A method and system for simulating visual conduction pathway impairment

Abstract

The present application relates to the field of medical education simulation training, more particularly, the present application relates to a visual conduction pathway injury simulation method and system, the method comprises: collecting the positioning data of students on the model containing multiple injury nodes, calculating the stability degree by comprehensively evaluating the data volatility and spatial weight, and screening out effective positioning data; dynamic clustering analysis is carried out on the effective data, the mis-touch mode is identified and mapped to the corresponding injury node; the misdiagnosis rate of each node is calculated based on the correct identification times and mis-touch mode information, and the teaching difficulties are identified and the training content is adaptively adjusted. The present application solves the problems of low interaction precision and rough evaluation of existing teaching tools through data reliability processing, mis-touch mode fine analysis and quantitative feedback closed loop, and significantly improves the data reliability, error diagnosis accuracy and personalized teaching effect of simulation training.

Description

Technical Field

[0001] This invention relates to the field of medical education simulation training. More specifically, this invention relates to a simulation method and system for visual conduction pathway damage. Background Technology

[0002] As a core component of neuroanatomy teaching, the visual pathway is complex in structure and exhibits diverse damage manifestations. Accurately understanding the correspondence between each node of this pathway and the corresponding visual field defects is crucial for the training of medical students and clinicians.

[0003] Currently, related teaching and training mainly rely on two-dimensional diagrams, static anatomical models, or simple electronic interactive tools. While these traditional methods can demonstrate basic structures, they generally suffer from problems such as susceptibility to environmental factors and coarse operational feedback when conducting high-precision interactive operation training. Students cannot obtain stable and accurate practical feedback, and teachers find it difficult to quantify and analyze students' operational errors and provide targeted guidance. This makes the training process prone to becoming a mere formality, failing to effectively establish solid spatial localization and clinical diagnostic skills.

[0004] Therefore, there is an urgent need to develop a simulation system that can achieve stable and accurate interaction, intelligent quantitative evaluation, and dynamic optimization of training, so as to improve the reliability and effectiveness of teaching and training. Summary of the Invention

[0005] To address the aforementioned technical problem of how to improve the reliability and effectiveness of teaching and training, this invention provides solutions in the following aspects.

[0006] In the first aspect, a simulation method for visual conduction path impairment includes:

[0007] During a simulation training operation, the positioning data of all students holding interactive detection devices on a visual transmission path model containing multiple damage nodes is collected, and the distance from the interactive detection device to each damage node is calculated based on this data to form a distance sequence of all distance combinations under each damage node.

[0008] Calculate the stability of the positioning data at all times from the start of the operation to the current time; remove the positioning data at abnormal times based on the set stability threshold to obtain valid positioning data;

[0009] Cluster all valid location data to find the spatial distribution pattern of repeated accidental touches during student operation to obtain multiple clusters. Map each cluster to the damage node that is closest to its geometric center and establish the association between the clusters and the damage nodes.

[0010] For each damaged node, the misclassification rate of the damaged node is calculated based on the number of times it is correctly identified and all the clusters associated with it.

[0011] Based on the comparison between the misclassification rate and the preset threshold, the damaged nodes with a misclassification rate exceeding the threshold are output and reinforced training is performed on them.

[0012] Preferably, the stability is the average of the contribution values ​​corresponding to all damaged nodes; wherein, the contribution value corresponding to a single damaged node is the product of the first factor and the second factor;

[0013] The first factor is an exponential function with the natural constant as the base, and the exponent of this exponential function is the negative of the standard deviation of the loss value sequence from the start of the operation to the current time of the damage node.

[0014] The second factor is the ratio of the distance from the interactive detection device to the damaged node at the current moment to the sum of the distances from the interactive detection device to all damaged nodes from the start of the operation to the current moment.

[0015] Preferably, obtaining the loss value sequence includes:

[0016] The distance sequence of each damaged node is fitted to obtain the fitting function;

[0017] Calculate the distance from each element in the distance sequence to the fitted function; this distance is the loss value. Combine all the loss values ​​to construct the loss value sequence for each damaged node.

[0018] Preferably, the clustering uses the K-means algorithm.

[0019] Preferably, the clustering process further includes:

[0020] The number of clusters k is set to increase from 1. The evaluation score of the clustering result corresponding to each k value is calculated. Then, the evaluation score sequence is constructed according to the increasing order. When the evaluation score sequence first shows a decreasing inflection point, the increment is stopped, and all clusters in the clustering result corresponding to the k value before the inflection point are taken as the optimal clusters.

[0021] Preferably, the acquisition of the evaluation score includes:

[0022] Obtain all clusters for any k value. For any cluster, calculate the ratio of the mean to the maximum distance of all valid localization data within the cluster to the damaged node associated with that cluster. Use this ratio as the first ratio. Calculate the mean of the first ratios for all clusters.

[0023] Calculate the hyperbolic tangent function value for the current k value, and use the product of the hyperbolic tangent function value for the current k value and the difference between 1 and the mean of the first ratios of all clusters corresponding to all k values ​​as the evaluation score of the clustering result corresponding to the current k value.

[0024] Preferably, the misclassification rate is 1 minus the product of the calculated accuracy component and the dispersion component; wherein, the accuracy component is obtained by calculating the number of times the damaged node is correctly identified and dividing it by the sum of the number of times it is correctly identified and the total number of valid localization data in all clusters associated with the damaged node; and the dispersion component is obtained by calculating the mean of the cosine similarity between the direction vectors of any two clusters in all clusters associated with the damaged node.

[0025] The direction vector is composed of the vectors from all valid localization data in all clusters associated with the damaged node to the damaged node.

[0026] Preferably, the number of times the device is correctly identified includes:

[0027] In a simulation operation, after the student finishes the operation, the spatial coordinates of the final operation point are determined based on the student's valid positioning data. The Euclidean distance between the spatial coordinates and the coordinates of the damaged node is calculated. If the Euclidean distance is less than the preset correct judgment threshold, it is judged as a correct identification, and the correct identification counter of the damaged node is incremented.

[0028] Secondly, a simulation system for visual transmission path impairment that implements the above method is provided, comprising:

[0029] The positioning and acquisition module is used to collect positioning data of the student's handheld interactive detection device on the visual transmission path model;

[0030] The data processing module is communicatively connected to the positioning and acquisition module, and is used to receive the positioning data, execute the steps of the method, and output damage node information and training adjustment instructions when the misjudgment rate exceeds a preset threshold.

[0031] The model interaction module is communicatively connected to the data processing module and is used to provide the visual transmission path model to students and receive operations from the interactive detection device.

[0032] The feedback control module is communicatively connected to the data processing module and the model interaction module. It is used to generate prompts based on the information of damaged nodes whose misjudgment rate exceeds a preset threshold, and to control the model interaction module to adjust the training content for the corresponding damaged nodes according to the training adjustment instructions.

[0033] The beneficial effects of this invention are:

[0034] This invention introduces a stability evaluation model based on data volatility and spatial cumulative weights, enabling reliable screening of raw positioning data. This effectively resists interference in the teaching environment, ensuring the accuracy of subsequent data sources and significantly improving the reliability of simulation training data. Furthermore, through a dynamic clustering algorithm for effective data, the system can automatically identify and map recurring spatial mis-touch patterns in student operations, achieving refined attribution and classification of errors. This allows for a deeper understanding of error diagnosis, moving beyond simply identifying whether an error occurred to determining the specific type of error, greatly enhancing the accuracy of error diagnosis. Finally, by integrating a misjudgment rate quantification model that combines accuracy and mis-touch pattern consistency, the system can intelligently identify teaching difficulties and drive adaptive adjustments to training content accordingly. This forms a quantitative feedback loop of "assessment-diagnosis-intervention," achieving truly personalized teaching and effectively improving training efficiency and learning outcomes. Attached Figure Description

[0035] Figure 1 This is a flowchart of steps S1-S4 in a simulation method for visual transmission path damage according to an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram of the structure of a simulation system for visual transmission path damage according to an embodiment of the present invention. Detailed Implementation

[0037] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0038] The application scenario of this invention is simulation training for visual pathway damage recognition in medical education. Students use standardized interactive detection devices (such as high-precision positioning probes) to operate on a standard physical model of the visual pathway. In each training session, students, following instructions, sequentially and coherently traverse and click on multiple or all damage nodes set in the model. Upon locating a damage node and deeming the location accurate, a confirmation signal is triggered (e.g., pressing a button). The system collects and processes the spatial positioning data at the final confirmation moment. Operational data from multiple students can be aggregated and analyzed within a unified coordinate system.

[0039] Reference Figure 1 A simulation method for visual transmission path impairment includes steps S1-S4, as detailed below:

[0040] S1: During a simulation training operation, the positioning data of all students' handheld interactive detection devices on the visual transmission path model containing multiple damage nodes is collected, and the distance from the interactive detection device to each damage node is calculated based on this data to form a distance sequence of all distance combinations under each damage node.

[0041] Accurate assessment of student actions requires obtaining the precise location of their handheld device in model space. However, common factors in the teaching environment, such as electromagnetic interference, can cause momentary jumps or deviations in the positioning signal. Therefore, the system needs to construct a reliable metric based on the original positioning data that reflects the real-time spatial relationship between the device and each preset anatomical location (i.e., the lesion node).

[0042] In one embodiment, high-precision positioning base stations (such as UWB or Bluetooth AOA) are deployed at or around each damaged node in the simulation model to form a positioning network. Interactive detection devices held by students act as mobile tags, transmitting signals.

[0043] During a simulation training operation, when a student completes positioning and triggers a confirmation signal, this moment is marked as the valid sampling moment for this operation. The positioning base station immediately performs high-precision sampling to obtain the three-dimensional coordinates of the tip of the interactive detection device in the unified coordinate system of the model. The three-dimensional coordinates of each damage node have been pre-calibrated and stored in the system.

[0044] For each damaged node, the Euclidean distance from the interactive detection device to the damaged node is calculated. As different students click on the damaged node, a distance sequence obtained from multiple students' clicks will be accumulated for any specific damaged node.

[0045] The above operations transform the student's spatial operations into a series of quantitative distance observations associated with each damage node, providing a foundational data stream for subsequent analysis.

[0046] S2: Calculate the stability of the positioning data at all times from the start of the operation to the current time; based on the set stability threshold, remove the positioning data at abnormal times to obtain valid positioning data.

[0047] Common factors in the teaching environment, such as electromagnetic interference, equipment signal obstruction, and hand tremors, can cause momentary anomalies in positioning data, known as "outliers." Directly using this data for evaluation can distort students' actual operational performance at damaged points and even mislead the identification of teaching difficulties. Therefore, it is essential to dynamically evaluate the reliability of data at each sampling moment and eliminate low-quality data.

[0048] In one embodiment, for each damaged node, a curve fitting is performed using the distance sequence acquired according to the S1 operation described above to obtain a fitting function. Then, the vertical distance from the current distance to the fitted curve is calculated, and this vertical distance is defined as the loss value of the node at the current time. The larger the loss value, the further the current distance deviates from the historical normal trend. All the above-mentioned loss values ​​of the damaged node from the start of the operation to the current time constitute its loss value sequence.

[0049] For any node at any given time, calculate its contribution value. The specific calculation process is as follows:

[0050] First, calculate the standard deviation of the loss value sequence of a single damage node from the start of the operation to the current time, and convert the standard deviation into the form of an exponential function to obtain the first factor.

[0051] For example, the first factor mentioned above can be expressed as a relational expression:

[0052]

[0053] In the formula, For the first factor mentioned above, Let be the standard deviation of the loss value sequence of a single damaged node from the start of the operation to the current time. It is an exponential function with the natural constant e as its base.

[0054] The larger the standard deviation, the less stable the distance observation history of the damaged node is. Therefore, the reliability of the relevant data of the damaged node at the current moment should be suppressed, that is, the smaller the first factor.

[0055] Then, the second factor is obtained by calculating the ratio of the distance from the interactive detection device to the damaged node at the current moment to the sum of the distances from the interactive detection device to all damaged nodes from the start of the operation to the current moment.

[0056] The second factor mentioned above assesses the relative magnitude of the current distance in history.

[0057] Finally, the product of the first and second factors calculated above is taken as the contribution value corresponding to the damaged node at the current moment. Similarly, the contribution values ​​corresponding to each damaged node at all moments from the start of the operation to the current moment can be obtained, and then the average of all contribution values ​​is calculated to obtain the stability of the localization data at each moment.

[0058] Furthermore, a stability threshold, such as 0.6, can be set. For each moment, the positioning data corresponding to the moment with a stability level greater than or equal to the stability threshold is retained and marked as valid positioning data. Moments with a stability level less than the stability threshold are judged as abnormal moments, and their corresponding positioning data needs to be removed.

[0059] In summary, by fusing historical volatility and distance relativity of damaged nodes in a dual assessment, dynamic cleaning of the original positioning data can be achieved. This not only filters out anomalies caused by sudden interference but also identifies "pseudo-precise points," thereby selecting highly reliable and effective positioning data.

[0060] S3: Cluster all valid positioning data to find the spatial distribution pattern of repeated accidental touches during student operation to obtain multiple clusters. Map each cluster to the damage node that is closest to its geometric center and establish the association between the clusters and the damage nodes.

[0061] Students' errors during hands-on activities are not entirely random. Due to the proximity of damaged nodes, morphological similarity, or blind spots in teaching cognition, students often repeatedly touch certain specific areas. These areas form spatial clusters. Traditional methods only record the number of errors, ignoring the spatial distribution patterns of errors, and therefore cannot diagnose the specific type of error (e.g., whether it is confusing the damaged node "optic nerve" with the damaged node "optic chiasm," or a blurred understanding of the boundaries of the damaged node "optic tract"). Identifying these spatial patterns is crucial for achieving precise teaching intervention.

[0062] In one embodiment, spatial clustering analysis is performed on all the filtered valid positioning data (i.e., touch points that are determined to be reliable but not necessarily correct) in S2 above, in order to discover recurring false touch patterns.

[0063] First, K-means clustering is performed on all valid location data, but the value of k is not fixed. That is, starting from k=1, the value of k is gradually increased to make multiple clustering attempts.

[0064] For each k value, the quality of the clustering results needs to be evaluated, that is, the evaluation score of the clustering results under each k value needs to be calculated.

[0065] The calculation process for the above evaluation scores is generally as follows:

[0066] First, calculate the geometric center of all clusters under a single k value. Then, calculate the Euclidean distance from this geometric center to each damaged node, and associate each cluster with the nearest damaged node. This association means that the accidental touch within the cluster was most likely to have occurred when attempting to click on that damaged node. Furthermore, map each cluster to the damaged node closest to its geometric center, establishing the association between clusters and damaged nodes.

[0067] Next, for any cluster under a single k value, the ratio of the mean to the maximum distance of all valid location data within the cluster to the associated damaged node is calculated as the first ratio. The larger the first ratio, the denser and more concentrated the points within the cluster, indicating a clear mis-touch habit pattern. The mean of the first ratios of all clusters is calculated to measure the clarity of the overall pattern of the clusters under a single k value.

[0068] Then, it is also necessary to calculate the hyperbolic tangent function value for the current k value, that is... The hyperbolic tangent function value approaches 1 as the value of k increases, in order to balance the tendency for the quantification of clusters to naturally lead to more compact individual clusters, and to prevent the algorithm from blindly pursuing smaller k values.

[0069] Finally, the evaluation score for the clustering result corresponding to the current k-value is obtained by multiplying the hyperbolic tangent function value of the current k-value by the difference between 1 and the mean of the first ratios of all clusters corresponding to all k-values.

[0070] Furthermore, according to the above operation, the value of k starts from 1 and increments, and the evaluation score of the clustering result corresponding to each k value is calculated. Then, the evaluation score sequence is constructed according to the incrementing order. When the evaluation score sequence first shows a decreasing inflection point, the increment stops, and all the clusters in the clustering result corresponding to the k value before the inflection point are taken as the optimal clusters.

[0071] Then, based on the above-mentioned optimal cluster, the association between the optimal cluster and the damaged node can be obtained.

[0072] Through the clustering described above, patterns that spatially cluster and recur are automatically identified and separated from all valid location data (i.e., touch points judged as reliable but not necessarily correct). These patterns are interpreted as students' systematic accidental touch habits. Clicks that do not form clusters, which may be accidental deviations or completely correct clicks, are not assigned to any cluster.

[0073] S4: For each damaged node, calculate the misclassification rate of the damaged node based on the number of times it is correctly identified and all the clusters associated with it. Based on the comparison between the misclassification rate and the preset threshold, output the damaged nodes whose misclassification rate exceeds the threshold and perform reinforcement training on them.

[0074] Simply having correlations is insufficient to guide instruction. A comprehensive indicator is needed that reflects both students' overall mastery of a particular point of weakness (correct / incorrect frequency) and reveals the nature of their error patterns (whether it is a systematic shift or random occurrence). Only such a composite indicator can distinguish between "persistent errors caused by conceptual confusion" and "random errors caused by accidental mistakes," thereby driving instructional interventions of varying intensities.

[0075] In one embodiment, by calculating a misclassification rate for each damaged node, this metric integrates information on both correct operation and error pattern. It can reflect the overall accuracy rate for a damaged node and reveal the characteristics of its error pattern, thereby distinguishing between "knowledge confusion" and "accidental mistakes" and driving differentiated teaching interventions.

[0076] The calculation process for the above false positive rate is generally as follows:

[0077] First, in a single simulation operation, after the student completes their operation, the 3D coordinates of the final operation point are determined based on their valid positioning data. The Euclidean distance between the 3D coordinates of the final operation point and the 3D coordinates of the damaged node is calculated. If this Euclidean distance is less than a preset correct judgment threshold (e.g., 2mm), it is considered a correct identification, and the correct identification counter for the damaged node is incremented. For example, if a damaged node, such as the optic chiasm, is clicked a total of 200 times, after calculating the Euclidean distance and comparing it with the correct judgment threshold, the number of times it was correctly clicked is determined to be 150.

[0078] Then, based on the above operations, the number of times a single damaged node was correctly identified is divided by the sum of that number of correct identifications and the total number of valid localization data in all clusters associated with that damaged node, to obtain the accuracy component. For example, for the optic chiasm, it was clicked 200 times, of which 150 times were correct clicks. The associated clusters 1 and 2 contain a total of 45 valid localization data (in fact, the valid localization data contained in the clusters formed after clustering represent a series of false clicks), so the total number is 195.

[0079] Next, the mean of the cosine similarity between the direction vectors of any two clusters in all clusters associated with the damaged node is calculated to obtain the dispersion component. The higher the dispersion component, the more consistent the spatial offset direction of different accidental touch patterns (e.g., they all lean to the same side), indicating that the error is systematic. If the accidental touch patterns are very dispersed, the lower the dispersion component, indicating that the error is more likely to be random.

[0080] The direction vector is composed of the vectors from all valid localization data in all clusters associated with the damaged node to the damaged node.

[0081] Finally, the calculated accuracy component is multiplied by the dispersion component, and the difference between 1 and the product is taken as the aforementioned misclassification rate.

[0082] Multiplying these two components means that the final assessment result is sensitive to both. A low misjudgment rate (i.e., high mastery) must simultaneously satisfy "high accuracy" and "high error consistency." This design can effectively filter out damage points that, although they have been wrong a few times, are made without any pattern (possibly accidental mistakes), thereby accurately identifying the real teaching difficulties that are not only frequently wrong but also make mistakes in a pattern (reflecting knowledge blind spots).

[0083] The threshold for the false positive rate is set at 0.2. For any damaged node, if its false positive rate is greater than this threshold, it indicates that the damaged node is a teaching difficulty and the reinforcement training mechanism is automatically triggered: in subsequent training, the frequency of the occurrence of the damaged node is increased, targeted prompts are provided or comparative training is conducted until its false positive rate drops below the threshold.

[0084] This invention also provides a simulation system for visual conduction path impairment. For example... Figure 2 As shown, the system includes:

[0085] Location acquisition module: Used to collect raw location data of students' handheld interactive detection devices on the visual transmission path model.

[0086] Data processing module: As the core of the system, it is connected to the positioning and acquisition module. It receives raw data, executes all the steps of the aforementioned method (S1-S4), calculates the misclassification rate of each damaged node, identifies the difficult points (damaged nodes) whose misclassification rate exceeds the preset threshold, and generates corresponding training adjustment instructions.

[0087] Model Interaction Module: This module presents the visual transmission path model to students and receives their input. It also connects to the data processing module and the feedback control module to execute specific interactions and displays.

[0088] Feedback and Adjustment Module: Connects to the data processing module and the model interaction module. Based on the difficulty node information and adjustment instructions output by the data processing module, it generates teaching prompts and controls the model interaction module to adaptively adjust subsequent training content (such as increasing the training frequency of difficult points).

[0089] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0090] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A simulation method for visual transmission path impairment, characterized in that, include: During a simulation training operation, the positioning data of all students holding interactive detection devices on a visual transmission path model containing multiple damage nodes is collected, and the distance from the interactive detection device to each damage node is calculated based on this data to form a distance sequence of all distance combinations under each damage node. The stability of the positioning data at all times from the start of the calculation operation to the current time is calculated. The stability is the average of the contribution values ​​corresponding to all damaged nodes. The contribution value corresponding to a single damaged node is the product of the first factor and the second factor. The first factor is an exponential function with the natural constant as the base, and the exponent of this exponential function is the negative of the standard deviation of the loss value sequence from the start of the operation to the current time of the damage node. The second factor is the ratio of the distance from the interactive detection device to the damaged node at the current moment to the sum of the distances from the interactive detection device to all damaged nodes from the start of the operation to the current moment; the acquisition of the loss value sequence includes: The distance sequence of each damaged node is fitted to obtain the fitting function; Calculate the distance from each element in the distance sequence to the fitted function; this distance is the loss value. Combine all the loss values ​​to construct the loss value sequence for each damaged node. Remove the location data at abnormal times based on the set stability threshold to obtain effective location data. Cluster all valid location data to find the spatial distribution pattern of repeated accidental touches during student operation to obtain multiple clusters. Map each cluster to the damage node that is closest to its geometric center and establish the association between the clusters and the damage nodes. For each damaged node, the misclassification rate of the damaged node is calculated based on the number of times it is correctly identified and all clusters associated with it. The misclassification rate is 1 minus the product of the calculated accuracy component and the dispersion component. Specifically, the accuracy component is obtained by dividing the number of times the damaged node is correctly identified by the sum of the number of times it is correctly identified and the total number of valid localization data in all clusters associated with the damaged node. The dispersion component is obtained by calculating the mean of the cosine similarity between the direction vectors of any two clusters in all clusters associated with the damaged node. The direction vector is composed of vectors from all valid localization data in all clusters associated with the damaged node to the damaged node. Based on the comparison between the misclassification rate and the preset threshold, the damaged nodes with a misclassification rate exceeding the threshold are output and reinforced training is performed on them.

2. The simulation method for visual transmission path impairment according to claim 1, characterized in that, The clustering was performed using the K-means algorithm.

3. The simulation method for visual transmission path impairment according to claim 1, characterized in that, The clustering process also includes: The number of clusters k is set to increase from 1. The evaluation score of the clustering result corresponding to each k value is calculated. Then, the evaluation score sequence is constructed according to the increasing order. When the evaluation score sequence first shows a decreasing inflection point, the increment is stopped, and all clusters in the clustering result corresponding to the k value before the inflection point are taken as the optimal clusters.

4. The simulation method for visual transmission path impairment according to claim 3, characterized in that, The evaluation score is obtained by: Obtain all clusters for any k value. For any cluster, calculate the ratio of the mean to the maximum distance of all valid localization data within the cluster to the damaged node associated with that cluster. Use this ratio as the first ratio. Calculate the mean of the first ratios for all clusters. Calculate the hyperbolic tangent function value for the current k value, and use the product of the hyperbolic tangent function value for the current k value and the difference between 1 and the mean of the first ratios of all clusters corresponding to all k values ​​as the evaluation score of the clustering result corresponding to the current k value.

5. The simulation method for visual transmission path impairment according to claim 1, characterized in that, The number of times it was correctly identified includes: In a simulation operation, after the student finishes the operation, the spatial coordinates of the final operation point are determined based on the student's valid positioning data. The Euclidean distance between the spatial coordinates and the coordinates of the damaged node is calculated. If the Euclidean distance is less than the preset correct judgment threshold, it is judged as a correct identification, and the correct identification counter of the damaged node is incremented.

6. A simulation system for visual conduction path impairment, used to implement the simulation method for visual conduction path impairment as described in any one of claims 1-5, characterized in that, include: The positioning and acquisition module is used to collect positioning data of the student's handheld interactive detection device on the visual transmission path model; The data processing module is communicatively connected to the positioning and acquisition module, and is used to receive the positioning data, execute the steps of the method, and output damage node information and training adjustment instructions when the misjudgment rate exceeds a preset threshold. The model interaction module is communicatively connected to the data processing module and is used to provide the visual transmission path model to students and receive operations from the interactive detection device. The feedback control module is communicatively connected to the data processing module and the model interaction module. It is used to generate prompts based on the information of damaged nodes whose misjudgment rate exceeds a preset threshold, and to control the model interaction module to adjust the training content for the corresponding damaged nodes according to the training adjustment instructions.

Images