subset.

[0139] S2-2, train the model to be trained according to each subset of data, and obtain a set of initial fitting results for multiple parameters to be fitted from each subset of data.

[0140] In an optional implementation, during each training iteration, the correlation coefficient between the measured efficacy data and the predicted efficacy data after light intervention therapy for each test subject is calculated to measure the training effectiveness of the model. The expression for this correlation coefficient is as follows:

[0141]

[0142] Among them, y n For the measured efficacy data of the nth test subject, f(x) n ) represents the efficacy prediction data for the nth test subject, μ y and μ f These represent the average efficacy and the predicted efficacy for n test subjects, respectively.

[0143] Thus, the regression performance of the model can be evaluated by the correlation between the efficacy prediction data and the measured efficacy data. Exemplarily, the preset condition set in this embodiment is -1 < CORR < 1. At this time, it represents the linear correlation strength between the measured efficacy data and the efficacy prediction data. Being close to 0 indicates weak correlation or even no correlation, being close to 1 indicates positive correlation, and being close to -1 indicates negative correlation. When the CORR value is less than 0, it means that the performance of the model is poor, which means that the efficacy prediction data obtained from larger measured efficacy data is smaller than the efficacy prediction data obtained from smaller measured efficacy data.

[0144] Thus, multiple initial fitting parameters of multiple initial efficacy prediction models can be obtained from multiple sub-datasets. Continuing to refer to Figure 3 , based on multiple groups of initial fitting results, step S2 further includes:

[0145] S2-3, according to multiple groups of initial fitting results, obtain the target fitting results of multiple fitting parameters to be fitted.

[0146] In this regard, it should be noted that the model to be trained includes multiple fitting parameters to be fitted. Exemplarily, in this embodiment, the difference in the Hamilton Depression Scale scores before and after the light intervention therapy for each tester can be used as a regression index for multiple linear regression, so as to obtain an efficacy prediction model. At this time, if the first depression data includes 6 indicators, the expression of the model to be trained can be expressed as:

[0147] y = a1x1 + a2x2 + a3x3 + a4x4 + a5x5 + a6x6 + b;

[0148] After training, the fitting results of multiple fitting parameters (a1, a2, a3, a4, a5, a6, b) in the above expression are obtained, so as to obtain an efficacy prediction model. In the above embodiment, it has been introduced that the first depression data and the second depression data of multiple testers are divided into multiple sub-datasets, which means that each sub-dataset will fit an initial efficacy prediction model, that is, a set of initial fitting results corresponding to multiple fitting parameters. Therefore, the computer can obtain the average fitting result of each fitting parameter according to multiple groups of initial fitting results; respectively use the average fitting result of each fitting parameter as the target fitting result of each fitting parameter.

[0149] Exemplarily, take the first fitting parameter a1 in the above expression as an example. Assuming that there are 5 sub-datasets in total, 5 initial fitting results of a1 can be obtained. The average of these 5 initial fitting results is calculated to obtain the final target fitting result of the fitting parameter a1.

[0150] S2-4, according to the target fitting results of multiple fitting parameters to be fitted, obtain an efficacy prediction model.

[0151] In summary, the efficacy prediction model obtained using the above embodiments can predict the difference in Hamilton Depression Rating Scale scores based on any patient's sFC and dFC values ​​before light intervention therapy, serving as predictive data for the efficacy of light intervention therapy for that patient. Doctors can then use this predictive data to determine whether the therapeutic effect is significant.

[0152] Based on the same inventive concept as the MRI-based method for predicting the efficacy of treatment for depression provided in this embodiment, this embodiment also provides a related device. Specifically, this embodiment also provides an MRI-based device for predicting the efficacy of treatment for depression, which includes at least one software functional module that can be stored in memory 21 or embedded in an electronic device. The processor 22 in the electronic device is used to execute the executable module stored in memory 21. For example, the software functional modules and computer programs included in a depression efficacy prediction model training device, etc. Please refer to... Figure 4 Functionally, a training device for predicting the efficacy of depression treatment can include:

[0153] Data processing module 11 acquires the first depression data and efficacy test data of multiple test subjects, wherein the first depression data of each test subject represents the depression characteristics data before treatment;

[0154] The model training module 12 trains the model to be trained based on the first depression data of multiple test subjects and the actual efficacy test data, so as to obtain an efficacy prediction model that meets the preset conditions. The first depression data of multiple test subjects is the input data when training the model to be trained, and the actual efficacy test data of multiple test subjects is the supervision data when training the model to be trained.

[0155] In this embodiment, the data processing module 11 is used to implement... Figure 1 In step S1, the model training module 12 is used to implement... Figure 1 Step S2 in the above process. For a detailed description of each module, please refer to the specific implementation method of the corresponding step, which will not be repeated in this embodiment.

[0156] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0157] It should also be understood that if the above embodiments are implemented as software functional modules and sold or used as independent products, they 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 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, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0158] Therefore, this embodiment also provides a storage medium, which is a computer-readable storage medium. This storage medium stores a computer program, which, when executed by a processor, implements the MRI-based method for predicting the efficacy of depressive therapy provided in this embodiment. The storage medium can be any medium capable of storing program code, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0159] Furthermore, this embodiment also provides an electronic device for implementing the above-described MRI-based method for predicting the efficacy of depressive therapy. For example... Figure 5 As shown, the electronic device may include a processor 22 and a memory 21. The memory 21 stores a computer program, and the processor reads and executes the computer program corresponding to the above-described embodiments in the memory 21 to implement the MRI-based method for predicting the efficacy of treatment for depression provided in this embodiment.

[0160] Continue to see Figure 5 The electronic device also includes a communication unit 23. The memory 21, processor 22 and communication unit 23 are electrically connected to each other directly or indirectly through system bus 24 to realize data transmission or interaction.

[0161] The memory 21 can be an information recording device based on any electronic, magnetic, optical, or other physical principles, used to record execution instructions, data, etc. In some embodiments, the memory 21 can be, but is not limited to, volatile memory, non-volatile memory, memory drive, etc.

[0162] In some embodiments, the volatile memory may be random access memory (RAM); in some embodiments, the non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory, etc.; in some embodiments, the storage drive may be a disk drive, solid-state drive, any type of storage disk (such as optical disc, DVD, etc.), or similar storage media, or a combination thereof.

[0163] The communication unit 23 is used to send and receive data over a network. In some embodiments, the network may include a wired network, a wireless network, a fiber optic network, a telecommunications network, an intranet, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network, or a near field communication (NFC) network, or any combination thereof. In some embodiments, the network may include one or more network access points. For example, the network may include wired or wireless network access points, such as base stations and / or network switching nodes, through which one or more components of the service request processing system can connect to the network to exchange data and / or information.

[0164] The processor 22 may be an integrated circuit chip with signal processing capabilities, and may include one or more processing cores (e.g., a single-core processor or a multi-core processor). By way of example only, the processor described above may include a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), an Application Specific Instruction-set Processor (ASIP), a Graphics Processing Unit (GPU), a Physics Processing Unit (PPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), a Programmable Logic Device (PLD), a controller, a microcontroller unit, a Reduced Instruction Set Computing (RISC) computer, or a microprocessor, or any combination thereof.

[0165] Understandable. Figure 5 The structure shown is for illustrative purposes only. Electronic devices may also have more advanced features. Figure 5 Showing more or fewer components, or having with Figure 5 The different configurations shown. Figure 5 The components shown can be implemented using hardware, software, or a combination thereof.

[0166] It should be understood that the apparatus and methods disclosed in the above embodiments can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0167] The above descriptions are merely various embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting the efficacy of treatment for depression based on MRI, characterized in that, The method includes: First and second images of the brains of multiple test subjects were acquired, wherein the first and second images of the brains of each test subject were medical brain images before and after light intervention therapy. Based on the brain biometric indicators in the first and second images of the brains of the multiple test subjects, characteristic brain regions that changed before and after brain treatment were identified. The brain biometric indicators are the static functional connectivity indicators and dynamic functional connectivity indicators of the amygdala before and after the entire brain treatment. The characteristic brain regions include the basolateral amygdala, the central medial amygdala, the superficial amygdala, and the amygdala in the layer amyloid transition zone. Based on the first depression data and the second depression data of each test subject, the measured efficacy data of each test subject were obtained. The first depression data and the second depression data of each test subject represent the static functional connectivity index and the dynamic functional connectivity index of the amygdala before and after treatment of the characteristic brain region, respectively. These data were obtained by analyzing the brain medical images of the test subjects before and after treatment. The first depression data and efficacy test data of the multiple test subjects are divided into multiple subsets using the leave-one-out method. The first depression data of the multiple test subjects is the input data when training the model to be trained, and the efficacy test data of the multiple test subjects is the supervision data when training the model to be trained. The model to be trained includes multiple parameters to be fitted, and each subset includes the first depression data and efficacy test data of a portion of the test subjects. The model to be trained is trained according to each of the said subsets of data to obtain a set of initial fitting results for the multiple parameters to be fitted for each of the said subsets of data. Based on the multiple sets of initial fitting results, the average fitting result of each parameter to be fitted is obtained; The average fitting result of each of the parameters to be fitted is taken as the target fitting result of each of the parameters to be fitted. Based on the target fitting results of the multiple parameters to be fitted, an efficacy prediction model is obtained, wherein the efficacy prediction model is used to predict the efficacy of depression treatment.

2. The MRI-based method for predicting the efficacy of treatment for depression according to claim 1, characterized in that, Based on the brain biometric indicators in the first and second images of the brains of the multiple test subjects, characteristic brain regions that changed before and after brain treatment were identified, including: Based on the first image of each test subject's brain, brain biometric indicators were obtained for each test subject before treatment. Based on the second image of each test subject's brain, the brain biometric indicators of each test subject after treatment were obtained; The brain biometrics of each test subject before and after treatment were compared to identify the characteristic brain regions.

3. A device for predicting the efficacy of treatment for depression based on MRI, characterized in that, The device includes: The data processing module is used to acquire first and second images of the brains of multiple test subjects, wherein the first and second images of the brains of each test subject are medical brain images before and after light intervention therapy. Based on the brain biometric indicators in the first and second images of the brains of the multiple test subjects, characteristic brain regions that changed before and after brain treatment were identified. The brain biometric indicators are the static functional connectivity indicators and dynamic functional connectivity indicators of the amygdala before and after the entire brain treatment. The characteristic brain regions include the basolateral amygdala, the central medial amygdala, the superficial amygdala, and the amygdala in the layer amyloid transition zone. Based on the first depression data and the second depression data of each test subject, the measured efficacy data of each test subject were obtained. The first depression data and the second depression data of each test subject represent the static functional connectivity index and the dynamic functional connectivity index of the amygdala before and after treatment of the characteristic brain region, respectively. These data were obtained by analyzing the brain medical images of the test subjects before and after treatment. The model training module divides the first depression data and efficacy test data of the multiple test subjects into multiple subsets using the leave-one-out method. The first depression data of the multiple test subjects is the input data for training the model to be trained, and the efficacy test data of the multiple test subjects is the supervision data for training the model to be trained. The model to be trained includes multiple parameters to be fitted, and each subset includes the first depression data and efficacy test data of a portion of the test subjects. The model to be trained is trained according to each of the said subsets of data to obtain a set of initial fitting results for the multiple parameters to be fitted for each of the said subsets of data. Based on the multiple sets of initial fitting results, the average fitting result of each parameter to be fitted is obtained; The average fitting result of each of the parameters to be fitted is taken as the target fitting result of each of the parameters to be fitted. Based on the target fitting results of the multiple parameters to be fitted, an efficacy prediction model is obtained, wherein the efficacy prediction model is used to predict the efficacy of depression treatment.

4. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the MRI-based method for predicting the efficacy of depressive therapy as described in any one of claims 1-2.

5. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing a computer program, which, when executed by the processor, implements the MRI-based method for predicting the efficacy of depression treatment as described in any one of claims 1-2.

Images