Strabismus detection method based on causal feature gating

Abstract

The invention discloses a strabismus detection method based on causal feature gating, and the method comprises the steps: carrying out the modeling analysis of a causal relationship between a key point coordinate and a strabismus category through a causal discovery algorithm, obtaining a causal relationship model used for determining the causal relationship, screening the key point coordinate based on the causal relationship model, and obtaining a strabismus detection result. Extracting a plurality of causal variables from a plurality of samples of the training sample set according to the causal variable set; and introducing a causal feature gating module between adjacent network layers of the classification model, training the classification model by using a plurality of causal variables, and performing adaptive adjustment on gating parameters in the training process of the classification model until the classification model converges or reaches a preset training round number, thereby obtaining a final classification model. According to the embodiment of the invention, the accuracy of strabismus classification can be improved, and the interpretability and medical application value of the model can be remarkably enhanced.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a strabismus detection method based on causal feature gating. Background Technology

[0002] Currently, deep learning technology has been widely applied in the field of medical image analysis, among which models such as convolutional neural networks (CNNs) have shown high recognition accuracy in ophthalmic disease detection and classification tasks. In strabismus classification research, common methods usually rely on end-to-end feature extraction and classification of eye images directly. The model can automatically learn the high-dimensional features of the image and complete the differentiation between strabismus and normal samples. Summary of the Invention

[0003] The purpose of this application is to provide a strabismus detection method based on causal feature gating, in order to solve the technical problem that existing strabismus determination models ignore causal relationships, resulting in poor reliability. The various technical effects of the preferred solutions among the many technical solutions provided in this application are detailed below.

[0004] To achieve the above objectives, this application provides the following technical solutions: The first aspect of this application provides a strabismus detection method based on causal feature gating, comprising: acquiring the coordinates of several key points in the eye region of a face; using a causal discovery algorithm to model and analyze the causal relationship between the key point coordinates and the strabismus category to obtain a causal relationship model for determining the causal relationship; filtering the key point coordinates based on the causal relationship model to obtain a set of causal variables; extracting multiple causal variables from several samples in a training sample set according to the set of causal variables; introducing a causal feature gating module between adjacent network layers of a classification model, the causal feature gating module including gating parameters for characterizing the contribution weights of corresponding causal variables, the causal feature gating module being used to weight and modulate each causal variable; training the classification model using multiple causal variables, while adaptively adjusting the gating parameters during the training process of the classification model until the classification model converges or reaches a preset number of training rounds to obtain a final classification model; using the final classification model to predict the sample to be predicted to obtain the prediction result, and simultaneously using SHAP (SHapley Additive)... The exPlanations interpretation method is used to interpret the prediction results to obtain visualization results, and the strabismus category is determined by combining the prediction results and the visualization results.

[0005] In some embodiments, the step of using a causal discovery algorithm to model and analyze the causal relationship between the keypoint coordinates and the strabismus category to obtain a causal relationship model for determining the causal relationship includes: generating several candidate causal relationship models based on the keypoint coordinates; constructing a corresponding structural equation model for each candidate causal relationship model; performing parameter estimation and model fitting on each structural equation model; performing a conditional independence test based on the fitting results; and selecting models that pass the conditional independence test; scoring all models that pass the conditional independence test; and obtaining the causal structure diagram of the causal relationship model with the best score as the causal structure between the keypoint coordinates and the strabismus category, used to determine the causal relationship.

[0006] In some embodiments, the step of filtering the key point coordinates based on the causal relationship model to obtain a set of causal variables includes: determining the target node corresponding to the strabismus category from the causal structure diagram, and obtaining a set of variables that form a Markov blanket with the target node, wherein the key point coordinate variables in the set of variables are determined as causal variables related to the strabismus category.

[0007] In some embodiments, the step of extracting multiple causal variables from several samples in the training sample set based on the set of causal variables includes: randomly selecting one sample from the same group of samples in the training sample set as the original sample, and the remaining samples as augmented samples; extracting the causal variables from the original sample and the augmented samples respectively; constructing the causal variables into a one-dimensional feature array, wherein the one-dimensional feature array of the original sample is used as the main input of the classification model, and the one-dimensional feature array of the augmented sample is used as the auxiliary input of the classification model.

[0008] In some embodiments, training the classification model using multiple causal variables while adaptively adjusting the gating parameters during the training process includes: forward propagating the multiple causal variables sequentially through linear transformations, causal feature gating modulation, and nonlinear activation functions at each layer of the classification model to obtain predicted values; calculating a loss function based on the difference between the predicted values ​​and the true labels, and constructing consistency constraints; calculating the gradient of the loss function with respect to the network parameters and the gating parameters in the classification model using a backpropagation algorithm, and jointly updating the network parameters and the gating parameters using an optimization algorithm.

[0009] In some embodiments, the Markov blanket variable of the target node includes the parent node of the target node, the child nodes of the target node, and the parent node of the child nodes of the target node.

[0010] In some embodiments, the classification model is a multilayer perceptron model, which includes at least one hidden layer.

[0011] In some embodiments, the structural equation model is represented by the following formula: in, Indicates the first Key point coordinate variables, Represents the coordinate variables of the key points The set of parent node variables, This represents the corresponding functional relationship. This represents mutually independent noise terms.

[0012] In some embodiments, the method includes: using the least squares method or the maximum likelihood estimation method to perform parameter estimation and model fitting on the structural equation model.

[0013] A second aspect of this application provides a processing apparatus, comprising: one or more processors; a memory for storing one or more computer programs, wherein the one or more processors are configured to execute the one or more computer programs stored in the memory to cause the one or more processors to perform the strabismus detection method based on causal feature gating as described above.

[0014] Implementing one of the above-mentioned technical solutions of this application has the following advantages or beneficial effects: In this application, firstly, a causal relationship model for determining causal relationships is obtained through a causal discovery algorithm, then the causal relationship model is used to filter out a set of causal variables from the key point coordinates, and then these causal vectors are extracted from the training sample set to train the classification model. At the same time, during the training process, a causal feature gating module is introduced to make the gating parameters adaptively updated, thereby realizing the joint optimization of the classification model network parameters and the gating parameters, and obtaining the final classification model for obtaining the prediction results of the strabismus category, and the prediction results are visualized.

[0015] In this context, by performing causal discovery and causal variable construction on key points of the eye, key causal features with clear medical semantics can be extracted. Simultaneously, a causal feature gating module is introduced into the classification model, enabling it to adaptively assign importance weights to causal variables. Combined with the SHAP interpretation method, the transparency and clinical usability of the prediction results are enhanced. Therefore, the embodiments of this application not only improve the accuracy of strabismus classification but also significantly enhance the interpretability and medical application value of the model. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a schematic flowchart of the strabismus detection method with causal feature gating according to an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, various exemplary embodiments described below will be referenced to the accompanying drawings, which form part of the exemplary embodiments and depict various exemplary embodiments that may be adopted to implement this application. Unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. It should be understood that they are merely examples of processes, methods, and apparatuses consistent with some aspects of this application disclosed as detailed in the appended claims, and other embodiments may be used, or structural and functional modifications may be made to the embodiments listed herein without departing from the scope and spirit of this application.

[0018] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," etc., indicate the orientation or positional relationship based on the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the referred element must have a specific orientation, or be constructed and operated in a specific orientation. The terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. The term "multiple" means two or more. The terms "connected" and "linked" should be interpreted broadly, for example, they can be fixed connections, detachable connections, integral connections, mechanical connections, electrical connections, communication connections, direct connections, indirect connections through an intermediate medium, and can be the internal connection of two elements or the interaction relationship between two elements. The term "and / or" includes any and all combinations of one or more of the related listed items. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0019] To illustrate the technical solutions described in this application, specific embodiments are provided below, showing only the parts related to the embodiments of this application.

[0020] like Figure 1 As shown, this application provides a strabismus detection method based on causal feature gating, including the following steps (steps S1 to S3): S1. Obtain the coordinates of several key points in the eye region of the face. Use the causal discovery algorithm to model and analyze the key point coordinates and strabismus category to obtain a causal relationship model for determining causal relationships. Based on the causal relationship model, filter the key point coordinates to obtain a set of causal variables. Extract multiple causal variables from several samples in the training sample set according to the set of causal variables.

[0021] Specifically, the process begins by acquiring a facial image of the subject to be detected. Then, based on a pre-defined facial landmark algorithm, the coordinates of several key points in the eye region of the face are extracted from the facial image. Alternatively, facial images can be obtained from publicly available general face recognition datasets, such as the CelebA dataset and the LFW dataset, or facial images generated using AI (artificial intelligence) technology. The eye region of the face can be located from the facial image, and then key point information for the eyes can be extracted. This key point information can include key point coordinates. Specifically, the Mediapiple framework can be used to extract the x and y coordinates of all key points in the eye region. The Mediapiple framework is a multimodal machine learning model application framework that can be used to perform visual perception tasks, such as face detection.

[0022] Subsequently, based on the training sample set, a causal discovery algorithm can be used to model and analyze the causal relationship between keypoint coordinates and strabismus categories, resulting in a causal relationship model. This model identifies causal variables directly associated with strabismus categories. Based on the causal relationship model, Markov blanket theory is used to filter the keypoint coordinates, obtaining a set of causal variables related to strabismus categories. Finally, the corresponding causal variables can be extracted from this set to construct the input for the classification model.

[0023] In some embodiments, a causal discovery algorithm is used to model and analyze keypoint coordinates and strabismus categories to obtain a causal relationship model for determining causal relationships. This may include: generating several candidate causal relationship models based on keypoint coordinates; constructing a corresponding structural equation model for each candidate causal relationship model; performing parameter estimation and model fitting for each structural equation model; performing conditional independence tests based on the fitting results; and selecting models that pass the conditional independence tests. All models that pass the conditional independence tests are scored, and the causal structure diagram of the causal relationship model with the best score is obtained as the causal structure between keypoint coordinates and strabismus categories to determine causal relationships.

[0024] In some embodiments, a structural equation model can be represented by the following formula: in, It can represent the first Key point coordinate variables, It can represent the coordinate variables of key points. The set of parent node variables, It can represent the corresponding functional relationship. It can represent mutually independent noise terms. Structural equation modeling can be used to describe the dependencies between causal variables.

[0025] Specifically, the least squares method or the maximum likelihood estimation method can be used to estimate parameters and fit the model of the structural equation model. The least squares method is used to minimize the sum of squared prediction errors, while the maximum likelihood estimation method is used to maximize the probability of observed data.

[0026] In some embodiments, parameter estimation and model fitting can be performed on each structural equation model, and a conditional independence test can be performed based on the fitting results to screen candidate causal relationship models that meet preset independence constraints. The candidate causal relationship models that pass the conditional independence test are scored, and the causal relationship model with the best score is selected according to a preset scoring criterion. The causal structure diagram corresponding to the best-scoring causal relationship model is used as the causal structure between keypoint coordinates and the squint category to determine the causal relationship. Further, it can be tested whether the error terms of the fitted model meet the independence assumption. If there is a statistical dependency between the error terms, it indicates that the causal structure diagram corresponding to that model is invalid. This is used to determine which models pass the independence test, thereby obtaining the models that meet the preset independence constraints, i.e., those that pass the conditional independence test.

[0027] In some embodiments, filtering keypoint coordinates based on a causal relationship model to obtain a set of causal variables may include: determining the target node corresponding to the strabismus category from the causal structure graph, and obtaining a set of variables that form a Markov blanket with the target node, wherein the keypoint coordinate variables in the variable set are determined as causal variables related to the strabismus category. Specifically, the Markov blanket variables of the target node corresponding to the strabismus category may include the parent node of the target node, the child nodes of the target node, and the parent nodes of the child nodes of the target node. Unlike existing high-dimensional features that rely on CNNs (Convolutional Neural Networks), the causal variables obtained through causal relationship filtering are low-dimensional, semantically clear, and medically interpretable. Furthermore, by eliminating redundant features through causal relationships, the consistency between features and disease representation and the model's generalization ability can be improved.

[0028] In some embodiments, extracting multiple causal variables from several samples in a training sample set based on a set of causal variables may include: randomly selecting one sample from the same group of samples in the training sample set as the original sample, and the remaining samples as augmented samples; extracting causal variables from the original sample and the augmented samples respectively; constructing a one-dimensional feature array for the causal variables, wherein the one-dimensional feature array of the original sample serves as the main input to the classification model, and the one-dimensional feature array of the augmented samples serves as the auxiliary input to the classification model. This can be achieved by grouping multiple face image samples belonging to the same target object acquired within the same time period into the same group of samples.

[0029] The original samples and augmented samples can be used together in the training process of the classification model. Compared with the high-dimensional image features extracted by CNN, the causal variables in this application can be represented more compactly and have high-level semantic information.

[0030] For example, after the causal discovery algorithm and causal relationship are determined, X causal variables that are causally related to the strabismus category are selected to form a causal variable set. Then, when extracting multiple causal variables from several samples in the training sample set, at least these X causal variables in the causal variable set can be extracted.

[0031] S2. Introduce a causal feature gating module between adjacent network layers of the classification model. The causal feature gating module includes gating parameters that characterize the contribution weights of the corresponding causal variables. The causal feature gating module is used to weight and modulate each causal variable, and the classification model is trained using multiple causal variables. At the same time, the gating parameters are adaptively adjusted during the training process of the classification model until the classification model converges or reaches the preset number of training rounds, thus obtaining the final classification model.

[0032] Specifically, the causal feature gating module may include learnable gating parameters that correspond one-to-one with causal variables, used to weight and modulate the causal feature representation input to the current network layer.

[0033] During the forward propagation of the classification model, the causal feature gating module can perform element-wise or channel-wise gating modulation operations on the feature vectors from the previous network layer to obtain the gated feature representation, and use the gated feature representation as the input of the current network layer.

[0034] During the training of the classification model, the gating parameters can be used as trainable parameters in backpropagation. By minimizing the preset loss function, the gating parameters and network parameters are jointly optimized, so that the gating parameters are adaptively adjusted during training. This dynamically enhances the features that contribute causally to the classification results and suppresses features that are weakly related to the classification task or are redundant, until the classification model converges or reaches the preset number of training rounds, and the final classification model is obtained.

[0035] To address the data imbalance issue caused by the relatively small number of strabismus samples and the large number of non-strabismus samples in real-world application scenarios, this application's embodiments employ a group-based training strategy during the model training phase: Specifically, multiple facial images of the same subject collected within the same time period are combined into a sample, where the samples within the group differ in terms of the type of shooting device, shooting angle, or shooting environment.

[0036] In each training iteration, one sample is selected from each group of samples as the original sample, and the remaining samples are used as augmented samples. The training data within the group is constructed based on the original sample and the augmented samples.

[0037] During model training, the causal variables corresponding to the same set of samples are input into the classification model to obtain multiple prediction results within the group. Based on the prediction results, a group consistency loss function is constructed. By minimizing the group consistency loss function, the gating parameters in the causal feature gating module and the network parameters in the classification model are jointly optimized. This allows the gating parameters to be adaptively adjusted during training, thereby enhancing the response capability to features with stable causal contributions and suppressing non-causal interference features introduced by changes in shooting conditions.

[0038] The gating parameter, ranging from 0 to 1, can be viewed as multiplying each causal variable by a soft mask, i.e., a causal mask. This mask adds a weight or probability to the original sample, effectively weighting the original information and a general "masked representation." Through adaptive updates of the gating parameter, the classification model can assign different weights based on the importance of the features.

[0039] In some embodiments, training a classification model using multiple causal variables while adaptively adjusting the gating parameters during the training process can include: Multiple causal variables are sequentially transformed through linear transformations, causal feature gating modulation, and nonlinear activation functions at each layer of the classification model, and forward propagation is performed to obtain predicted values. A loss function is calculated based on the difference between the predicted values ​​and the true labels, and group consistency constraints are constructed. The gradient of the loss function with respect to the network parameters and gating parameters in the classification model is calculated using the backpropagation algorithm, and the network parameters and gating parameters are jointly updated using an optimization algorithm. Causal variables are sequentially transformed through linear transformations and nonlinear activation functions at each layer of the classification model, progressively extracting, transforming, and combining features at each layer to obtain a higher-level, more abstract representation.

[0040] Specifically, causal variables from the main input and auxiliary input can be input into the classification model, allowing them to pass through linear transformations, causal feature gating modulation, and nonlinear activation functions at each layer of the classification model in sequence, performing forward propagation to obtain the corresponding predicted values. Then, the loss function can be calculated based on the difference between the predicted values ​​and the true labels, and group consistency constraints can be constructed based on the prediction results of samples in the same group. Finally, the gradient of the loss function with respect to the network parameters and causal feature gating parameters in the classification model can be calculated using the backpropagation algorithm, and the network parameters and causal feature gating parameters can be jointly updated using an optimization algorithm.

[0041] In some embodiments, the classification model can be a multilayer perceptron (MLP) model, which includes at least one hidden layer.

[0042] During each training iteration, multiple sets of samples from the training sample set can be input simultaneously, from which original samples and augmented samples can be randomly selected, thereby enriching the diversity of the training data. In other embodiments, the classification model can be logistic regression, random forest, support vector machine, or a graph neural network-based model.

[0043] During training, the classification model can simultaneously utilize both original and augmented samples to optimize the gating parameters. The causal feature gating module can adjust the gating parameters based on the consistency constraints of the group data to ensure that the prediction results corresponding to the same group of samples remain consistent within a preset range. The consistency constraints make the high-level feature representations of the original and augmented samples as close as possible. The causal feature gating module can adjust the gating weight of each causal variable according to the synergistic effect of the consistency constraints and the task objective, thereby highlighting key causal features while maintaining feature consistency and improving the robustness and generalization ability of the final classification model.

[0044] The causal feature gating module can adjust the gating parameters based on the consistency constraints of group data during model training, so that the prediction results corresponding to the same group of samples remain consistent within a preset range. This guides the gating parameters to enhance the response to stable causal features and suppress non-causal features introduced by changes in collection conditions.

[0045] S3. Use the final classification model to predict the sample to be predicted and obtain the prediction results. At the same time, use the SHAP interpretation method to interpret the prediction results and obtain the visualization results. Combine the prediction results and visualization results to determine the strabismus category.

[0046] Specifically, after the model training is completed, the final classification model obtained from the training can be used to predict the sample to be predicted and obtain the corresponding strabismus classification prediction result. At the same time, based on the SHAP interpretation method, feature contribution analysis is performed on the prediction result to generate a visualization result of the degree of influence of causal variables on the classification result. By combining the prediction result and the visualization result, the strabismus category of the object to be detected is determined.

[0047] Specifically, the SHAP (SHapley Additive exPlanations) method is a unified approach to interpreting the predictions of any machine learning model. It can be used to calculate the contribution of each causal variable to the prediction, and the visualization results can show the weight of each causal variable in the final classification decision. In this case, the visualization results allow for an intuitive understanding of the model's judgment criteria, improving the credibility of the model's results.

[0048] In this application, a causal relationship model for determining causal relationships is first obtained through a causal discovery algorithm. Then, the causal relationship model is used to filter out a set of causal variables from the key point coordinates. These causal vectors are then extracted from the training sample set to train a classification model. During the training process, a causal feature gating module is introduced to adaptively update the gating parameters, thereby achieving joint optimization of the classification model network parameters and the gating parameters. This results in a final classification model for obtaining the prediction results of strabismus categories, and the prediction results are then visualized.

[0049] In this context, by performing causal discovery and causal variable construction on key points of the eye, key causal features with clear medical semantics can be extracted. Simultaneously, a causal feature gating module is introduced into the classification model, enabling it to adaptively assign importance weights to causal variables. Combined with the SHAP interpretation method, the transparency and clinical usability of the prediction results are enhanced. Therefore, the embodiments of this application not only improve the accuracy of strabismus classification but also significantly enhance the interpretability and medical application value of the model.

[0050] Those skilled in the art will understand that all or part of the features / steps of the above-described method embodiments can be implemented by methods, data processing systems, or computer programs. These features may be implemented without hardware, entirely in software, or in a combination of hardware and software. The aforementioned computer program may be stored in one or more computer-readable storage media. When the computer program is executed (e.g., by a processor), it performs the steps of the above-described strabismus detection method embodiments based on causal feature gating.

[0051] The aforementioned storage media capable of storing program code include: static hard disks, solid-state hard disks, random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), optical storage devices, magnetic storage devices, flash memory, magnetic disks or optical disks, and / or combinations of the above devices, that is, they can be implemented by any type of volatile or non-volatile storage devices or combinations thereof.

[0052] This application also provides a processing device embodiment, including one or more processors and a memory; wherein the memory is used to store one or more computer programs, and the one or more processors are used to execute the one or more computer programs stored in the memory, so that the processors execute the features / steps of the strabismus detection method embodiment based on causal feature gating described above.

[0053] The above description is merely a preferred embodiment of this application. Those skilled in the art will understand that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of this application. Furthermore, under the teachings of this application, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of this application. Therefore, this application is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of this application.

Claims

1. A strabismus detection method based on causal feature gating, characterized in that, include: The coordinates of several key points in the eye region of a face are obtained. The causal relationship between the key point coordinates and the strabismus category is modeled and analyzed using a causal discovery algorithm to obtain a causal relationship model for determining the causal relationship. The key point coordinates are filtered based on the causal relationship model to obtain a set of causal variables. Multiple causal variables are extracted from several samples in the training sample set according to the set of causal variables. A causal feature gating module is introduced between adjacent network layers of the classification model. The causal feature gating module includes gating parameters for characterizing the contribution weights of corresponding causal variables. The causal feature gating module is used to perform weighted modulation on each causal variable and train the classification model using multiple causal variables. At the same time, the gating parameters are adaptively adjusted during the training process of the classification model until the classification model converges or reaches a preset number of training rounds to obtain the final classification model. The final classification model is used to predict the sample to be predicted, and the prediction result is obtained. At the same time, the SHAP interpretation method is used to interpret the prediction result to obtain the visualization result. The strabismus category is determined by combining the prediction result and the visualization result.

2. The strabismus detection method based on causal feature gating according to claim 1, characterized in that, The step of using a causal discovery algorithm to model and analyze the causal relationship between the keypoint coordinates and the strabismus category to obtain a causal relationship model for determining the causal relationship includes: generating several candidate causal relationship models based on the keypoint coordinates; constructing a corresponding structural equation model for each candidate causal relationship model; performing parameter estimation and model fitting on each structural equation model; performing a conditional independence test based on the fitting results; and selecting models that pass the conditional independence test; scoring all models that pass the conditional independence test; and obtaining the causal structure diagram of the causal relationship model with the best score as the causal structure between the keypoint coordinates and the strabismus category, which is used to determine the causal relationship.

3. The strabismus detection method based on causal feature gating according to claim 2, characterized in that, The step of filtering the key point coordinates based on the causal relationship model to obtain a set of causal variables includes: determining the target node corresponding to the strabismus category from the causal structure diagram, and obtaining a set of variables that form a Markov blanket with the target node, wherein the key point coordinate variables in the set of variables are determined as causal variables related to the strabismus category.

4. The strabismus detection method based on causal feature gating according to claim 1, characterized in that, The step of extracting multiple causal variables from several samples in the training sample set based on the set of causal variables includes: randomly selecting one sample from the same group of samples in the training sample set as the original sample, and the remaining samples as augmented samples; extracting the causal variables from the original sample and the augmented samples respectively; constructing the causal variables into a one-dimensional feature array, wherein the one-dimensional feature array of the original sample is used as the main input of the classification model, and the one-dimensional feature array of the augmented sample is used as the auxiliary input of the classification model.

5. The strabismus detection method based on causal feature gating according to claim 1, characterized in that, The step of training the classification model using multiple causal variables while adaptively adjusting the gating parameters during the training process includes: forward propagating the multiple causal variables sequentially through linear transformations, causal feature gating modulation, and nonlinear activation functions at each layer of the classification model to obtain predicted values; calculating a loss function based on the difference between the predicted values ​​and the true labels, and constructing consistency constraints; calculating the gradient of the loss function with respect to the network parameters and the gating parameters in the classification model using a backpropagation algorithm, and jointly updating the network parameters and the gating parameters using an optimization algorithm.

6. The strabismus detection method based on causal feature gating according to claim 3, characterized in that, The Markov blanket variables of the target node include the parent node of the target node, the child nodes of the target node, and the parent node of the child nodes of the target node.

7. The strabismus detection method based on causal feature gating according to claim 1, characterized in that, The classification model is a multilayer perceptron model, which includes at least one hidden layer.

8. The strabismus detection method based on causal feature gating according to claim 2, characterized in that, The structural equation model is expressed by the following formula: in, Indicates the first Key point coordinate variables, Represents the coordinate variables of the key points The set of parent node variables, This represents the corresponding functional relationship. This represents mutually independent noise terms.

9. The strabismus detection method based on causal feature gating according to claim 2, characterized in that, include: The least squares method or the maximum likelihood estimation method are used to estimate the parameters and fit the model of the structural equation model.

10. A processing apparatus, characterized in that, include: One or more processors; A memory for storing one or more computer programs, and one or more processors for executing the one or more computer programs stored in the memory to cause the one or more processors to perform the strabismus detection method based on causal feature gating as described in any one of claims 1-9.

Images