Brain-Computer Interface-Based Eye Health Management System

Abstract

This invention relates to brain-computer interfaces, specifically to an eye health management system based on a brain-computer interface. A non-invasive multimodal signal acquisition device collects the user's multimodal signals. An effective signal acquisition module performs joint denoising on the multimodal signals to obtain an effective signal reflecting eye fatigue. A feature parameter extraction module extracts feature parameters strongly correlated with eye fatigue from the effective signal. A feature vector generation module calculates the contribution of each feature parameter to eye fatigue and generates a feature vector focusing on eye fatigue. An eye fatigue detection model training module trains the eye fatigue detection model to generate a pre-trained model. An eye fatigue level output module analyzes the feature vector based on the user's individual physiological baseline and outputs an eye fatigue level adapted to individual differences. This invention overcomes the shortcomings of not being able to extract effective signals accurately reflecting eye fatigue and the difficulty in scientifically and accurately monitoring and managing eye health.

Description

Technical Field

[0001] This invention relates to brain-computer interfaces, and more specifically to an eye health management system based on brain-computer interfaces. Background Technology

[0003] Traditional methods of monitoring eye health mainly rely on users' subjective feedback, such as their perceived level of eye fatigue. This approach lacks objectivity and accuracy, making it difficult to accurately determine the actual state and degree of eye fatigue. While some objective monitoring methods, such as simple intraocular pressure measurements and visual acuity tests, can provide some information, they suffer from limitations such as relying on single monitoring indicators, failing to comprehensively reflect the state of eye fatigue, and not being able to monitor changes in eye fatigue in real time and dynamically.

[0004] With the continuous development of technology, brain-computer interface (BCI) technology has brought new ideas and solutions to eye health management. By collecting multimodal signals related to eye use, such as electroencephalogram (EEG) and electrooculogram (EOG), richer and more objective physiological information can be obtained. However, these multimodal signals are easily affected by external environmental interference and individual differences, resulting in problems such as high noise and unstable quality. Effective signal processing technology is needed to extract effective signals that accurately reflect the state of eye fatigue, thereby achieving scientific and precise monitoring and management of eye health. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the existing technology, the present invention provides an eye health management system based on brain-computer interface, which can effectively overcome the shortcomings of the existing technology that cannot extract effective signals that accurately reflect the state of eye fatigue, and that it is difficult to scientifically and accurately monitor and manage eye health.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The brain-computer interface-based eye health management system includes a control unit. The control unit collects the user's multimodal signals through a non-invasive multimodal signal acquisition device and uses an effective signal acquisition module to perform joint noise reduction on the multimodal signals to obtain an effective signal that reflects the state of eye fatigue. The control unit extracts feature parameters that are strongly correlated with eye fatigue from the effective signals through a feature parameter extraction module and uses a feature vector generation module to generate a feature vector focusing on eye fatigue by measuring the contribution of each feature parameter to eye fatigue. The control unit constructs an eye fatigue detection model through an eye fatigue detection model construction module, and trains the eye fatigue detection model using an eye fatigue detection model training module to generate a pre-trained eye fatigue detection model. The control unit then analyzes the feature vector based on the pre-trained eye fatigue detection model and the user's individual physiological baseline using an eye fatigue level output module, outputting an eye fatigue level adapted to individual differences. Finally, the control unit uses an eye fatigue assistance adjustment module to provide personalized eye fatigue assistance adjustments to the user based on the eye fatigue level.

[0007] Preferably, the non-invasive multimodal signal acquisition device includes an eyeglass frame, an electroencephalogram (EEG) signal acquisition unit, a brain oxygenation signal acquisition unit, a heart rate signal acquisition unit, an optometry monitoring unit, a multimodal auxiliary sensor, and a neuromodulation unit. The eyeglass frame serves as the main structure of the non-invasive multimodal signal acquisition device. The EEG signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. It is set on the inner side of the temple, in the mastoid region behind the ear and the extension of the forehead. It has a total of 8 channels, of which 6 channels acquire EEG and 2 channels acquire EOG, which can accurately acquire EEG signals and eye movement signals. The brain oxygen signal acquisition unit, with the fNIRS sensor set at the forehead of the frame, is used to monitor brain oxygen saturation signals. The heart rate signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. The PPG sensor is located at the end of the temple, behind the ear, and is used to monitor heart rate and heart rate variability signals to assist in emotion assessment. The eye vision monitoring unit includes a distance sensor and an ambient light sensor, which monitors eye distance, duration and light intensity in real time. It is equipped with a dynamic focusing lens, which can initially adjust the focus according to the eye distance to relieve eye fatigue. It also has a reserved interface for monitoring tear physical parameters, and can support monitoring of biochemical indicators of eye fatigue through OTA upgrades in the future. Multimodal auxiliary sensors, including a six-axis gyroscope, a temperature sensor, and a machine vision miniature camera, capture eye features in real time, including scanning speed, gaze duration, pupil dynamic changes, and spontaneous blinking frequency, to achieve multi-dimensional personal data fusion and improve assessment accuracy. The neuromodulation unit, based on the vestibular nerve electrical stimulation (VeNS) technology, uses a low-intensity transcranial electrical stimulation (ES) module, which is placed on the inner side of the temple of the endoscope. It uses a symmetrical biphasic rectangular wave with a current intensity of 0.5~2mA, a frequency of 10~20Hz, and a pulse width of 100~500μs. The electrodes utilize nano-coating technology to improve adhesion to the skin, reduce noise interference during movement, and meet the needs of personal daily activities.

[0008] Preferably, the effective signal acquisition module performs joint denoising on the multimodal signals to acquire an effective signal that reflects the state of eye fatigue, including: Physiological events, including rapid blinking, fixation drift, and head movement, are detected using eye movement and pupil signals, while simultaneously marking time windows in EEG and heart rate variability signals that are affected by physiological events. Based on the interference characteristics of each modal signal, joint denoising of multimodal signals is performed by combining marked time windows; By comparing the correlation between the multimodal signals before and after denoising, if the deviation exceeds the preset threshold, the above process is repeated until the requirements are met, and finally an effective signal that can reflect the state of eye fatigue is obtained.

[0009] Preferably, the joint denoising of multimodal signals based on the interference characteristics of each modal signal and combined with marked time windows includes: For EEG signals, independent component analysis was used to separate and remove electromyographic artifacts, while signal segments affected by electrooculography artifacts were corrected or removed. For eye movement and pupil signals, ambient light interference is eliminated through filtering, and image shift caused by head movement is corrected by combining image registration algorithm. For heart rate variability signals, motion artifacts are removed by bandpass filtering, while disturbed abnormal heartbeat intervals are smoothed out.

[0010] Preferably, the feature parameter extraction module extracts feature parameters strongly correlated with eye fatigue from the effective signal, including: For EEG signals, frequency domain features reflecting the degree of visual cortex fatigue are extracted, including the power spectral density ratio of alpha waves to beta waves and the power spectral density ratio of theta waves to beta waves. For eye movement signals and pupil signals, temporal features reflecting the fatigue state of eye muscles and pupil accommodation are extracted, including blink frequency, pupil diameter change rate and fixation point drift amplitude. For heart rate variability signals, frequency domain features reflecting autonomic nervous system dysfunction caused by fatigue are extracted, including the ratio of low-frequency to high-frequency power and total power.

[0011] Preferably, the feature vector generation module generates a feature vector of focusing eye fatigue by measuring the contribution of each feature parameter to eye fatigue, including: The contribution of each feature parameter to eye fatigue is learned through a feature weight allocation mechanism, and the weight of each feature parameter is dynamically adjusted according to the strength of the correlation between each feature parameter and eye fatigue. Each feature parameter is multiplied by its corresponding weight and then concatenated to generate a feature vector of eye fatigue caused by focusing.

[0012] Preferably, the eye fatigue detection model training module trains the eye fatigue detection model to generate a pre-trained eye fatigue detection model, including: During initial calibration, multimodal signals under normal eye use conditions are collected, corresponding feature vectors are extracted, and an initial model is trained to establish an individual physiological baseline for the user. During the online learning phase, multimodal signals are continuously collected and corresponding feature vectors are extracted during the user's daily use. The model parameters are dynamically updated through incremental learning algorithms to adapt the model to changes in the user's individual eye habits.

[0013] Preferably, the eye fatigue level output module is based on a pre-trained eye fatigue detection model, analyzes the feature vector in conjunction with the user's individual physiological baseline, and outputs an eye fatigue level adapted to individual differences, including: The feature vectors are input into the input layer of the eye fatigue detection model as the initial data for model analysis. The pre-trained eye fatigue detection model learns the correlation between multimodal signals through the interaction layer, performs in-depth analysis on the input feature vector, and mines the potential correlation between features to assess the state of eye fatigue. The eye fatigue detection model classifies the analysis results through the output layer and finally outputs an eye fatigue level that is adapted to individual differences.

[0014] Preferably, the eye-use assistance adjustment module provides personalized eye-use assistance adjustments for the user based on the level of eye fatigue, including: When eye fatigue levels exceed a preset threshold, the system sends a triple warning—a slight vibration of the temples, an app notification, and a voice alert—to parents, allowing for timely intervention in unhealthy eye-use behaviors.

[0015] Preferably, the eye-use assistance adjustment module provides personalized eye-use assistance adjustments for the user based on the level of eye fatigue, including: Visual Relaxation Mode: When no adjustment is made after the reminder is triggered, the "Visual Relaxation Mode" is activated: the lens is focused to a distant viewing state, and at the same time, the visual cortex relaxation area is activated through low-intensity transcranial electrical stimulation to help relieve eye fatigue and promote eye muscle relaxation training. Brain-controlled training linkage: We recommend one "personal flying ball eye muscle training" session per day. By controlling the up, down, left, and right movements of the flying ball through brain waves, we can train the flexibility of the eye muscles, promote choroidal circulation, help improve vision, and enhance interactivity. Cervical spine coordination training: The "head movement control" section is added to the "personal flying ball eye muscle training" to guide users to turn their heads to control the trajectory of the flying ball, activate the cervical spine, relieve cervical spine fatigue, strengthen the neck muscles, and solve cervical spine problems caused by prolonged sitting. Health monitoring: Daily eye health monitoring reports are generated to provide data support for optometrists and help develop personalized myopia prevention and control plans.

[0016] Compared with existing technologies, the brain-computer interface-based eye health management system provided by this invention has the following beneficial effects: 1) Accurately capture fatigue signals and improve monitoring accuracy. The system employs a non-invasive multimodal signal acquisition device, capable of simultaneously acquiring multi-dimensional signals such as EEG, eye movement, brain oxygenation, heart rate, and visual acuity. The effective signal acquisition module utilizes combined denoising technology, marking interference time windows with eye movement and pupil signals to specifically remove artifacts from each modality, ensuring the acquisition of effective signals that reflect the state of eye fatigue. The feature parameter extraction module extracts feature parameters strongly correlated with eye fatigue from the effective signals, providing rich data support for accurate assessment of eye fatigue and greatly improving the accuracy of eye fatigue monitoring. 2) Adapt to individual differences and provide personalized services During initial calibration, the system collects multimodal signals from users under normal eye-use conditions to establish an individual physiological baseline. During the online learning phase, it dynamically updates model parameters using an incremental learning algorithm, enabling the model to adapt to changes in individual user eye-use habits. The eye fatigue level output module analyzes the feature vectors based on the individual physiological baseline, outputting an eye fatigue level tailored to individual differences. The eye-use assistance adjustment module provides personalized eye-use assistance adjustments based on different eye fatigue levels, meeting the individual needs of different users and improving user experience and eye health management effectiveness. 3) Multi-dimensional coordinated regulation to achieve comprehensive health management This system not only focuses on monitoring and adjusting eye fatigue, but also comprehensively considers the connection between eye health and overall health. In addition to adjusting eye fatigue, the eye-assisted adjustment also incorporates a cervical spine coordination training component, guiding users to rotate their heads to control the trajectory of the flying ball, activating the cervical spine, relieving cervical spine fatigue, strengthening neck muscles, and solving cervical spine problems caused by prolonged sitting. At the same time, the system generates daily eye health monitoring reports, providing data support for optometrists to help develop personalized myopia prevention and control plans, and achieving comprehensive management of users' eye health and overall health. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is a schematic diagram of the system of the present invention; Figure 2 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] The following describes the specific functional modules of the brain-computer interface-based eye health management system provided by this invention, using concrete examples (such as...). Figure 1 (as shown) and technical effects.

[0021] The system functional modules include: a control unit, which collects the user's multimodal signals through a non-invasive multimodal signal acquisition device, and uses an effective signal acquisition module to perform joint denoising on the multimodal signals to obtain an effective signal that can reflect the state of eye fatigue. The control unit extracts feature parameters that are strongly correlated with eye fatigue from the effective signals through a feature parameter extraction module, and uses a feature vector generation module to generate a feature vector of focusing eye fatigue by measuring the contribution of each feature parameter to eye fatigue. The control unit constructs an eye fatigue detection model through an eye fatigue detection model construction module, and trains the eye fatigue detection model using an eye fatigue detection model training module to generate a pre-trained eye fatigue detection model. Based on the pre-trained eye fatigue detection model and combined with the user's individual physiological baseline, the control unit analyzes the feature vector through an eye fatigue level output module, outputs an eye fatigue level adapted to individual differences, and uses an eye fatigue assistance adjustment module to provide personalized eye fatigue assistance adjustment for the user according to the eye fatigue level.

[0022] I. Non-invasive multimodal signal acquisition device The non-invasive multimodal signal acquisition device includes an eyeglass frame, an electroencephalogram (EEG) signal acquisition unit, a brain oxygen signal acquisition unit, a heart rate signal acquisition unit, an optometry monitoring unit, a multimodal auxiliary sensor, and a neuromodulation unit. The eyeglass frame serves as the main structure of the non-invasive multimodal signal acquisition device. The EEG signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. It is set on the inner side of the temple, in the mastoid region behind the ear and the extension of the forehead. It has a total of 8 channels, of which 6 channels acquire EEG and 2 channels acquire EOG, which can accurately acquire EEG signals and eye movement signals. The brain oxygen signal acquisition unit, with the fNIRS sensor set at the forehead of the frame, is used to monitor brain oxygen saturation signals. The heart rate signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. The PPG sensor is located at the end of the temple, behind the ear, and is used to monitor heart rate and heart rate variability signals to assist in emotion assessment. The eye vision monitoring unit includes a distance sensor and an ambient light sensor, which monitors eye distance, duration and light intensity in real time. It is equipped with a dynamic focusing lens, which can initially adjust the focus according to the eye distance to relieve eye fatigue. It also has a reserved interface for monitoring tear physical parameters, and can support monitoring of biochemical indicators of eye fatigue through OTA upgrades in the future. Multimodal auxiliary sensors, including a six-axis gyroscope, a temperature sensor, and a machine vision miniature camera, capture eye features in real time, including scanning speed, gaze duration, pupil dynamic changes, and spontaneous blinking frequency, to achieve multi-dimensional personal data fusion and improve assessment accuracy. The neuromodulation unit, based on the vestibular nerve electrical stimulation (VeNS) technology, uses a low-intensity transcranial electrical stimulation (ES) module, which is placed on the inner side of the temple of the endoscope. It uses a symmetrical biphasic rectangular wave with a current intensity of 0.5~2mA, a frequency of 10~20Hz, and a pulse width of 100~500μs. The electrodes utilize nano-coating technology to improve adhesion to the skin, reduce noise interference during movement, and meet the needs of personal daily activities.

[0023] II. Effective Signal Acquisition Module The effective signal acquisition module performs joint denoising on multimodal signals to acquire effective signals that reflect eye fatigue status, including: Physiological events, including rapid blinking, fixation drift, and head movement, are detected using eye movement and pupil signals, while simultaneously marking time windows in EEG and heart rate variability signals that are affected by physiological events. Based on the interference characteristics of each modal signal, joint denoising of multimodal signals is performed by combining marked time windows; By comparing the correlation between the multimodal signals before and after denoising, if the deviation exceeds the preset threshold, the above process is repeated until the requirements are met, and finally an effective signal that can reflect the state of eye fatigue is obtained.

[0024] Specifically, based on the interference characteristics of each modal signal, joint denoising of the multimodal signals is performed by combining marked time windows, including: For EEG signals, independent component analysis was used to separate and remove electromyographic artifacts, while signal segments affected by electrooculography artifacts were corrected or removed. For eye movement and pupil signals, ambient light interference is eliminated through filtering, and image shift caused by head movement is corrected by combining image registration algorithm. For heart rate variability signals, motion artifacts are removed by bandpass filtering, while disturbed abnormal heartbeat intervals are smoothed out.

[0025] III. Feature Parameter Extraction Module The feature parameter extraction module extracts feature parameters strongly correlated with eye fatigue from the effective signal, including: For EEG signals, frequency domain features reflecting the degree of visual cortex fatigue are extracted, including the power spectral density ratio of alpha waves to beta waves and the power spectral density ratio of theta waves to beta waves. For eye movement signals and pupil signals, temporal features reflecting the fatigue state of eye muscles and pupil accommodation are extracted, including blink frequency, pupil diameter change rate and fixation point drift amplitude. For heart rate variability signals, frequency domain features reflecting autonomic nervous system dysfunction caused by fatigue are extracted, including the ratio of low-frequency to high-frequency power and total power.

[0026] IV. Feature Vector Generation Module The feature vector generation module calculates the contribution of each feature parameter to eye fatigue, generating a feature vector for focusing eye fatigue, including: The contribution of each feature parameter to eye fatigue is learned through a feature weight allocation mechanism, and the weight of each feature parameter is dynamically adjusted according to the strength of the correlation between each feature parameter and eye fatigue. Each feature parameter is multiplied by its corresponding weight and then concatenated to generate a feature vector of eye fatigue caused by focusing.

[0027] The above technical solution employs a non-invasive multimodal signal acquisition device, which can simultaneously acquire multi-dimensional signals such as EEG, eye movement, brain oxygenation, heart rate, and visual acuity. The effective signal acquisition module uses combined denoising technology to mark interference time windows using eye movement and pupil signals, specifically removing artifacts from each modality signal to ensure the acquisition of effective signals that reflect the state of eye fatigue. The feature parameter extraction module extracts feature parameters that are strongly correlated with eye fatigue from the effective signals, providing rich data support for accurate assessment of eye fatigue and greatly improving the accuracy of eye fatigue monitoring.

[0028] V. Eye Fatigue Detection Model Training Module The eye fatigue detection model training module trains the eye fatigue detection model to generate a pre-trained eye fatigue detection model, including: During initial calibration, multimodal signals under normal eye use conditions are collected, corresponding feature vectors are extracted, and an initial model is trained to establish an individual physiological baseline for the user. During the online learning phase, multimodal signals are continuously collected and corresponding feature vectors are extracted during the user's daily use. The model parameters are dynamically updated through incremental learning algorithms to adapt the model to changes in the user's individual eye habits.

[0029] VI. Eye Fatigue Level Output Module The eye fatigue level output module is based on a pre-trained eye fatigue detection model. It analyzes the feature vector by combining the user's individual physiological baseline and outputs an eye fatigue level adapted to individual differences, including: The feature vectors are input into the input layer of the eye fatigue detection model as the initial data for model analysis. The pre-trained eye fatigue detection model learns the correlation between multimodal signals through the interaction layer, performs in-depth analysis on the input feature vector, and mines the potential correlation between features to assess the state of eye fatigue. The eye fatigue detection model classifies the analysis results through the output layer and finally outputs an eye fatigue level that is adapted to individual differences.

[0030] VII. Eye-Assisting Adjustment Module The eye-use assistance module provides personalized eye-use assistance adjustments based on the user's level of eye fatigue, including: When eye fatigue levels exceed a preset threshold, the system sends a triple warning—a slight vibration of the temples, an app notification, and a voice alert—to parents, allowing for timely intervention in unhealthy eye-use behaviors.

[0031] The eye-use assistance module provides personalized eye-use assistance adjustments based on the user's level of eye fatigue, including: Visual Relaxation Mode: When no adjustment is made after the reminder is triggered, the "Visual Relaxation Mode" is activated: the lens is focused to a distant viewing state, and at the same time, the visual cortex relaxation area is activated through low-intensity transcranial electrical stimulation to help relieve eye fatigue and promote eye muscle relaxation training. Brain-controlled training linkage: We recommend one "personal flying ball eye muscle training" session per day. By controlling the up, down, left, and right movements of the flying ball through brain waves, we can train the flexibility of the eye muscles, promote choroidal circulation, help improve vision, and enhance interactivity. Cervical spine coordination training: The "head movement control" section is added to the "personal flying ball eye muscle training" to guide users to turn their heads to control the trajectory of the flying ball, activate the cervical spine, relieve cervical spine fatigue, strengthen the neck muscles, and solve cervical spine problems caused by prolonged sitting. Health monitoring: Daily eye health monitoring reports are generated to provide data support for optometrists and help develop personalized myopia prevention and control plans.

[0032] The above technical solution collects multimodal signals from users under normal eye use conditions during initial calibration to establish an individual physiological baseline for each user. During the online learning phase, it dynamically updates model parameters through an incremental learning algorithm, enabling the model to adapt to changes in individual user eye use habits. The eye fatigue level output module analyzes the feature vectors in conjunction with the individual user physiological baseline and outputs an eye fatigue level that adapts to individual differences. The eye use assistance adjustment module provides personalized eye use assistance adjustments for users based on different eye fatigue levels, meeting the individual needs of different users and improving the user experience and eye health management effectiveness.

[0033] Meanwhile, the above-mentioned technical solutions not only focus on monitoring and adjusting eye fatigue, but also comprehensively consider the relationship between eye health and overall health. In addition to adjusting eye fatigue, the eye-assisted adjustment also incorporates a cervical spine coordination training component, guiding users to rotate their heads to control the trajectory of the flying ball, activating the cervical spine, relieving cervical spine fatigue, strengthening neck muscles, and solving cervical spine problems caused by prolonged sitting. At the same time, the system generates daily eye health monitoring reports, providing data support for optometrists to assist in the development of personalized myopia prevention and control plans, and achieving comprehensive management of users' eye health and overall health.

[0034] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A brain-computer interface-based eye health management system, characterized in that: The system includes a control unit, which acquires the user's multimodal signals through a non-invasive multimodal signal acquisition device and performs joint noise reduction on the multimodal signals using an effective signal acquisition module to obtain an effective signal that reflects the state of eye fatigue. The control unit extracts feature parameters that are strongly correlated with eye fatigue from the effective signal through a feature parameter extraction module and generates a feature vector of focusing eye fatigue by using the contribution of each feature parameter to eye fatigue through a feature vector generation module. The control unit constructs an eye fatigue detection model through an eye fatigue detection model construction module, and trains the eye fatigue detection model using an eye fatigue detection model training module to generate a pre-trained eye fatigue detection model. The control unit then analyzes the feature vector based on the pre-trained eye fatigue detection model and the user's individual physiological baseline using an eye fatigue level output module, outputting an eye fatigue level adapted to individual differences. Finally, the control unit uses an eye fatigue assistance adjustment module to provide personalized eye fatigue assistance adjustments to the user based on the eye fatigue level.

2. The brain-computer interface-based eye health management system according to claim 1, characterized in that: The non-invasive multimodal signal acquisition device includes an eyeglass frame, an electroencephalogram (EEG) signal acquisition unit, a brain oxygen signal acquisition unit, a heart rate signal acquisition unit, an optometry monitoring unit, a multimodal auxiliary sensor, and a neuromodulation unit. The eyeglass frame serves as the main structure of the non-invasive multimodal signal acquisition device. The EEG signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. It is set on the inner side of the temple, in the mastoid region behind the ear and the extension of the forehead. It has a total of 8 channels, of which 6 channels acquire EEG and 2 channels acquire EOG, which can accurately acquire EEG signals and eye movement signals. The brain oxygen signal acquisition unit, with the fNIRS sensor set at the forehead of the frame, is used to monitor brain oxygen saturation signals. The heart rate signal acquisition unit uses dry electrodes that incorporate the properties of flexible biomaterials. The PPG sensor is located at the end of the temple, behind the ear, and is used to monitor heart rate and heart rate variability signals to assist in emotion assessment. The eye vision monitoring unit includes a distance sensor and an ambient light sensor, which monitors eye distance, duration and light intensity in real time. It is equipped with a dynamic focusing lens, which can initially adjust the focus according to the eye distance to relieve eye fatigue. It also has a reserved interface for monitoring tear physical parameters, and can support monitoring of biochemical indicators of eye fatigue through OTA upgrades in the future. Multimodal auxiliary sensors, including a six-axis gyroscope, a temperature sensor, and a machine vision miniature camera, capture eye features in real time, including scanning speed, gaze duration, pupil dynamic changes, and spontaneous blinking frequency, to achieve multi-dimensional personal data fusion and improve assessment accuracy. The neuromodulation unit, based on the vestibular nerve electrical stimulation (VeNS) technology, uses a low-intensity transcranial electrical stimulation (ES) module, which is placed on the inner side of the temple of the endoscope. It uses a symmetrical biphasic rectangular wave with a current intensity of 0.5~2mA, a frequency of 10~20Hz, and a pulse width of 100~500μs. The electrodes utilize nano-coating technology to improve adhesion to the skin, reduce noise interference during movement, and meet the needs of personal daily activities.

3. The brain-computer interface-based eye health management system according to claim 1, characterized in that: The effective signal acquisition module performs joint denoising on the multimodal signals to acquire an effective signal that reflects the state of eye fatigue, including: Physiological events, including rapid blinking, fixation drift, and head movement, are detected using eye movement and pupil signals, while simultaneously marking time windows in EEG and heart rate variability signals that are affected by physiological events. Based on the interference characteristics of each modal signal, joint denoising of multimodal signals is performed by combining marked time windows; By comparing the correlation between the multimodal signals before and after denoising, if the deviation exceeds the preset threshold, the above process is repeated until the requirements are met, and finally an effective signal that can reflect the state of eye fatigue is obtained.

4. The brain-computer interface-based eye health management system according to claim 3, characterized in that: The method of jointly denoising multimodal signals based on the interference characteristics of each modal signal and combining the marked time window includes: For EEG signals, independent component analysis was used to separate and remove electromyographic artifacts, while signal segments affected by electrooculography artifacts were corrected or removed. For eye movement and pupil signals, ambient light interference is eliminated through filtering, and image shift caused by head movement is corrected by combining image registration algorithm. For heart rate variability signals, motion artifacts are removed by bandpass filtering, while disturbed abnormal heartbeat intervals are smoothed out.

5. The brain-computer interface-based eye health management system according to claim 3, characterized in that: The feature parameter extraction module extracts feature parameters strongly correlated with eye fatigue from the effective signal, including: For EEG signals, frequency domain features reflecting the degree of visual cortex fatigue are extracted, including the power spectral density ratio of alpha waves to beta waves and the power spectral density ratio of theta waves to beta waves. For eye movement signals and pupil signals, temporal features reflecting the fatigue state of eye muscles and pupil accommodation are extracted, including blink frequency, pupil diameter change rate and fixation point drift amplitude. For heart rate variability signals, frequency domain features reflecting autonomic nervous system dysfunction caused by fatigue are extracted, including the ratio of low-frequency to high-frequency power and total power.

6. The brain-computer interface-based eye health management system according to claim 5, characterized in that: The feature vector generation module generates a feature vector for focusing on eye fatigue by calculating the contribution of each feature parameter to eye fatigue, including: The contribution of each feature parameter to eye fatigue is learned through a feature weight allocation mechanism, and the weight of each feature parameter is dynamically adjusted according to the strength of the correlation between each feature parameter and eye fatigue. Each feature parameter is multiplied by its corresponding weight and then concatenated to generate a feature vector of eye fatigue caused by focusing.

7. The brain-computer interface-based eye health management system according to claim 6, characterized in that: The eye fatigue detection model training module trains the eye fatigue detection model to generate a pre-trained eye fatigue detection model, including: During initial calibration, multimodal signals under normal eye use conditions are collected, corresponding feature vectors are extracted, and an initial model is trained to establish an individual physiological baseline for the user. During the online learning phase, multimodal signals are continuously collected and corresponding feature vectors are extracted during the user's daily use. The model parameters are dynamically updated through incremental learning algorithms to adapt the model to changes in the user's individual eye habits.

8. The brain-computer interface-based eye health management system according to claim 7, characterized in that: The eye fatigue level output module is based on a pre-trained eye fatigue detection model. It analyzes the feature vector by combining the user's individual physiological baseline and outputs an eye fatigue level adapted to individual differences, including: The feature vectors are input into the input layer of the eye fatigue detection model as the initial data for model analysis. The pre-trained eye fatigue detection model learns the correlation between multimodal signals through the interaction layer, performs in-depth analysis on the input feature vector, and mines the potential correlation between features to assess the state of eye fatigue. The eye fatigue detection model classifies the analysis results through the output layer and finally outputs an eye fatigue level that is adapted to individual differences.

9. The brain-computer interface-based eye health management system according to claim 8, characterized in that: The eye-use assistance adjustment module provides personalized eye-use assistance adjustments for users based on their eye fatigue level, including: When eye fatigue levels exceed a preset threshold, the system sends a triple warning—a slight vibration of the temples, an app notification, and a voice alert—to parents, allowing for timely intervention in unhealthy eye-use behaviors.

10. The brain-computer interface-based eye health management system according to claim 9, characterized in that: The eye-use assistance adjustment module provides personalized eye-use assistance adjustments for users based on their eye fatigue level, including: Visual Relaxation Mode: When no adjustment is made after the reminder is triggered, the "Visual Relaxation Mode" is activated: the lens is focused to a distant viewing state, and at the same time, the visual cortex relaxation area is activated through low-intensity transcranial electrical stimulation to help relieve eye fatigue and promote eye muscle relaxation training. Brain-controlled training linkage: We recommend one "personal flying ball eye muscle training" session per day. By controlling the up, down, left, and right movements of the flying ball through brainwaves, we can train the flexibility of the eye muscles, promote choroidal circulation, help improve vision, and enhance interactivity. Cervical spine coordination training: The "head movement control" section is added to the "personal flying ball eye muscle training" to guide users to turn their heads to control the trajectory of the flying ball, activate the cervical spine, relieve cervical spine fatigue, strengthen the neck muscles, and solve cervical spine problems caused by prolonged sitting. Health monitoring: Daily eye health monitoring reports are generated to provide data support for optometrists and help develop personalized myopia prevention and control plans.

Images