System for detecting mental diseases through eye movement based on artificial intelligence technology

Abstract

The invention provides an artificial intelligence technology-based eye movement mental disease detection system, which comprises an image acquisition unit, and is characterized in that the image acquisition unit comprises a camera device, an image processing module, a control module and a data storage module; the camera equipment is used for collecting eye actions of a to-be-tested person, including eyeball actions and eyeball actions, and collecting head actions of the to-be-tested person at the same time; the image processing module is used for judging whether the head of the to-be-detected person moves or not according to the eye actions and the head actions collected by the camera equipment; judging whether the head shakes or not by analyzing the head actions, if so, judging whether the eye actions are influenced by the head shaking or not, filtering eye pictures influenced by the head shaking, and retaining the eye pictures during the head rest period; and carrying out image analysis on the retained eye picture in the head static period, carrying out image analysis on the eye action subareas where the eye actions are located, and outputting an eye fundus image, eyeball actions and eyeball actions.

Description

Technical Field

[0001] The present invention belongs to the field of disease detection, and in particular relates to a system for detecting mental illness by eye movement based on artificial intelligence technology. Background Art

[0002] As an emerging biometric identification method, eye movement detection technology has gradually attracted attention in the field of mental health in recent years. Traditional diagnosis of mental illness mainly relies on clinical interviews and self-reports, which is not only time-consuming and labor-intensive, but also affected by a variety of subjective factors. With the development of artificial intelligence technology, eye movement detection is considered to be an objective and real-time evaluation method.

[0003] By analyzing an individual's eye movement patterns during specific tasks, such as fixation time and eye saccade frequency, researchers can detect potential mental health problems. The basis of this technology is the advancement of eye tracking equipment, combined with the application of machine learning algorithms, which makes the analysis of eye movement data more accurate and efficient. However, although eye movement detection technology has shown potential in screening for mental illness, there are still some obvious shortcomings. First of all, the accuracy and reliability of eye movement detection systems remain an issue that needs to be addressed. The eye movement characteristics of different individuals vary, which may lead to misdiagnosis or missed diagnosis.

[0004] Environmental factors and the sensitivity of the equipment can also affect the detection results. For example, the accuracy of eye tracking may decrease in noisy or light-changing environments. Existing eye movement detection systems usually require high technical barriers and complex operations, especially for medical workers. This limits their promotion in clinical applications, especially in resource-poor areas. Data privacy and ethical issues are also important challenges facing eye movement detection technology. The collection and storage of personal eye movement data requires compliance with strict privacy protection standards, but in actual operations, how to balance the relationship between data utilization and personal privacy protection remains a difficult problem.

[0005] Eye movement detection technology has not yet formed unified standards and specifications. Different research institutions or companies use different equipment and algorithms, which makes it difficult to compare and verify the results. This lack of standardization has resulted in unsatisfactory clinical application results for eye movement detection technology and has limited its widespread promotion in the field of mental health. Therefore, although eye movement detection technology has shown certain application prospects in the early screening and assessment of mental illness, in order to overcome current technical defects, in-depth research and exploration are still needed in algorithm optimization, equipment improvement, user training, and ethical standards. Through continuous technological innovation and standardization construction, eye movement detection is expected to play a more important role in future mental health assessments. Summary of the invention

[0006] The present invention proposes a system for detecting mental illness by eye movement based on artificial intelligence technology. Traditional methods of diagnosing mental illness often rely on subjective evaluation, which is easily affected by the doctor's personal experience and the patient's self-report, resulting in inaccurate diagnosis results. The system uses an image acquisition unit and a camera to monitor the eye and head movements of the person to be tested in real time, which can objectively record eye movement data and reduce the interference of subjective factors, thereby improving the accuracy and reliability of diagnosis.

[0007] The technical solution of the present invention is implemented as follows: a system for detecting mental illness by eye movement based on artificial intelligence technology, comprising an image acquisition unit, the image acquisition unit comprising a camera, an image processing module, a control module and a data storage module; the camera is used to acquire the eye movements of the person to be tested, including eye movements and eyeball movements, and at the same time acquire the head movements of the person to be tested; the image processing module determines whether the head of the person to be tested moves through the eye movements and head movements acquired by the camera; determines whether there is head shaking by analyzing the head movements, and if there is head shaking, determines whether the eye movements are affected by the head shaking, and filters the eye pictures affected by the head shaking, and retains the eye pictures during the period when the head is still; performs image analysis on the retained eye pictures during the period when the head is still, performs image analysis on the eye movement partitions where the eye movements are located, and outputs fundus images, eyeball movements and eyeball movements; The control module encodes the eye movement partition where the eye movement is located, and outputs the code to the data storage module; the data storage module is used to store the eye movement partition code, fundus image, eyeball movement, eyeball movement and eye movement sample library; the eyeball movement includes eye blinking movement; the eyeball movement includes pupil deflection movement, pupil expansion and contraction movement, and pupil static movement; the eye movement includes pupil deflection, pupil expansion, pupil contraction, eye blinking, pupil size unchanged in a natural state, and eyelid movement; the eye movement sample library includes eye movements of sample patients with mental illness and eye movements of healthy people; the control module uploads the encoded eye movement partition to the data analysis unit; The data analysis unit includes a code reading module, a classification module, a matching module, a first data analysis module and a second data analysis module; the code reading module is used to read the eye movement partition code; the classification module obtains the eye movement partition code of the sample to be analyzed, and classifies the eye movement data of the sample to be analyzed; the matching module matches the eye movement samples of the sample to be analyzed with all eye movements in the eye movement sample library one by one; the first data analysis module is used to match and analyze the eye movement data of the sample to be analyzed with similar eye movements in the eye movement sample library, and output the analysis results; the second data analysis module compares and analyzes the eye movement of the sample to be analyzed with normal eye movements and eye movements of different types of mental patients according to the eye movement characteristics, and outputs the analysis results; the data analysis unit transmits the comparative analysis results.

[0008] Compared with existing eye movement detection technology, this system has obvious advantages in many aspects. First, traditional eye movement detection technology often only focuses on the movement of the eyeballs, while ignoring the influence of head movements. This system can more comprehensively understand the eye movement characteristics of the person being tested by comprehensively analyzing the movement of the eyes and the head. This multi-dimensional analysis method makes the diagnosis results more accurate and can effectively reduce errors caused by factors such as head shaking.

[0009] The existing technology is relatively simple in data storage and processing, and most systems cannot effectively integrate a large amount of eye movement data for in-depth analysis. However, this system uses a data storage module to systematically store information such as eye movement partition coding, fundus images, eye movements, and eyeball movements to form a complete eye movement sample library. This innovation makes subsequent data analysis more efficient, supports the application of deep learning algorithms, and provides more powerful data support for the diagnosis of mental illness. The impact of head movement on eye movement data is a common problem. Traditional methods are difficult to effectively distinguish eye movement changes caused by head shaking when recording eye movements, which can easily lead to erroneous analysis results. The system uses an image processing module to determine in real time whether the head of the person to be tested has moved, and can filter out the eye images affected by head shaking to ensure the accuracy of the analysis results. This real-time monitoring and filtering function of head movements significantly improves the effectiveness of eye movement data analysis.

[0010] The design of the system includes a comprehensive eye movement sample library, covering eye movement data of patients with mental illness and healthy people. This feature provides rich reference data for subsequent matching and classification analysis, which can help identify potential characteristics of mental illness and achieve early diagnosis and intervention. The establishment of this sample library solves the problem of lack of sufficient control data in the past. The existing technology lacks unified standards in the output and interpretation of analysis results, which often makes it difficult to compare the results of different studies. The system can conduct systematic comparative analysis of the samples to be analyzed and provide clear analysis results through the classification, matching and analysis modules of the data analysis unit. This structured analysis method improves the operability and readability of the data and provides more intuitive decision support for clinicians.

[0011] Existing technologies are deficient in real-time and convenience, and many systems require a long time for data processing and result analysis. However, this system can achieve real-time monitoring and rapid feedback through efficient image processing and data analysis modules, greatly improving work efficiency. This high efficiency makes clinical applications more flexible and can meet rapidly changing medical needs.

[0012] As a preferred embodiment, when performing image analysis on the eye movement partition where the eye movement is located, the collected image is filtered, the fundus movement part is identified and analyzed, and the behavior to be tested is compared and analyzed with the sample images in the eye movement sample library.

[0013] As a preferred implementation, the eye movement sample library includes an eye movement sample library of healthy persons and an eye movement sample library of various mental illnesses.

[0014] As a preferred implementation, the eye movement sample library contains basic data on the movement trajectories of blinking, nystagmus, nystagmus, strabismus, horizontal gaze, oblique gaze, eye droop, eye twitch, eye twitch, proptosis, ectropion and entropion, and simultaneously defines the time and frequency information of the eye movements; different movements of the eyes at different times, the duration, number and strength of the movements are recorded using different data respectively.

[0015] As a preferred embodiment, the data analysis unit simultaneously performs analysis modeling when performing analysis, and constructs pupil movement states under different eye curve states through models for regional analysis, wherein the curves include eyeball curves and pupil curves, and the eyeball and pupil are respectively divided into different regions through regional analysis, and the eyeball and pupil are divided into regions, and the positions of eye movement features are recorded, and the eye movement features are stored, wherein the eye movement features include eye curve features and pupil features, and by matching the eye action sample library, when the user has abnormal points, the eye or pupil state is analyzed by regional analysis, and different regions are divided for the eyeball area and the pupil area, and the eyeball area and the pupil area are analyzed, and in the eyeball area, one of the curves is selected, and the position information of the eyeball area is analyzed with the curve, and if the analysis shows that the eyeball area matches the curve successfully, the eyeball state is abnormal at this time; if the analysis shows that the eyeball area does not match the curve successfully, the eyeball state is calibrated, and the calibration is abnormal, and if the pupil area matches the curve successfully, the pupil state is calibrated, and the pupil state is abnormal at this time.

[0016] As a preferred embodiment, the first data analysis module includes a matching module, an analysis module and a correction module; the matching module is used to match the eye movement of the sample to be analyzed with the same type of eye movement samples; the analysis module is used to obtain the matching degree between the sample to be tested and the same type of eye movement samples, and output data A according to the matching degree; if the data A is less than the threshold value A, the correction module processes and corrects the eye movement area coding of the sample to be tested, and stores the corrected data, repeating multiple times until the data A is greater than the threshold value A; if the data A is greater than the threshold value A, the corrected eye movement of the sample to be tested is matched and stored with the same type of eye movement samples in the eye sample action library.

[0017] As a preferred embodiment, the second data analysis module includes a normal eye movement analysis module and a mental illness eye movement analysis module; the normal eye movement analysis module is used to compare and analyze the eye movements of the samples to be analyzed with the eye movement sample library of healthy persons, and output the comparison results; the mental illness eye movement analysis module compares and analyzes the eye movements of the samples to be analyzed with the eye movement sample library of various mental illnesses, and outputs the results.

[0018] After adopting the above technical scheme, the beneficial effect of the present invention is that the introduction of the system has brought significant positive effects on the early screening and diagnosis of mental illness. Objective data collection and analysis methods make the diagnosis of mental illness more scientific and reliable. Through the quantitative analysis of eye movement characteristics, doctors can more accurately identify potential mental health problems, thereby achieving early intervention and reducing the risk of worsening of the disease. The multi-dimensional analysis capability of the system makes the identification of different types of mental illness more accurate. There may be obvious differences in the eye movement characteristics of patients with different mental illnesses. By comparing and analyzing with healthy people and other mental patients in the sample library, doctors can obtain more detailed diagnostic basis. This differentiated analysis method is conducive to the formulation of personalized treatment plans and improves the pertinence and effectiveness of treatment. The efficiency and convenience of the system make clinical applications more popular. Traditional diagnosis of mental illness often takes a lot of time and manpower, while this system can achieve fast real-time analysis, saving resources for medical institutions and improving diagnostic efficiency. This feature can not only improve the patient's medical experience, but also bring positive effects to the operation and management of medical institutions. The establishment and continuous improvement of the eye movement sample library has promoted the progress of mental illness research. By accumulating a large amount of eye movement data from different types of patients, researchers can deeply analyze the relationship between eye movement characteristics and mental illness, providing rich basic data for subsequent research. This long-term impact will promote technological progress and application development in the entire field of mental health. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0020] Figure 1 It is a flow chart of the system of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0022] like Figure 1As shown, a system for detecting mental illness by eye movement based on artificial intelligence technology includes an image acquisition unit, and the image acquisition unit includes a camera, an image processing module, a control module and a data storage module; the camera is used to acquire the eye movements of the person to be tested, including eye movements and eyeball movements, and at the same time acquire the head movements of the person to be tested; the image processing module determines whether the head of the person to be tested moves through the eye movements and head movements acquired by the camera; determines whether there is head shaking by analyzing the head movements, and if there is head shaking, determines whether the eye movement is affected by the head shaking, and filters the eye pictures affected by the head shaking, and retains the eye pictures during the period when the head is still; performs image analysis on the eye pictures during the period when the head is still, performs image analysis on the eye movement partitions where the eye movements are located, and outputs fundus images, eyeball movements and eyeball movements; The control module encodes the eye movement partition where the eye movement is located, and outputs the code to the data storage module; the data storage module is used to store the eye movement partition code, fundus image, eyeball movement, eyeball movement and eye movement sample library; the eyeball movement includes eye blinking movement; the eyeball movement includes pupil deflection movement, pupil expansion and contraction movement, and pupil static movement; the eye movement includes pupil deflection, pupil expansion, pupil contraction, eye blinking, pupil size unchanged in a natural state, and eyelid movement; the eye movement sample library includes eye movements of sample patients with mental illness and eye movements of healthy people; the control module uploads the encoded eye movement partition to the data analysis unit; The data analysis unit includes a code reading module, a classification module, a matching module, a first data analysis module and a second data analysis module; the code reading module is used to read the eye movement partition code; the classification module obtains the eye movement partition code of the sample to be analyzed, and classifies the eye movement data of the sample to be analyzed; the matching module matches the eye movement samples of the sample to be analyzed with all eye movements in the eye movement sample library one by one; the first data analysis module is used to match and analyze the eye movement data of the sample to be analyzed with similar eye movements in the eye movement sample library, and output the analysis results; the second data analysis module compares and analyzes the eye movement of the sample to be analyzed with normal eye movements and eye movements of different types of mental patients according to the eye movement characteristics, and outputs the analysis results; the data analysis unit transmits the comparative analysis results.

[0023] This technology is based on artificial intelligence technology and aims to assist in the diagnosis of mental illness through eye movement detection. Its working principle and workflow can be divided into several main stages: image acquisition, data storage, data analysis and result output. The following is a detailed description of the specific operation steps of each stage.

[0024] Preparation of camera equipment: First, the system needs to be equipped with high-performance camera equipment that can capture the eye and head movements of the person to be tested at a high frequency. The equipment should have high definition and good light adaptability to ensure the quality of the collected images. Image acquisition: After the person to be tested sits in front of the system, the camera equipment begins to collect the movements of his eyes and head in real time. The system uses built-in algorithms to detect various movements of the eyeballs and eyeballs, including blinking, deflection, expansion, and contraction of the pupils. At the same time, any movement of the head will also be recorded. Head movement judgment: The image processing module will analyze the collected images in real time to determine whether the head of the person to be tested is moving. If head shaking is detected, the system will mark the eye images during these periods to avoid eye movement data errors caused by head movement. Filter and retain data: The system will automatically filter out the eye images collected during the period of head movement and only retain the eye movement data during the period when the head is still. This process ensures the accuracy of subsequent analysis.

[0025] Eye movement partition coding: The control module partitions and codes the eye movements during the period of head stillness, with the purpose of storing different types of eye movement information in a structured manner for subsequent analysis. Data storage: All coded information, fundus images, eye movements, eyeball movements, and eye movement samples are stored in the data storage module. The data here includes eye movement sample libraries of healthy people and patients with mental illness for subsequent comparative analysis.

[0026] Code reading module: In the data analysis unit, the code reading module first reads the stored eye movement partition code. This process provides basic information for subsequent classification and matching. Classification module: After reading the code, the classification module will classify it according to the eye movement partition code of the sample to be analyzed. This step is to classify different types of eye movement data into different categories for more detailed analysis. Matching module: After the classification is completed, the matching module will match the eye movement samples of the sample to be analyzed with all the eye movements in the eye movement sample library one by one. The purpose of this step is to find samples with higher similarity for more in-depth analysis.

[0027] First data analysis module: Through the matching results, the first data analysis module matches and analyzes the eye movement data of the sample to be analyzed with the similar eye movements in the sample library, and outputs the analysis results. This result will provide a basis for the preliminary judgment of mental illness. Second data analysis module: This module is mainly responsible for comparing and analyzing the eye movements of the sample to be analyzed with normal eye movements and the eye movements of different types of mental patients. Its output results will help clinicians understand the differences between the eye movement characteristics of the person to be tested and those of normal people and patients with mental illness.

[0028] After the above analysis, the data analysis unit will transmit the comparative analysis results to the relevant application system or doctor. This process ensures that the doctor can obtain the analysis results in a timely manner and make judgments based on clinical experience. Doctors can use the analysis results output by the system in combination with other clinical information to make more accurate diagnoses of mental illnesses and develop treatment plans. Through the operations in the above stages, this technology can effectively collect and analyze the eye movement data of the persons being tested, providing a scientific basis for the diagnosis of mental illness. The core of this system is to use artificial intelligence technology to conduct a comprehensive analysis and comparison of eye movement information, helping doctors to make more accurate diagnoses and judgments in clinical practice. With the continuous development and improvement of technology, the system is expected to play a greater role in the field of mental health in the future. When performing image analysis on the eye action partition where the eye action is located, the collected image is filtered, the fundus action part is identified and analyzed, and the behavior to be tested is compared and analyzed with the sample images in the eye action sample library. The eye action sample library includes the eye action sample library of healthy people and the eye action sample library of various mental diseases.

[0029] This technical solution uses filtering processing technology in the process of fundus image analysis to effectively identify and analyze the fundus movement part. This process has several significant differences compared with the prior art. The prior art usually relies on more basic image processing methods when analyzing fundus images, which may lack pertinence and accuracy. However, this technical solution can effectively remove noise and enhance the characteristics of fundus movements through filtering processing, making the subsequent identification and analysis process more accurate, thereby improving the effectiveness and reliability of eye movement data. This technical solution compares and analyzes the behavior to be tested with the sample images in the eye movement sample library, and the sample library not only includes eye movement samples of healthy people, but also covers samples of patients with various mental illnesses. This multi-dimensional sample library design enables the system to more comprehensively consider the characteristics of different types of eye movements during comparative analysis, thereby improving the accuracy of identifying mental illnesses.

[0030] In contrast, existing technologies may only rely on a single source of sample libraries, resulting in limitations and one-sidedness in the analysis results. The processing flow of this technical solution combines advanced image processing algorithms with machine learning technology, which significantly improves the intelligence of the system in data analysis, and can automatically adapt to different changes in fundus image features, and more flexibly respond to the eye movements of different patients. However, existing technologies often rely on manual experience and lack intelligent analysis methods, which can easily lead to problems of strong subjectivity and poor stability. This technical solution shows higher precision, accuracy and adaptability than existing technologies in terms of filtering processing of fundus image analysis, diversification of sample libraries, and intelligent analysis processes, providing more effective support for the auxiliary diagnosis of mental illness.

[0031] The eye movement sample library contains basic data on the movement trajectories of blinking, nystagmus, nystagmus, strabismus, horizontal gaze, oblique gaze, eye droop, eye twitch, eye twitch, eye twitch, eye proptosis, ectropion and entropion, and defines the time and frequency information of the eye movements at the same time; different movements of the eyes at different times, the duration, number and strength of the movements are recorded using different data respectively.

[0032] The sample library of this technical solution contains detailed basic data of various eye movements, such as blinking, nystagmus, nystagmus, strabismus and other movement trajectories. This diverse action record not only enriches the content of the sample library, but also provides a more comprehensive reference for subsequent analysis. However, the existing technology often focuses on a single or a few eye movement features, lacks comprehensive capture of complex eye movement behaviors, resulting in insufficient depth and breadth of analysis. Secondly, this technical solution has made detailed definitions in recording the time and frequency information of eye movements, which can accurately capture the movement performance of the eyes at different times, including the duration, number and strength of the movement. This quantitative analysis of time and frequency can provide richer dynamic features for eye movement behavior. In contrast, the existing technology has relatively low attention in this regard, usually only focusing on static eye movement characteristics, and failing to deeply explore the changing laws of eye movement behavior. Through comprehensive recording of time and frequency, this technical solution can reveal the changing trend of eye movement behavior, thereby providing more effective data support for the diagnosis of mental illness. In addition, the introduction of basic motion trajectory data in the sample library enables the technical solution to perform more accurate pattern recognition and analysis based on historical data, improving the ability to detect various eye movement anomalies. However, existing technologies often lack systematicity and standardization in the construction of sample libraries, which may lead to insufficient data reliability and availability, thus affecting the final analysis results.

[0033] When the data analysis unit is performing analysis, it simultaneously performs analysis modeling, and constructs pupil movement states under different eye curve states through the model for regional analysis, wherein the curves include eyeball curves and pupil curves, and the eyeball and pupil are respectively divided into different regions through regional analysis, and the eyeball and pupil are divided into regions, and the positions of eye movement features are recorded, and the eye movement features are stored, wherein the eye movement features include eye curve features and pupil features, and by matching the eye action sample library, when an abnormal point appears in the user, the eye or pupil state is analyzed by regional analysis, and different regions are divided for the eyeball area and the pupil area, and the eyeball area and the pupil area are analyzed, and one of the curves is selected in the eyeball area, and the position information of the eyeball area is analyzed with the curve, and if the analysis shows that the eyeball area matches the curve successfully, the eyeball state is abnormal at this time; if the analysis shows that the eyeball area does not match the curve successfully, the eyeball state is calibrated, and the calibration is abnormal, and if the pupil area matches the curve successfully, the pupil state is calibrated, and the pupil state is abnormal at this time.

[0034] This solution introduces the concept of synchronous analysis modeling, and performs regional analysis by constructing the curve states of different eyes (including eyeball curves and pupil curves). This method provides a systematic analysis of eye movement data, so that the dynamic behavior of the eye can be described and predicted through an accurate mathematical model. However, the existing technology usually lacks such a systematic modeling process, and often relies on simple threshold judgment or static analysis, resulting in an insufficiently in-depth and comprehensive understanding of the eye movement state. Secondly, this technical solution clearly divides the eyeball and pupil into regions through regional analysis, and records the location of eye movement features, including eye curve features and pupil features. This regional division allows eye movement analysis to conduct in-depth research on specific regions, while the existing technology often regards eye movement behavior as a whole and fails to fully consider the impact of different regions on eye movement features. Through this detailed regional analysis, the system can more accurately capture the abnormal state of the eyeball and pupil, providing a more reliable basis for diagnosis. This technical solution adopts a more rigorous matching mechanism in the detection and calibration of abnormal states. When the user has an abnormal point, the system analyzes the state of the eyeball and pupil through regional analysis. If the analysis result successfully matches the model curve, it is calibrated as a normal state; if it does not match, it is calibrated as abnormal. The advantage of this method is that it not only considers the static characteristics of eye movement, but also combines the dynamic model to provide a more accurate basis for judging abnormal states. In contrast, the existing technology relies on a single static standard for abnormality detection, which often leads to misjudgment or missed judgment, reducing the accuracy and reliability of diagnosis. Finally, the solution also optimizes the storage and subsequent analysis of eye movement features, which can effectively accumulate and use historical data for continuous improvement of the model, while the existing technology often lacks such a data feedback mechanism, resulting in slow model updates and difficulty in adapting to new eye movement features.

[0035] The first data analysis module includes a matching module, an analysis module and a correction module; the matching module is used to match the eye movement of the sample to be analyzed with the same type of eye movement samples; the analysis module is used to obtain the matching degree between the sample to be tested and the same type of eye movement samples, and output data A according to the matching degree; if the data A is less than the threshold value A, the correction module processes and corrects the eye movement area coding of the sample to be tested, and stores the corrected data, repeating multiple times until the data A is greater than the threshold value A; if the data A is greater than the threshold value A, the corrected eye movement of the sample to be tested is matched and stored with the same type of eye movement samples in the eye sample movement library.

[0036] This scheme introduces a hierarchical structure of matching module, analysis module and correction module, making the data analysis process more systematic and modular. The function of the matching module is to match the eye movements of the sample to be analyzed with similar samples, which is a common function in the prior art, but the uniqueness of this scheme lies in its subsequent analysis and correction process. The analysis module not only obtains the matching degree, but also quantifies the accuracy of the matching by outputting data A. This mechanism is relatively rare in the prior art, which usually simply performs matching without further quantitative analysis.

[0037] This technical solution adopts a dynamic correction strategy when processing matching results. When data A is less than threshold A, the correction module will process and correct the eye movement area code of the sample to be tested, and store the corrected data. This process allows the system to continuously iteratively improve the matching effect of the sample until the predetermined matching standard is reached. This method avoids the limitations brought by static processing and improves the flexibility and accuracy of analysis. However, the existing technology often gives the final result after the first match, lacks further analysis and adjustment of unmatched samples, and easily leads to erroneous diagnosis or analysis results. The corrected sample data will be repeatedly processed until the threshold A is reached. This iterative optimization process enables the system to improve the matching accuracy through continuous learning and improvement. In contrast, the sample processing in the prior art is often single and lacks a feedback mechanism, and cannot be self-adjusted according to actual conditions, thereby affecting the overall analysis effect. When data A is greater than threshold A, the system matches and stores the corrected sample with similar samples in the eye sample action library. This link ensures the validity and utilization of the data. The existing technology is usually simple in sample storage and subsequent utilization, lacking a systematic and structured storage strategy.

[0038] The second data analysis module includes a normal eye movement analysis module and a mental illness eye movement analysis module; the normal eye movement analysis module is used to compare and analyze the eye movements of the sample to be analyzed with the eye movement sample library of healthy persons, and output the comparison results; the mental illness eye movement analysis module compares and analyzes the eye movements of the sample to be analyzed with the eye movement sample library of various mental illnesses, and outputs the results.

[0039] The solution sets up normal eye movement analysis and mental illness eye movement analysis as two independent modules, making the data analysis process more systematic and targeted. The normal eye movement analysis module focuses on comparing the eye movements of the samples to be analyzed with the sample library of healthy people. This process can help identify the characteristics of normal eye movements and provide a benchmark for subsequent diagnosis. However, existing technologies often mix normal and abnormal analysis together, lacking a clear distinction, resulting in an insufficient understanding of the normal state and an inability to effectively use the data of healthy samples for comparative analysis.

[0040] The mental illness eye movement analysis module focuses on comparing the eye movements of the samples to be analyzed with the sample libraries of various mental illnesses. By comparing with sample data related to various mental illnesses, the module can more accurately identify potential abnormal eye movement features. This classification and analysis method allows the system to be more targeted during the diagnosis process. Compared with the existing technology that usually relies on a single abnormal sample library or lacks sample diversity, this solution can effectively improve the recognition rate of mental illness through a multi-level sample library design.

[0041] Both modules of this technical solution emphasize the output of comparison results. This function not only helps clinicians quickly obtain analysis information, but also provides data support for the final diagnosis. Existing technologies are often simple in outputting results, lack detailed and structured information, and are difficult to meet clinical needs. Through this modular and targeted design, this technical solution significantly improves the accuracy and effectiveness of eye movement analysis, thereby providing a more scientific and reliable basis for the auxiliary diagnosis of mental illness.

[0042] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A system for detecting mental illness by eye movement based on artificial intelligence technology, characterized in that: The image acquisition unit includes an image acquisition unit, which includes a camera, an image processing module, a control module and a data storage module; the camera is used to collect the eye movements of the person to be tested, including eye movements and eyeball movements, and collect the head movements of the person to be tested at the same time; the image processing module determines whether the head of the person to be tested moves through the eye movements and head movements collected by the camera; determines whether there is head shaking by analyzing the head movements, and if there is head shaking, determines whether the eye movement is affected by the head shaking, and filters the eye pictures affected by the head shaking, and retains the eye pictures during the period when the head is still; performs image analysis on the retained eye pictures during the period when the head is still, performs image analysis on the eye movement partition where the eye movement is located, and outputs the fundus image, eyeball movement and eyeball movement; The control module encodes the eye movement partition where the eye movement is located, and outputs the code to the data storage module; The data storage module is used for storing eye movement partition codes, fundus images, eye movements, eyeball movements and eye movement sample libraries; The eye movement includes eye blinking movement; The eyeball movements include pupil deflection movements, pupil dilation and contraction movements, and pupil static movements; The eye movements include pupil deflection, pupil dilation, pupil contraction, eye blinking, pupil size remaining unchanged in a natural state, and eyelid movement; The eye movement sample library includes the eye movements of sample patients with mental illness and the eye movements of healthy people; the control module uploads the encoded eye movement partitions to the data analysis unit; The data analysis unit includes a code reading module, a classification module, a matching module, a first data analysis module and a second data analysis module; the code reading module is used to read the eye movement partition code; the classification module obtains the eye movement partition code of the sample to be analyzed, and classifies the eye movement data of the sample to be analyzed; the matching module matches the eye movement samples of the sample to be analyzed with all the eye movements in the eye movement sample library one by one; the first data analysis module is used to match and analyze the eye movement data of the sample to be analyzed with the same type of eye movements in the eye movement sample library, and output the analysis result; The second data analysis module compares and analyzes the eye movements of the sample to be analyzed with normal eye movements and eye movements of different types of mental patients according to the eye movement characteristics, and outputs the analysis results; The data analysis unit transmits the comparative analysis result.

2. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 1, characterized in that: When performing image analysis on the eye action partition where the eye action is located, the fundus action part is identified and analyzed by filtering the collected image, and the behavior to be tested is compared and analyzed with the sample images in the eye action sample library.

3. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 1, characterized in that: The eye movement sample library includes an eye movement sample library of healthy persons and an eye movement sample library of various mental illnesses.

4. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 3, characterized in that: The eye movement sample library contains basic data on the movement trajectories of blinking, nystagmus, nystagmus, strabismus, horizontal gaze, oblique gaze, eye droop, eye twitch, eye twitch, eye twitch, eye proptosis, ectropion and entropion, and defines the time and frequency information of the eye movements at the same time; different movements of the eyes at different times, the duration, number and strength of the movements are recorded using different data respectively.

5. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 1, characterized in that: When the data analysis unit is performing analysis, it simultaneously performs analysis modeling, and constructs pupil movement states under different eye curve states through the model for regional analysis, wherein the curves include eyeball curves and pupil curves, and the eyeball and pupil are respectively divided into different regions through regional analysis, and the eyeball and pupil are divided into regions, and the positions of eye movement features are recorded, and the eye movement features are stored, wherein the eye movement features include eye curve features and pupil features, and by matching the eye action sample library, when an abnormal point appears in the user, the eye or pupil state is analyzed by regional analysis, and different regions are divided for the eyeball area and the pupil area, and the eyeball area and the pupil area are analyzed, and one of the curves is selected in the eyeball area, and the position information of the eyeball area is analyzed with the curve, and if the analysis shows that the eyeball area matches the curve successfully, the eyeball state is abnormal at this time; if the analysis shows that the eyeball area does not match the curve successfully, the eyeball state is calibrated, and the calibration is abnormal, and if the pupil area matches the curve successfully, the pupil state is calibrated, and the pupil state is abnormal at this time.

6. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 1, characterized in that: The first data analysis module includes a matching module, an analysis module and a correction module; the matching module is used to match the eye movement of the sample to be analyzed with the same type of eye movement samples; the analysis module is used to obtain the matching degree between the sample to be tested and the same type of eye movement samples, and output data A according to the matching degree; If the data A is less than the threshold A, the correction module processes and corrects the eye movement area code of the sample to be tested, and stores the corrected data, and repeats the process multiple times until the data A is greater than the threshold A; If the data A is greater than the threshold A, the corrected eye movement of the sample to be tested is matched and stored with similar eye movement samples in the eye movement sample library.

7. The system for detecting mental illness by eye movement based on artificial intelligence technology as claimed in claim 1, characterized in that: The second data analysis module includes a normal eye movement analysis module and a mental illness eye movement analysis module; the normal eye movement analysis module is used to compare and analyze the eye movements of the sample to be analyzed with the eye movement sample library of healthy persons, and output the comparison results; the mental illness eye movement analysis module compares and analyzes the eye movements of the sample to be analyzed with the eye movement sample library of various mental illnesses, and outputs the results.

Images