Ciliary muscle accommodation-based child visual function training system and method

Abstract

The scheme belongs to the technical field of children's vision training, and particularly relates to a children's visual function training system and method based on ciliary muscle accommodation. The children's visual function training system based on ciliary muscle accommodation comprises a cooperative acquisition module and a dynamic training module. The cooperative acquisition module acquires real-time eyeball anterior segment images through an inbuilt image acquisition unit to identify eye features, acquires eye muscle pressure deformation data through an inbuilt pressure acquisition unit, and then correlates and analyzes the eye features and the eye muscle pressure deformation data through an inbuilt data fusion unit. A mapping model is established according to the analysis result, and the change in the lens refractive power is calculated. The calculation result is processed as the ciliary muscle accommodation amount. The dynamic training module determines training parameters in combination with individual data of children and historical data acquired by the cooperative acquisition module, and dynamically adjusts the training duration and light wave parameters according to the ciliary muscle accommodation amount, thereby solving the problems of poor training effect of children and safety risks in the training process.

Description

TECHNICAL FIELD

[0001] The present scheme belongs to the technical field of children's vision training, and particularly relates to a children's visual function training system and method based on ciliary muscle accommodation. BACKGROUND

[0002] In the field of children's visual function training, real-time accurate acquisition of ciliary muscle state and dynamic adaptation of training plan are the core problems to improve the effect, and most of the current technical solutions have significant shortcomings.

[0003] As the key smooth muscle for regulating the focal length of the lens inside the eyeball, the contraction or relaxation state of the ciliary muscle directly determines the visual accommodation ability (contraction to make the lens convex when looking near, relaxation to make the lens flat when looking far). From the physiological mechanism, the movement of the ciliary muscle is accurate to calculate the deformation of the lens, because the lens itself is directly pulled by the suspension ligament through the contraction and relaxation of the ciliary muscle, and then realizes the change of its curvature, and there is a direct mechanical linkage between the two; but conversely, it is not accurate to calculate the accommodation of the ciliary muscle through the deformation of the lens, because the deformation of the lens may also be affected by other factors such as the elasticity of the lens itself and the tension of the suspension ligament, and single lens deformation data cannot completely reflect the real contraction or relaxation state of the ciliary muscle.

[0004] And the anatomical characteristics of the ciliary muscle located inside the eyeball and in a ring shape make its detection itself have a congenital hindrance, the ring structure tightly wraps the lens, and is adjacent to the choroid and suspension ligament, so it is difficult for conventional optical equipment to penetrate the eyeball wall to accurately capture the activity of muscle fibers; in addition, the contraction amplitude of muscle fibers is only microns, and subtle movements are easily covered by interference signals such as iris tremor and corneal reflection, making it difficult to extract effective information alone. This leads to two technical bottlenecks in real-time state acquisition: first, it cannot be directly measured, and needs to be inferred through indirect indicators such as lens curvature change and eyeball micro-movement, and the existing technology has single collection means, only through the pre-optic image to infer the accommodation of the ciliary muscle to the lens, it is difficult to establish a precise correlation between indirect indicators and the real state of the ciliary muscle; second, the ciliary muscle of children is not yet mature, with a large accommodation amplitude (for example, 12-14D for children aged 6-8 years, and about 8D for adults) and easy spasm, and its state fluctuates more severely with training time and eye use intensity, and traditional static collection (such as single optometry data) cannot capture dynamic changes, resulting in insufficient data timeliness.

[0005] More importantly, the state of the ciliary muscle plays a decisive role in the formulation and adjustment of training plans. Most current testing methods still revolve around the lens. Whether it's obtaining the refractive power of the lens through optometry equipment or measuring changes in lens curvature using optical instruments, these are essentially tests of visual acuity-related "outcome indicators," reflecting only the current visual level (such as myopia degree or degree of refractive error), but failing to address the dynamic functional state of the ciliary muscle behind the visual results. If the ciliary muscle is in a "spasmodic state" (e.g., the ciliary muscle cannot relax after continuous near work), continued high-intensity training will exacerbate fatigue and even lead to an increase in myopia; if it is in an "active state" (e.g., a sensitive accommodative response), low-intensity training cannot fully activate the accommodative reserve. Most current training programs generally ignore this connection, and most training devices rely on human experience for parameter adjustments (e.g., manually modifying based on visual improvement), failing to acquire the ciliary muscle state in real time through technological means. For example, when a child's ciliary muscle pressure-related signal (extraocular muscle deformation) indicates fatigue, the training device cannot automatically reduce the training intensity, easily causing temporary blurred vision after training. Summary of the Invention

[0006] The purpose of this solution is to provide a visual function training system and method for children based on ciliary muscle accommodation, in order to solve the problems of lack of real-time acquisition of ciliary muscle status and the disconnect between ciliary muscle status and training plan, which leads to poor training results or safety risks for children.

[0007] To achieve the above objectives, this solution provides a visual function training system for children based on ciliary muscle accommodation, comprising:

[0008] The collaborative acquisition module includes an image acquisition unit, a pressure acquisition unit, and a data fusion unit;

[0009] The image acquisition unit acquires images of the anterior segment of the eye in real time and identifies ocular feature parameters in the images of the anterior segment of the eye.

[0010] The pressure acquisition unit includes a pressure sensor located on the skin surface at the attachment point of the extraocular muscles on the outer side of the eye socket, which collects pressure deformation data of the extraocular muscles caused by eyeball convergence movement in real time.

[0011] The data fusion unit acquires ocular feature parameters and pressure deformation data collected at the same time, and performs correlation analysis. It establishes a mapping model based on the changes in pressure deformation data and lens curvature. The lens refractive power change data is indirectly calculated based on the ocular feature parameters using a thin lens refractive power conversion algorithm as the first data. The pressure deformation data is input into the mapping model to calculate the real-time state of the ciliary muscle, and the lens refractive power change data in the ocular feature parameters is calculated based on the real-time state as the second data.

[0012] The data fusion unit compares the first data with the second data, if the comparison result is within a preset error threshold, the real-time state is processed into a real-time adjustment amount of ciliary muscle according to the collection time; if the comparison result is out of the error threshold range, the image collection unit and the pressure collection unit are used to obtain the eye feature parameters and the pressure deformation data again, and the first data and the second data are calculated and compared;

[0013] The dynamic training module obtains individual data of the child and historical data of the cooperative collection module, formulates training parameters according to a preset training period, and dynamically adjusts the training parameters according to the real-time adjustment amount of the ciliary muscle; the real-time state includes a fatigue state and an active state, and the training parameters include a training time length and a light wave combination; the training time length, the target frequency, the proportion of different types of light in the light wave combination, and the contrast ratio are dynamically adjusted according to the real-time state; and the training mode in the control parameters is adjusted according to the continuous real-time state.

[0014] In addition, a ciliary muscle adjustment-based child visual function training method using the ciliary muscle adjustment-based child visual function training system is provided.

[0015] The principle and technical effect of the scheme are that the scheme relies on the dynamic adaptation mechanism of the ciliary muscle state and the training parameters, and constructs a closed loop combining perception, analysis, and control through multi-dimensional collection and intelligent control. The cooperative collection unit serves as a state perception layer, captures internal features such as lens curvature and iris displacement through the eye image collection unit, synchronously obtains external correlation signals of eye muscle deformation through an orbital pressure sensor, establishes a mapping model of lens adjustment and eye muscle pressure through the data fusion unit, breaks through the bottleneck that the ciliary muscle cannot be directly measured, and realizes indirect determination of the tension, relaxation, and fatigue states; the dynamic training module adjusts the training parameters based on the real-time state, reduces the target switching frequency and increases the proportion of green light for relaxation in the fatigue state, and increases the frequency and the proportion of red light for strengthening training in the active state, while avoiding overtraining through automatic mode switching; individual differences such as the age and amblyopia degree of the child are combined to generate a personalized scheme including the time length and the light wave combination, and a complete closed loop is formed. The scheme solves the problem that the traditional technology cannot quantify the dynamic changes of the ciliary muscle of the child, improves the state determination accuracy, realizes dynamic and accurate training, reduces the risk of ciliary muscle spasm, and improves the adjustment amplitude.

[0016] The scheme can further improve the reliability of state determination through multi-source data fusion. The cooperative collection of the anterior segment image and the pressure deformation data can offset the interference factors of single data (such as the influence of the elasticity of the lens itself on the anterior segment image of the eye), so that the ciliary muscle state determination is more in line with the actual situation, and even in the scene where the child's eye moves frequently, the determination effect remains stable.

[0017] The dynamic parameter regulation and multispectral coordination principle in the scheme can enhance the training coordination. The multispectral module regulates the proportion of different types of light in the light wave combination, such as regulating the proportion of green light in the light wave combination, adjusting the pupil diameter to enhance the linkage between the ciliary muscle and the visual system; regulating the proportion of red light in the light wave combination, stimulating choroid blood circulation, both of which form a synergistic effect with ciliary muscle training, improve the efficiency of eye blood circulation, provide physiological support for the activation of visual cells in amblyopic children, and make the training effect more lasting.

[0018] The scheme combines contrast regulation with dynamic training, which can strengthen the stability of the ciliary muscle in complex visual environments. For example, by dynamically adjusting the contrast of the picture to simulate complex visual environments, and training in real-time with the state of the ciliary muscle, the adaptability of the ciliary muscle under different visual conditions can be improved, further consolidating the adjustment function.

[0019] The scheme combines personalized scheme generation with dynamic adaptation, which can accurately match the development differences of children. For different stages of ciliary muscle development in children aged 6-12, combined with age, amblyopia degree and historical data, the training scheme is customized to avoid the limitations of general mode, shorten the amblyopia correction period, and achieve better myopia prevention and control effect.

[0020] The state perception and mode switching linkage in the scheme can deepen the training safety guarantee. By identifying ciliary muscle fatigue in advance (rather than relying solely on external behavior monitoring), the training mode is automatically switched, for example, according to the fact that the ciliary muscle is in a fatigued state for several times in a row, the training mode is switched to a brain training mode, avoiding temporary visual decline caused by excessive load, significantly improving the safety and compliance of children's training, and reducing training resistance.

[0021] In summary, the scheme realizes real-time ciliary muscle state judgment through multi-source closed loop, dynamically adapts training parameters, accurately improves the effect, and solves the problem of poor training effect or safety risk caused by lack of real-time state acquisition and training disconnection.

[0022] Further, the data fusion unit is also used for precision verification of the AC / A constant according to synchronous acquisition of pressure deformation data and eye feature parameters. When the individual error of AC / A is ≤±0.3Δ / D and the error of the mapping model after calibration by UBM and open field optometry is ≤±0.25D, the objective accommodation response is recorded by the infrared video optometer or open field automatic optometer as the golden standard of accommodation, and the instantaneous convergence is measured by the prism cover method or synoptophore as the golden standard of convergence, to ensure that the AC / A obtained from the golden standards of accommodation and convergence has a difference of ≤±0.3Δ / D with the synchronous acquisition of pressure deformation data and eye feature parameters. Then, the accommodation stimulus is applied at different viewing distances and repeated multiple times, and the stimulus response curve is drawn. If the linear fitting R² of the stimulus response curve is <0.9 or the intercept is >0.5Δ, then the segmented fitting or quadratic curve correction is performed. The phoria at far and near distances is measured by the cover and uncover methods. For those with a phoria of >8Δ, a separate model is established. For those with a difference in dioptric power of >1D, the average value of the left and right eye AC / A is taken after calibration, and the repeated measurement on the next day ensures that the difference between two AC / A values is <±0.2Δ / D.

[0023] The present scheme forms a multi-link closed-loop verification mechanism through the precision verification of the AC / A constant by the data fusion unit. The AC / A error is controlled within ±0.3Δ / D, and the mapping model error is locked within ±0.25D, which directly improves the accuracy of the ciliary muscle state calculation and provides reliable data support for the dynamic training module. At the same time, the scheme automatically corrects the abnormal stimulus response curve according to the characteristics of the large accommodation fluctuation of children, avoids the failure of linear assumption, and enhances the stability of the scheme. For special children with a phoria of >8Δ or a difference in dioptric power of >1D, a separate model is established, and the consistency is ensured by repeated measurement on the next day, realizing individualized adaptation and solving the problem of insufficient adaptation of traditional methods for special groups. Moreover, the verified AC / A can refresh the training parameters in real time, and cooperates with the dynamic training module to form a zero-delay closed loop of verification and regulation, which not only improves the training accuracy, but also reduces the deviation caused by individual differences through multi-link verification. This multi-effect collaborative closed loop design is difficult to achieve by a single verification method.

[0024] Further, the dynamic training module comprises a dynamic adjustment unit and a multi-spectrum adjustment unit; when the real-time state of the ciliary muscle is detected as a fatigue state, the target switching frequency is reduced to 8-12 times per minute, and the proportion of 500-550 nm green light is increased to 60% to promote relaxation; when the active state is detected, the target switching frequency is increased to 15-20 times per minute, and the proportion of 620-660 nm red light is increased to 50% to strengthen the training; the multi-spectrum adjustment unit is used to emit 430-680 nm specific light waves, wherein the green light adjusts the pupil diameter by alternating light and dark brightness, and synchronously trains the ciliary muscle and the pupil coordination, and the red light is used for light nutrition supplement to stimulate the acceleration of choroid blood circulation, which realizes training strengthening by the characteristics of light waves, and enhances the linkage of the visual system by pupil adjustment.

[0025] The dynamic adjustment unit of the scheme accurately adjusts the target switching frequency and the proportion of light waves according to the real-time state of the ciliary muscle, which can promote relaxation when tired and strengthen training when active, and ensures the pertinence and effectiveness of the training; the specific light waves emitted by the multi-spectrum adjustment unit not only realize the ciliary muscle and pupil coordination training and choroid blood circulation stimulation by the characteristics of green light and red light respectively, but also enhance the linkage of the visual system through pupil adjustment, which realizes the dual functions of training strengthening and linkage. For example, when children perform near vision training, the green light alternately bright and dark makes the pupil constantly contract and expand, and in this process, the ciliary muscle also adjusts accordingly, thereby strengthening the cooperation of the two, which can make the children's eye adjustment more smooth when quickly switching the distance of viewing objects, and it is not easy to appear temporary blurred vision, which is different from the effect of only focusing on single parameter adjustment before, and significantly improves the comprehensiveness and practicality of the training.

[0026] Further, the dynamic training module further comprises a contrast regulation unit, a mode switching unit and a scheme generation unit; the contrast regulation unit is used to dynamically reduce the contrast of the training picture in the training parameters to strengthen the adjustment stability of the ciliary muscle in complex visual environment; the mode switching unit supports automatic switching between object training and brain training, and switches to brain training for 1-2 minutes when the data fusion unit determines that the ciliary muscle is in a fatigue state for three times in succession; the scheme generation unit outputs the daily training time and light wave combination mode in combination with the age of the child, the degree of amblyopia and the historical data of the coordination acquisition module, realizes precise regulation of the training and recovery effect, and is linked with the dynamic adjustment unit, so that the generated scheme can be updated in real time according to the state of the ciliary muscle; the degree of amblyopia is obtained by matching the eye feature parameters with the stored amblyopia features.

[0027] The contrast regulation unit simulates a complex visual environment by dynamically reducing the contrast of the training picture, and strengthens the accommodation stability of the ciliary muscle under different light and clarity; the mode switching unit automatically switches to brain training when the ciliary muscle is continuously fatigued, which not only avoids damage to the ciliary muscle caused by overtraining, but also relaxes the eye muscles through short-term brain training, thereby achieving a dynamic balance between training and recovery; the scheme generation unit generates a personalized scheme in combination with the age of the child, the degree of amblyopia and historical data, and realizes real-time updating in linkage with the dynamic adjustment unit, wherein the degree of amblyopia is obtained by matching the eye feature parameters, thereby ensuring the adaptability of the scheme, and the scheme generation unit integrates and analyzes multi-dimensional data, so that the scheme has both pertinence and flexibility.

[0028] In particular, the automatic switching mechanism of the mode switching unit not only converts the training mode, but also constructs an active defense system for eye protection. For example, when the child continuously performs 20 minutes of object training, the data fusion unit determines that the ciliary muscle is in a fatigued state for three times in succession, and the mode switching unit immediately starts 1 minute of brain training. Through visual imagination training, the ciliary muscle is restored to vitality in a load-free state. This predictive protection combined with alternating training reduces the risk of ciliary muscle spasm compared to the traditional manual switching mode. This effect is particularly significant in the training scenario after the child has been learning for a long time at close range. For example, when the child performs visual function training after school, the scheme can timely sense eye fatigue and switch modes to avoid aggravating the eye burden due to continuous training.

[0029] Further, the training device includes a shell, a display component and a fastening component. The display component is arranged inside the shell and fixedly connected with the shell. The fastening component is fixedly connected with the outside of the shell and used for fixing the shell of the training device around the eye. The shell is provided with an orbital hole matched with the orbit. The outside of the shell around the orbital hole is provided with a pressure detection area corresponding to the skin surface of the lateral orbital extraocular muscle attachment point. A plurality of pressure sensors of the pressure acquisition unit are fixedly connected with the pressure detection area. The image acquisition unit is a binocular miniature camera integrated above the orbital hole inside the training device. The binocular miniature camera is provided with an infrared fill light. The inside of the shell is also fixedly connected with the display component. The display component is used for receiving the training parameters provided by the dynamic training module and displaying the training picture.

[0030] The pressure detection area around the orbital hole matched with the eye socket on the shell enables the pressure sensor to accurately correspond to the eye muscle attachment point, improves the accuracy of pressure deformation data collection, and provides a reliable basis for data fusion unit analysis; the dual micro cameras are matched with infrared fill light to ensure that the anterior segment images of the eyeball can be clearly collected in different light environments, ensuring the stability of the extraction of eye feature parameters; the display component is fixedly connected with the shell, can stably receive training parameters and display pictures, and is fixed on the eye with the fastening component to ensure the stability of the device position during training and reduce the interruption of training caused by shaking.

[0031] The adaptive design of the pressure detection area and the orbital hole not only accurately collects pressure data, but also maintains good contact between the sensor and the skin when the child is moving slightly, such as turning and shaking on the chair while sitting for a long time during home training, so that the training device does not cause the pressure sensor to deviate from the detection position due to shaking, ensuring the continuity of data collection and improving the flexibility of the device. At the same time, the integrated design of the display component, camera, pressure sensor, and shell makes the training device integrated, without the need for additional equipment, so that children can quickly start training without complex installation and debugging at home, and can also start training quickly without professional guidance. The coordinated work of each component improves the effective training ratio in the same training time, achieving better training results.

[0032] Further, the pressure collection unit is integrated with a contact sensing assembly for synchronously collecting intraocular pressure data; the image collection unit analyzes the corneal reflection image to obtain the tear film breakup time, forming a cooperative verification method for intraocular pressure and tear film state; the cooperative verification method uses the intraocular pressure fluctuation threshold and the tear film breakup time threshold as auxiliary verification conditions, and correlates and analyzes the pressure deformation data collected by the pressure sensor and the eye feature parameters obtained by the image collection unit; when the intraocular pressure exceeds the preset fluctuation range, the data fusion unit corrects the weight of the associated parameters in the mapping model, and the associated parameters are the weight coefficients in the mapping model for correlating the lens curvature and the extraocular muscle pressure; when the tear film breakup time is shorter than the preset breakup threshold, the multispectral adjustment unit adjusts the light wave ratio to improve the ocular surface state.

[0033] The scheme can realize the cooperative verification method by the pressure acquisition unit integrated with the contact sensing assembly to collect intraocular pressure data and the image acquisition unit to analyze the tear film breakup time. The pressure deformation data and the eye feature parameters can be associated and analyzed. When the intraocular pressure exceeds the preset fluctuation range, the data fusion unit corrects the weight of the associated parameters in the mapping model, improves the accuracy of the correlation between the lens curvature and the extraocular muscle pressure, and provides a more reliable basis for the accommodation calculation of the ciliary muscle. When the tear film breakup time is shorter than the preset threshold, the multispectral adjustment unit adjusts the light wave proportion to improve the ocular surface state and ensure the image acquisition clarity. The dynamic correction of the associated parameter weight and the improvement of the ocular surface state are linked to realize real-time response through the data fusion unit, which not only ensures the correlation accuracy of the pressure deformation data and the eye feature parameters, but also quickly optimizes the ocular surface environment through light wave adjustment to reduce the training interruption caused by the fluctuation of the ocular surface state. For example, when children train in a dry environment, the tear film breakup time is shortened to 3 seconds. After the multispectral adjustment unit increases the green light proportion, the tear film state is improved, and the data fusion unit automatically corrects the associated parameter weight due to the tear film change to avoid the false correlation of pressure and optical data under the influence of the tear film. Even if the children blink frequently, the training parameters can still be stably output by the scheme, which not only ensures the continuity of the training, but also improves the training accuracy. The real-time linkage adjustment of multi-dimensional data enables the training to be efficiently carried out in complex environments.

[0034] Further, the flexible Ag / AgCl electrode is fixedly connected to the tragus corresponding position outside the shell and the corneal limbus corresponding to the edge of the orbital hole of the training device, the sampling frequency of the flexible Ag / AgCl electrode is 250 Hz, and the corneal surface potential change is collected; when the ciliary muscle contracts, the epithelial ion pump activity is enhanced to make the corneal surface potential drift to the negative value by 0.5-1.2 mV; the data fusion unit associates and analyzes the corneal surface potential change data with the intraocular pressure data, the tear film breakup time and the pressure deformation data; when the linear deviation of the potential drift amplitude and the accommodation amount exceeds the preset potential accommodation deviation, the associated parameter weight of the mapping model is corrected synchronously, the associated parameter is the weight coefficient in the mapping model for correlating the lens curvature and the extraocular muscle pressure, and the dynamic training module controls the multispectral adjustment unit to adjust the red light proportion to enhance the ion pump activity.

[0035] The scheme collects the potential change of the corneal surface by setting flexible Ag / AgCl electrodes at the corresponding positions of the training device shell, and analyzes the data in association with intraocular pressure, tear film break-up time and pressure deformation. When the linear deviation of the potential drift amplitude and the accommodation amount exceeds the preset potential accommodation deviation, the mapping model correlation parameter weight is corrected synchronously and the red light proportion is adjusted. The contraction state of the ciliary muscle is accurately reflected by the potential change, which provides a reliable basis for the accommodation amount calculation and improves the accuracy of the data fusion unit analysis. The ciliary muscle function is improved by adjusting the red light proportion to enhance the activity of the ion pump. The variable of the potential change of the corneal surface can realize the dual functions of verifying the ciliary muscle state and assisting the optimization of the training intervention. For example, when the epithelial ion pump activity of children decreases due to eye fatigue, the linear deviation of the potential drift amplitude and the accommodation amount exceeds the preset value. The mapping model correlation parameter weight is corrected to ensure the accuracy of the calculation, and the red light proportion is increased to enhance the activity of the ion pump, so as to promote the recovery of the ciliary muscle contraction function. Even if the children do not show obvious fatigue symptoms during the training, the potential change can also be used to intervene in advance to avoid further decline of the accommodation function, so that the training is more forward-looking and targeted.

[0036] Further, the data fusion unit performs four-dimensional correlation analysis on the corneal surface potential change data, intraocular pressure data, tear film break-up time and pressure deformation data, and establishes a dynamic coupling model combined with potential, intraocular pressure, tear film and pressure. When the linear deviation of the potential drift amplitude and the accommodation amount exceeds the preset potential accommodation deviation, and the intraocular pressure exceeds the fluctuation range or the tear film break-up time is shorter than the preset break-up threshold, the mapping model correlation parameter weight is corrected synchronously, the red light and green light proportions of the multi-spectral adjustment unit are adjusted, and the contrast regulation unit operation is suspended.

[0037] The scheme performs four-dimensional correlation analysis on the corneal surface potential change, intraocular pressure, tear film break-up time and pressure deformation data by the data fusion unit, and establishes a dynamic coupling model. When a specific condition is met, multiple operations are triggered synchronously, which improves the comprehensiveness and accuracy of data correlation, provides more reliable basis for the correction of the mapping model correlation parameter weight, and ensures the effectiveness and comfort of the training by adjusting the light proportion of the multi-spectral adjustment unit and suspending the contrast regulation. The dynamic coupling model realizes cross-dimensional data mutual verification. For example, when children train in a dry environment, the tear film break-up time is shortened, and the potential drift is abnormal. The dynamic coupling model can determine whether the potential drift is caused by ciliary muscle contraction by using tear film data, avoiding misjudgment of a single data. If it is determined that it is a tear film problem, the green light proportion is adjusted to improve the ocular surface state, and the contrast regulation is suspended to prevent further decline of visual clarity. The adjustment of the red light proportion can enhance the activity of the ion pump and promote the recovery of the ciliary muscle function. This multi-data driven cooperative intervention method enables the training to be accurate and efficient even in complex eye conditions, significantly improving the adaptability and training effect of the scheme.

[0038] Further, the dynamic coupling model introduces a dynamic baseline calibration coefficient K, K is calculated based on the initial drift value of corneal surface potential, the baseline value of intraocular pressure, and the tear film breakage reference time according to the formula K = 0.3 x potential baseline + 0.2 x eye pressure baseline + 0.5 x tear film baseline, and is automatically updated every 30 seconds; the K value is embedded in the reliability score calculation of the dynamic coupling model, when the K fluctuation is less than or equal to 5%, the score weight is distributed in the original proportion; when the K fluctuation is more than 5%, the score weight of intraocular pressure and tear film data is increased by 20%; at the same time, the K value is linked with the multi-spectral adjustment unit, when the K fluctuation is more than 5%, in addition to the red light and green light adjustment, the light intensity is linearly increased with the deviation of K value, and the error threshold is triggered to shrink to ±0.2Δ / D by the data fusion unit.

[0039] By introducing the dynamic baseline calibration coefficient K in the dynamic coupling model, dynamic calibration and cross-unit collaborative intervention of multi-dimensional data are realized, K value is automatically updated every 30 seconds, which can adapt to the changes of children's eye state in real time, when K fluctuation is less than or equal to 5%, the original score weight distribution is maintained to ensure the stability of data correlation analysis; when K fluctuation is more than 5%, the score weight of intraocular pressure and tear film data is increased and the multi-spectral adjustment unit is linked to enhance the light supplement and shrink the error threshold, which not only improves the influence of key data in abnormal state, but also ensures the image acquisition quality through light supplement, and the error threshold shrinkage improves the strictness of data verification. Among them, K value as a unified variable can realize multiple functions of data reference calibration, score weight adjustment, light intensity control and error threshold adjustment, for example, when children train at home in the evening, the indoor light gradually darkens, which causes the tear film data to fluctuate and the K value to fluctuate by 6%, the tear film data score weight is immediately increased, the light intensity is increased by 12%, and the error threshold is shrunk to ±0.2Δ / D, which avoids the influence of insufficient light on image acquisition, ensures the data reliability through more stringent error standards, and at the same time, the red light and green light adjustment is still normal to improve the ocular surface state and ciliary muscle function, so that the training accuracy in complex environment is not affected, and the effective training data proportion is improved to a high level. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 It is a schematic diagram of the function module of the children's visual function training system based on ciliary muscle regulation in the embodiment of the application.

[0041] Figure 2 It is a related flowchart when the data fusion unit verifies the accuracy of AC / A constant in the embodiment of the application. DETAILED DESCRIPTION

[0042] The concept and technical effects of the present application will be described clearly and completely in combination with embodiments, so as to fully understand the purposes, features and effects of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, but not all the embodiments, and other embodiments obtained by those skilled in the art without creative labor based on the embodiments of the present application all belong to the protection scope of the present application:

[0043] As shown in the figure, the child visual function training system based on ciliary muscle accommodation comprises: Figure 1

[0044] The cooperative acquisition module comprises an image acquisition unit, a pressure acquisition unit and a data fusion unit.

[0045] The image acquisition unit acquires the anterior segment image of the eyeball in real time, and identifies the eye feature parameters in the anterior segment image of the eyeball.

[0046] The pressure acquisition unit comprises a pressure sensor arranged on the skin surface corresponding to the attachment point of the extraocular muscle on the lateral side of the orbit, and acquires the pressure deformation data of the extraocular muscle caused by the convergence movement of the eyeball in real time.

[0047] The data fusion unit obtains the eye feature parameters and the pressure deformation data acquired at the same time, and performs correlation analysis, establishes a mapping model through the pressure deformation data change and the lens curvature change, indirectly calculates the lens power change data as first data from the eye feature parameters through a thin lens refractive power conversion algorithm, inputs the pressure deformation data into the mapping model to calculate the real-time state of the ciliary muscle, and calculates the lens power change data in the eye feature parameters as second data according to the real-time state.

[0048] The data fusion unit compares the first data with the second data, if the comparison result falls within a preset error threshold, the real-time state is processed into the real-time accommodation amount of the ciliary muscle according to the acquisition time, if the comparison result exceeds the error threshold range, the image acquisition unit and the pressure acquisition unit acquire the eye feature parameters and the pressure deformation data again, and calculate and compare the first data and the second data.

[0049] The dynamic training module obtains the individual data of the child and the historical data of the cooperative acquisition module, formulates training parameters according to a preset training period, and dynamically adjusts the training parameters according to the real-time accommodation amount of the ciliary muscle; the real-time state comprises a fatigue state and an active state, and the training parameters comprise a training duration and a light wave combination; the training duration, the target frequency, the proportion of different types of light in the light wave combination and the contrast are dynamically regulated according to the real-time state; and the training mode in the regulation parameters is regulated according to the continuous real-time state.

[0050] ​Specifically, the thin lens power conversion algorithm realizes the calculation of the lens power change through the three steps of "parameter collection - modeling calculation - dynamic output". First, the image acquisition unit non-invasively captures the anterior segment image of the eyeball, extracts the morphological parameters such as the anterior surface radius of curvature, the posterior surface radius of curvature, and the central thickness d, and simultaneously calls the built-in children's exclusive medium refractive index database (including lens refractive index, aqueous and vitreous refractive index) in the data fusion unit; Then, based on the "thin lens assumption" (for example, the thickness of the child's lens is 3-4mm, which is much smaller than the focal length of 10-20mm in the accommodation state), the modeling is simplified. First, the marker formula (thin lens power calculation formula) is used to calculate the basic value of the lens optical power, and then combined with the axial length L measured by the binocular parallax algorithm, the actual refractive power is obtained by compensating the difference in the development of the child's eye axis through the correction formula; Finally, it is synchronized with the sampling frequency of the collaborative acquisition module, outputs the refractive power change data, and filters abnormal values such as "continuous 3 times sampling deviation > ±0.1D" (for example: the minimum single accommodation amplitude of the ciliary muscle of children aged 6-12 is about 0.2D, and ±0.1D is less than the minimum physiological accommodation, to avoid filtering the real accommodation signal; and the refractive power fluctuation caused by blinking, eye shaking and other interference has instantaneousness, non-continuity, and the probability of continuous 3 times exceeding ±0.1D is extremely low, while the real ciliary muscle adjustment fluctuation, such as fatigue spasm, has persistence), to ensure the contrast accuracy of the input data fusion unit.

[0051] The marker formula is: F=(n-1)×[(1 / R1)-(1 / R2)+(n-1)×d / (n×R1×R2)]; wherein F is the lens power (unit: D); n is the lens refractive index (for example: children take 1.41-1.42, and then adjust with age); R1 is the lens anterior surface radius of curvature; R2 is the lens posterior surface radius of curvature; d is the central thickness of the lens).

[0052] From the optical theory, the essence of dioptric power F is the deflection ability of the lens to the light, which follows the two laws of "focal length is inversely proportional to dioptric power" and "light deflection superposition". The total dioptric power of a thin lens is formed by the deflection superposition of the front and back surfaces, and the marker formula is derived based on this feature, which can directly establish the correlation between the lens shape parameters and the dioptric power. For example, when the ciliary muscle contracts to make the lens more convex (R1 decreases), F increases, which is consistent with the physiological regulation mechanism of "lens dioptric power increases when looking at near". From the physiological adaptation of the eyeball, the thickness of the child's lens naturally meets the thin lens assumption, which is only 1 / 2-1 / 5 of the focal length (in the field of child eye physiology and ophthalmic optics, the ratio of the thickness of the child's lens to the focal length in the adjusted state is only 1 / 2-1 / 5, which is a conclusion based on the measurement of child lens physiological parameters, optical focal length calculation and industry consensus, and is common knowledge in the professional field), and there is no adult nuclear sclerosis phenomenon, the refractive index distribution is approximately uniform, without complex gradient refractive index modeling, which further verifies the rationality of the thin lens simplification, and ensures that the algorithm calculation logic conforms to the physical law and is consistent with the structural characteristics of the child's eye.

[0053] The data fusion unit is further configured to verify the accuracy of the AC / A constant according to the synchronous acquisition of the pressure deformation data and the eye feature parameters, wherein the AC / A constant is the ratio of accommodative convergence to accommodation, with a unit of "Δ / D", which is a core physiological parameter of ophthalmic optics, representing the amount of binocular convergence function when the eye produces 1 diopter (D) of accommodation, and is used to reflect the coordination between accommodation and convergence (the stronger the accommodation, the more obvious the convergence of the binocular visual axis). When the individual error of AC / A is less than or equal to ±0.3Δ / D (many ophthalmic studies have pointed out that an AC / A measurement error of more than ±0.5Δ / D will significantly affect the accuracy of diagnosis of strabismus, phoria, etc., and the present scheme sets a more stringent ±0.3Δ / D, which is higher than the conventional diagnostic accuracy requirement, to ensure that the training intervention is based on a reliable physiological model), and the error of the mapping model is less than or equal to ±0.25D after being calibrated by UBM (ultrasound biomicroscope) and open field optometry, the following conditions are met: Figure 2As shown, first, the objective accommodation response is recorded by an infrared video optometry instrument or an open field automatic refractometer as the accommodation amount gold standard, and the instantaneous convergence amount is measured by prism cover method or synoptophore as the convergence amount gold standard, to ensure that the AC / A obtained from the accommodation amount gold standard and the convergence amount (the total convergence function amount, which refers to the physiological response strength when the visual axes of both eyes converge from infinity to a certain near target, and the essence is the visual axis deflection amount caused by the contraction of the extraocular muscles, measured in prism degrees Δ) gold standard has a difference of ≤±0.3Δ / D from the simultaneously collected pressure deformation data and eye feature parameters; then, the accommodation stimulus is applied at different viewing distances and repeated multiple times, and the stimulus response curve is drawn, if the linear fitting R² of the stimulus response curve is <0.9 (R² is the goodness of fit of the regression model, in normal cases, the accommodation and convergence of the human eye are in an approximately linear relationship, especially in children, the AC / A ratio is stable, and the linear fitting is good. In visual science and strabismus / phoria evaluation, R²≥0.9 is the general statistical threshold for judging whether the “stimulus-response relationship is stable”; if R²<0.9, it means that there is a nonlinear accommodation response, for example, accommodation spasm, excessive / inadequate convergence, and the linear model will mislead the training parameter setting) or the intercept is >0.5Δ (Δ is the unit symbol of prism degree), then segmented fitting or quadratic curve correction is performed; the far and near distance phoria amounts are measured by covering and uncovering methods, and for those with a phoria amount >8Δ, a separate model is established, and for those with a refractive power difference >1D, the left and right eye AC / A values are separately calibrated and then averaged, and the difference between two consecutive days is ensured to be <±0.2Δ / D (many studies have shown that the normal children or adults AC / A naturally fluctuates within a range of ≤±0.2Δ / D in a short period of time (such as 24-48 hours); if the fluctuation range exceeds this range, it may indicate that there is accommodation spasm, fatigue, measurement error or strabismus fluctuation).

[0054] Specifically, when the data fusion unit starts the AC / A constant precision verification, it first calls the image acquisition unit and the pressure acquisition unit to simultaneously collect 3 sets of eye feature parameters and pressure deformation data, with an interval of 2 seconds for each set, and automatically removes abnormal data generated due to blinking or violent eye movement during the interval. Under the premise that the AC / A individual error is ≤±0.3Δ / D and the mapping model has an error of ≤±0.25D after calibration, the infrared video optometry instrument is controlled to complete the accommodation response recording within 5 seconds, and the prism cover method is triggered to automatically switch the prism degree number, with 1 collection of convergence amount per second, and the average value of 10 seconds is taken as the convergence amount gold standard (assuming that 6-8 switching periods are needed to eliminate single-point errors, the head position deviation is <2° and the accommodation drift is <0.1D within 10 seconds, and the convergence amount standard error is reduced from ±0.3Δ to ±0.1Δ by taking the average of 10 points, which meets the total budget of ±0.3Δ / D), to ensure that the AC / A obtained from the accommodation amount and convergence amount gold standards has a difference of ≤±0.3Δ / D from the simultaneously collected data.

[0055] Different viewing distances were set as 20 cm (corresponding to 5D), 33 cm (approximately equal to 3D), 50 cm (corresponding to 2D), and 100 cm (corresponding to 1D). Under each viewing distance, 1D, 2D, 3D, and 5D adjustment stimuli were applied, respectively. Each stimulus level was repeated 5 times. During the collection process, the picture contrast was fixed at 50% by the contrast regulation unit to reduce the interference of environmental light on data stability.

[0056] The FCC regulation on electromagnetic exposure (RF Exposure) of mobile devices specifies that 20 cm is the common test distance, especially for devices worn on the body / near the head (such as mobile phones, Bluetooth earphones, etc.). The "Guidelines for Prevention and Control of Myopia in Children and Adolescents (2021)" clearly states that the eyes should be one foot (approximately 33 cm) away from the book during reading and writing. ISO 9241-5:2022 (Ergonomics of human-system interaction — Part 5: Workstation layout and postural requirements) recommends a viewing distance range: usually 50 cm to 100 cm (adjust according to screen size, resolution, and user vision). According to the adjustment requirement (D) = 1 / viewing distance (m), the adjustment requirements corresponding to 20 cm, 33 cm, 50 cm, and 100 cm can be obtained.

[0057] The eye movement frequency of children aged 6-12 years is higher than that of adults (about 3-5 times per second). Single collection (such as 300 ms / frame) may capture "blinking moments" (resulting in missing eye feature parameters) and "eye deviation from the target" (resulting in refractive power measurement deviation). This type of error can reach ±0.15D per single collection. By processing the 5 collection data using the "Grubbs criterion" (anomaly detection algorithm), 1-2 abnormal values (such as deviation values caused by blinking) can be effectively removed, and the mean error of the remaining 3-4 valid data can be reduced to within ±0.05D. Fixing the picture contrast at 50% in the central "adjustment platform area" is the most economical integer value with the smallest error and lowest individual difference.

[0058] When measuring the amount of phoria, the multi-spectral adjustment unit first emits white light to calibrate eye position, then performs covering and uncovering operations, with a 3-second interval between each operation. The mean value is taken after 3 consecutive measurements. For those with a phoria of >8Δ, a separate model is established. For those with a refractive power difference of >1D, the left and right eye AC / As are calibrated separately, and the average value is taken. The difference between two AC / As is ensured to be <±0.2Δ / D. For individuals who need to be modeled separately, the dynamic training module temporarily saves their historical training parameters. After the AC / A verification is completed, the new parameters are fused with the historical data to generate a transitional training plan.

[0059] In one specific embodiment of the present scheme, the (data fusion unit) greatly improves the accuracy and reliability of AC / A constant verification through multi-link precise operation, providing a solid data foundation for subsequent training. Synchronous acquisition and rejection of abnormal data ensure the quality of the initial data, avoiding the interference of abnormal values with the verification results; repeated collection of multiple viewing distances and multiple stimulus levels and fixed contrast reduce the influence of environmental factors on the data, making the stimulus response curve more realistic; with the help of the multi-spectral adjustment unit to calibrate the eye position and multiple measurements to take the average, the accuracy of the data such as the phoria is improved; special individuals are modeled separately and historical parameters are fused to generate a transition plan, ensuring the continuity and adaptability of the training.

[0060] For example, a 7-year-old child has moderate phoria (phoria 9Δ, and when the latent phoria exceeds 9Δ in clinical data, symptoms such as visual fatigue, reading difficulty, or intermittent diplopia may occur), in the verification, first through the multi-spectral adjustment unit to issue white light to accurately calibrate the eye position, then perform covering and uncovering operations and take the average of 3 consecutive measurements, accurately obtain the phoria of the child, and then model it separately. When collecting data at different viewing distances, the contrast control unit fixes the picture contrast at 50%, avoiding the interference of outdoor light changes on the collection, making the AC / A constant more accurate. After verification is completed, the dynamic training module fuses its historical training parameters to generate a transition plan, so that the child will not be uncomfortable due to too large changes in the initial application of new parameters, ensuring the training effect and improving the child's cooperation, which is difficult to achieve by traditional systems that rely only on single data collection and general models.

[0061] The dynamic training module includes a dynamic adjustment unit and a multi-spectral adjustment unit. When the dynamic adjustment unit detects that the real-time state of the ciliary muscle is a fatigue state, the target switching frequency is reduced to 8-12 times / minute, and the proportion of 500-550nm green light is increased to 60% to promote relaxation; when the active state is detected, the target switching frequency is increased to 15-20 times / minute, and the proportion of 620-660nm red light is increased to 50% to strengthen training. The multi-spectral adjustment unit is used to emit 430-680nm specific light waves, among which green light adjusts the pupil diameter through light and dark brightness alternation to train the ciliary muscle and pupil coordination, and red light is used for light nutrition supplementation to stimulate choroid blood circulation acceleration. This unit not only achieves training enhancement through light wave characteristics, but also enhances the linkage of the visual system through pupil adjustment.

[0062] Specifically, the dynamic adjustment unit has a built-in state recognition method, which receives the real-time ciliary muscle adjustment amount output by the data fusion unit and completes state determination within 1 second; when switching target frequency, a step transition is adopted (such as increasing or decreasing 2 times / minute each time, with an interval of 5 seconds), to avoid sudden changes in frequency causing eye discomfort. The green light bright-dark alternating cycle of the multi-spectrum adjustment unit is linked to the target switching frequency (such as 8 times / minute corresponding to 4 seconds of bright-dark cycle), and when alternating, the brightness gradually changes from 300 lux to 800 lux and then falls back, triggering the image acquisition unit to record the pupil diameter change at the same time, which is used as the evaluation basis for the effectiveness of the coordinated training. The intensity of red light is dynamically adjusted according to the choroid blood circulation simulation data (based on pressure deformation data for indirect calculation), and the intensity is increased by 10%-20% when the blood circulation is slow. The dynamic adjustment unit and the data fusion unit communicate in real time, feed back the light wave action data to the mapping model, optimize the accuracy of ciliary muscle state determination, and realize the closed-loop enhancement of training, evaluation, and model optimization combination.

[0063] For example, in the training of an 8-year-old child with amblyopia, the dynamic adjustment unit detects an active state, and the target frequency is stepped up from 10 times / minute to 15 times / minute, and the red light proportion is simultaneously increased to 50%. The multi-spectrum adjustment unit alternates green light brightness at a frequency of 15 times / minute, and the image acquisition unit records the synchronicity of pupil diameter changes with green light. The data fusion unit adjusts the mapping model accordingly. When the child continuously fixates on the training target, causing the choroid blood circulation to slow down, the red light intensity is automatically increased by 15%, which not only strengthens the training effect but also speeds up the eye metabolism, shortening the time to meet the training standard in a single training compared to the traditional fixed light wave scheme.

[0064] The dynamic training module further comprises a contrast regulation unit, a mode switching unit and a scheme generation unit. The contrast regulation unit is configured to dynamically reduce the contrast of the training picture in the training parameters to strengthen the accommodation stability of the ciliary muscle in a complex visual environment. The mode switching unit supports automatic switching between physical training and brain training. Physical training drives eye physiological responses through the content viewed by the child during the training process (set by the administrator), the definition of the viewed content, and the dynamic changes of the viewed content; for example, using flip cards to train the flexibility of the ciliary muscle, using convergence / divergence balls to guide the convergence / divergence of the visual axes of the two eyes to improve the convergence function, and using moving light points to train the smoothness of the eye movement. Brain training is to let the child gaze at a preset neutral gray screen (set by the administrator) and perform psychological imagery tasks such as “imagining looking far and near”, “imagining tracking light points”, “imagining flip cards” and the like under the guidance of a preset voice (set by the administrator). In this process, the dynamic training module collects the brain activity state of the child in real time through electroencephalogram signals, and simultaneously implements dual intervention of red light ion pump activation and alpha wave band neural feedback. When the data fusion unit determines that the ciliary muscle is in a fatigued state for three consecutive times, the brain training is switched to for 1-2 minutes. The scheme generation unit outputs the daily training time and light wave combination mode in combination with the age of the child, the degree of amblyopia and the historical data of the collaborative acquisition module, realizes precise regulation of the training and recovery effect, and is linked with the dynamic adjustment unit, so that the generated scheme can be updated in real time according to the state of the ciliary muscle; and the degree of amblyopia is obtained by matching the eye feature parameters with the stored amblyopia features.

[0065] More specifically, the contrast regulation unit reduces the picture contrast in stages by 5% gradient (such as 50% initially and 20% at the lowest), each stage lasts for 2 minutes, and the eye muscle pressure data of the pressure acquisition unit is received synchronously, and the contrast is paused and maintained at the current level when the pressure increase exceeds 10% (when the accommodation pressure increase exceeds 10%, the ciliary muscle may be over-tensioned, and maintaining the current contrast can avoid compensation disorders of visual function caused by parameter mutation). The mode switching unit starts green light warning (such as 500-550nm green light proportion temporarily increased to 70%) 10 seconds before switching. The brain training stage closes the target switching function of the dynamic adjustment unit, and only retains the basic red light irradiation (such as 620-660nm proportion 30%) of the multi-spectrum adjustment unit. The scheme generation unit has an amblyopia feature database, determines the degree of amblyopia through the matching degree of the eye feature parameters of the image acquisition unit (such as ≥85% to determine the corresponding level), divides the daily training time by age (such as 20 minutes / day for 6-8 years old and 30 minutes / day for 9-12 years old), and is linked with the real-time state of the dynamic adjustment unit. In the active state, the single training segment is extended by 1-2 minutes, and in the fatigued state, a 30-second green light relaxation segment is inserted. The scheme generation unit and the data fusion unit share historical data, so that the scheme update optimizes the error threshold of the mapping model, improves the cooperative accuracy of state determination and training scheme.

[0066] In one specific embodiment of the present scheme, the eye feature matching degree of a 10-year-old child with moderate amblyopia is 92%. During training, the contrast control unit is reduced from 50% to 30%, and the eye muscle pressure increases by 15% during this period. The contrast is immediately suspended; after the data fusion unit determines the fatigue state for three consecutive times, the mode switching unit first warns with a high proportion of green light, and then switches to brain training for 1.5 minutes, during which only the basic red light irradiation is retained; the scheme generation unit combines its 9-year-old age file and historical data to initially set 30 minutes of daily training. Due to the more active state, the single training segment is automatically extended to 3 minutes, and a green light relaxation segment is inserted between two active states, which not only strengthens the ability to adjust to complex environments but also avoids excessive fatigue through mode switching, and the training efficiency is significantly improved compared to the fixed scheme.

[0067] The training device further comprises a shell, a display component, and a fastening component. The display component is arranged inside the shell and fixedly connected to the shell. The fastening component is fixedly connected to the outside of the shell and used to fix the shell of the training device around the eye. The shell is provided with an orbital hole matched with the eye socket. The outside of the shell around the orbital hole is provided with a pressure detection area corresponding to the skin surface of the eye muscle attachment point outside the eye socket. A plurality of pressure sensors of the pressure acquisition unit are fixedly connected to the pressure detection area. The image acquisition unit is a binocular miniature camera integrated above the orbital hole inside the training device. The binocular miniature camera is provided with an infrared fill light. The inside of the shell is further fixedly connected with the display component. The display component is used to display the training parameters and training pictures provided by the dynamic training module.

[0068] A dry electrode electroencephalogram sensor is embedded and fixed in the area where the shell is attached to the forehead to collect the electroencephalogram signals of the frontal lobe and motor cortex in real time. The data fusion unit filters and amplifies the electroencephalogram signals and identifies the frequency bands such as the alpha band (e.g. captures 8-13 Hz as the alpha band). When it is identified that the child enters a stable visual image state (e.g. alpha band power > 5 μV² / Hz, μV² / Hz is the power spectral density unit, which is how many microvolts of power contained in the frequency bandwidth of 1 Hz in the electroencephalogram signal), the infrared fill light emits light with a preset fixed pulse (set by the administrator) to help the ciliary muscle relax and adjust the tension to zero. At the same time, the data fusion unit analyzes the alpha band power in real time and converts it into dynamic feedback of the display component.

[0069] Specifically, the training device uses a liquid lens (response time < 50 ms) or a miniature electric focusing lens with a diameter of 5-8 mm. The curvature of the lens is controlled by voltage to achieve continuous adjustment of 1D-5D diopter power (accuracy up to ±0.1D), meeting the viewing distance simulation requirements from 20 cm (5D) to 100 cm (1D).

[0070] Specifically, the training device shell is made of food-grade silicone material with a thickness of 2 mm. The food-grade silicone material is soft and skin-friendly, suitable for the delicate skin of children. The 2 mm thickness can ensure structural support while being flexible, conforming to the eye contour of different children, reducing the foreign body sensation when wearing, improving comfort, and reducing children's resistance.

[0071] The edge of the orbital hole is rounded by 0.5 mm to avoid damage to the skin around the eye caused by sharp edges rubbing against the skin. This further improves the safety of wearing, especially for active and lively children.

[0072] The pressure detection area is a 5 mm wide ring-shaped protrusion around the hole, with 8 flexible pressure sensors (generally, the flexible pressure sensor diameter can be selected as 3 mm, and special diameter flexible pressure sensors can also be customized according to the specific situation of children) embedded, with 2 for the medial rectus muscle, 2 for the lateral rectus muscle, 2 for the superior rectus muscle, and 2 for the inferior rectus muscle. The ring-shaped protrusion can enhance the adhesion of the sensor to the skin, and the symmetrical distribution can comprehensively capture the pressure deformation of each extraocular muscle; making the pressure data collection more comprehensive and accurate, providing a reliable basis for data fusion unit analysis, and reducing errors compared to single-point collection. The breathable non-woven fabric layer on the contact side of the pressure sensor and the skin is made of bamboo fiber material, and each piece can be replaced. Bamboo fiber is breathable and antibacterial, and the replaceable design is convenient for cleaning and hygiene, and can also reduce skin heat and allergy caused by long-term wearing, while facilitating maintenance.

[0073] The dual-eye miniature camera has a resolution of 12 million pixels, the lens axis has a 15° angle with the anterior segment of the eyeball, and the infrared fill light (850nm) has 3 brightness levels (other levels can also be customized according to the specific situation of children), which is automatically adjusted by the image acquisition unit according to the ambient light intensity (≤300 lux, high level is turned on). 12 million pixels and 15° angle can clearly capture the details of the anterior segment of the eyeball, and 3 brightness levels can adapt to different environmental light, allowing stable collection of high-quality images in bright and dark environments, effectively extracting eye feature parameters, and improving image efficiency.

[0074] The display component can be a 2.5-inch OLED screen with a refresh rate of 60Hz, or other screens with higher refresh rates and better display effects can be replaced according to the specific requirements of parents. The display component is electrically connected to the dynamic training module. The OLED screen has bright colors, fast response, and low delay, ensuring that the training picture and parameter adjustment are synchronized, allowing children to watch the training picture more clearly and smoothly, improving training concentration, and reducing training deviation caused by delay.

[0075] The fastening component can be an elastic headband made of nylon material, or other flexible materials according to the needs of parents. The length of the elastic headband can be adjusted to 40-60 cm. The pressure feedback module at the connection between the shell has a sampling frequency of 10 Hz. When the pressure exceeds 2 kPa, the display brightness is reduced by 10%, and the multi-spectrum adjustment unit increases the proportion of green light of 500-550 nm by 5%. (kPa is the abbreviation of "kilopascal", which is a unit of pressure. 1 mmHg is the pressure generated by a mercury column 1 mm high under standard gravity. 2 kPa is about 15 mmHg, which corresponds to the peak pressure of the extraocular muscle. When it exceeds, it means that the headband is too tight or the child is tense and frowns. Immediately reduce the brightness by 10% and add 5% green light to remove local pressure and relax the tension, avoiding the vicious cycle of "wearing tighter and more tired". The adjustable headband can adapt to different head circumferences. The pressure feedback can avoid wearing too tight. While ensuring the stability of the device, the brightness is reduced and the proportion of green light is increased to double adjust, relieve eye discomfort, and enable children to complete the training in a comfortable state, and improve the training efficiency with the dynamic training module.

[0076] In one specific embodiment of the present scheme, when a 6-year-old child with refractive power disparity wears the training device, the silicone shell is attached to the eye socket, and the 8 sensors in the pressure detection area accurately correspond to the attachment points of the extraocular muscles. When the ambient light is dim, the infrared fill light automatically adjusts to high gear, and the binocular camera clearly captures the anterior segment image of the eyeball. The headband is adjusted to the appropriate length, and the pressure feedback module monitors in real time. When the child frowns due to tension, the wearing pressure reaches 2.2 kPa, the display brightness is reduced by 10%, and the proportion of green light is increased. This not only relieves eye discomfort, but also ensures clear training images, and improves the effective duration of single training with parameter adjustment of the dynamic training module.

[0077] Specifically, the pressure acquisition unit is integrated with a contact sensing component for synchronous acquisition of intraocular pressure data; the image acquisition unit analyzes the corneal reflection image to obtain the tear film breakup time, forming a cooperative verification method for intraocular pressure and tear film state. The cooperative verification method takes the intraocular pressure fluctuation threshold and the tear film breakup time threshold as auxiliary verification conditions, and correlates and analyzes the pressure deformation data collected by the pressure sensor and the eye feature parameters obtained by the image acquisition unit. When the intraocular pressure exceeds the preset fluctuation range, the data fusion unit corrects the weight of the associated parameters in the mapping model, and the associated parameters are the weight coefficients in the mapping model for correlating the lens curvature and the extraocular muscle pressure; when the tear film breakup time is shorter than the preset breakup threshold, the multi-spectrum adjustment unit adjusts the light wave proportion to improve the ocular surface state.

[0078] Specifically, the contact sensing component is a micro-tonometer with an accuracy of ±0.5 mmHg, integrated next to the corresponding sensor of the superior rectus muscle in the pressure detection area, and triggered synchronously with the pressure sensor at a sampling frequency of 5 Hz. When the image acquisition unit analyzes the tear film breakup time, it takes 3 frames of corneal reflection images every 0.5 seconds, calculates the breakup time through a gray scale gradient algorithm, and the error is ≤0.3 seconds.

[0079] The initial values of the correlation parameters are set as follows: the influence weight of extraocular muscle pressure is 0.5, the response coefficient of lens curvature is 1.2Δ / D / kPa, and the intraocular pressure reference value is 15 mmHg. When the intraocular pressure exceeds the range of 12-18 mmHg, for every 1 mmHg deviation, the intraocular pressure correction factor is adjusted by ±0.05, and the output value of the mapping model is corrected synchronously. When the tear film breakup time is <5 seconds, in addition to adjusting the pressure influence weight and the curvature response coefficient, the multi-spectral adjustment unit increases the proportion of green light from 30% to 50% in the 500-550 nm range, and re-detects the tear film state after 2 minutes. At this time, the dynamic training module pauses the gradient contrast reduction operation of the contrast adjustment unit.

[0080] The above specific content is linked with the scheme generation unit of the dynamic training module. When the correlation parameters are corrected for 3 consecutive times, the scheme generation unit automatically shortens the single training segment by 1 minute and prolongs the brain training interval, effectively reducing the visual load accumulation rate of children during training.

[0081] In the above content, the synchronous triggering and collection of the micro-tonometer and the pressure sensor is not simply parallel data acquisition, but through the time anchor point of 5 Hz sampling frequency, the intraocular pressure fluctuation and the deformation of extraocular muscle pressure form a transient check of the mechanical correlation between the intraocular and extraocular. This millisecond-level time alignment mechanical coupling analysis breaks through the traditional cognition that intraocular pressure and extraocular muscle pressure are independent variables. For example, when children rapidly rotate their eyeballs, this scheme can distinguish between pressure changes caused by extraocular muscle contraction and passive fluctuations in intraocular pressure, avoiding misjudgment of the adjustment amount by the mapping model.

[0082] Secondly, the coordination of multi-spectral adjustment triggered by tear film breakup time and the suspension of contrast adjustment exceeds the conventional logic of simply improving the ocular surface state. When the proportion of green light is increased to 50%, it not only promotes tear secretion through photochemical action, but also reduces the accommodation load of the ciliary muscle due to the wavelength characteristics of green light. The suspension of gradient contrast reduction avoids the sudden drop in visual clarity when the tear film is unstable, forming a double protection of ocular surface repair and accommodation load buffering.

[0083] Finally, the dynamic adjustment of the training scheme is triggered by the continuous correlation correction of the parameters, realizing the predictive intervention of data anomalies and scheme adaptation. The scheme generation unit is not only adjusted according to the real-time state, but also predicts the potential cumulative risk of load through the parameter correction frequency, actively shortens the training section and prolongs the brain training interval. This forward-looking regulation based on data trends breaks the traditional mode of intervention after fatigue occurs, improving the control accuracy of visual load.

[0084] Specifically, the flexible Ag / AgCl electrode is fixedly connected at the tragus corresponding position outside the shell of the training device and the corneal limbus corresponding to the edge of the orbital hole, the electrode area is ≤10mm² and the contact resistance is ≤5kΩ; the flexible Ag / AgCl electrode has a sampling frequency of 250Hz (250Hz is only used for "slow change" ciliary muscle electric potential index, and high-frequency ERP is not used), which is used to collect the potential change of the corneal surface.

[0085] When the ciliary muscle contracts, the activity of epithelial ion pump is enhanced, which makes the potential of the corneal surface drift to the negative value by 0.5-1.2mV.

[0086] The data fusion unit analyzes the corneal surface potential change data in association with intraocular pressure data, tear film break-up time and pressure deformation data. When the linear deviation of potential drift amplitude and accommodation amount exceeds the preset potential accommodation deviation, the correlation parameter weight of the mapping model is corrected synchronously, the correlation parameter is the weight coefficient in the mapping model for correlating the lens curvature and the extraocular muscle pressure, and the dynamic training module controls the multi-spectral adjustment unit to adjust the red light proportion to enhance the ion pump activity.

[0087] More specifically, the flexible Ag / AgCl electrode adopts a 0.1mm ultra-thin base, the electrode at the tragus is in the form of a 5mm diameter circle, and the electrode at the corneal limbus is in the form of a 10mm arc, which respectively matches the contact area requirements of the two anatomical structures to ensure the stability of the potential signal collection. The flexible Ag / AgCl electrodes at the tragus and the corneal limbus are fixed by medical double-sided tape, and the contact resistance with the skin is ≤5kΩ, which ensures the low-loss transmission of weak potential signals.

[0088] The preset potential adjustment deviation is 5%, and the associated parameter correction rule is: for every 1% of the potential drift deviation, the extraocular muscle pressure influence weight is ±0.02, and the lens curvature response coefficient is ±0.1Δ / D / kPa. For every 1 kiloPascal (kPa) change in extraocular muscle pressure, the change in the "adjustment set and adjustment ratio (AC / A constant)" associated with the lens curvature adjustment is ±0.1 diopter / degree. Assuming that the extraocular muscle pressure of children aged 6-12 years ranges from 0.5 to 2 kPa, the AC / A constant of children ranges from 3 to 5Δ / D, and the AC / A fluctuation caused by pressure is generally ≤0.3Δ / D, combining the pressure fluctuation range and the upper limit of AC / A fluctuation, it is calculated that the "AC / A change caused by unit pressure change" is about ±0.1Δ / D / kPa; the lens curvature response coefficient is determined according to the actual range of extraocular muscle pressure and the range of AC / A constant of children. The proportion of red light (620-660 nm) in the multi-spectrum adjustment unit is initially set to 20%, and for every 1% increase in deviation beyond the limit, the proportion of red light is increased by 2% (upper limit 40%) for every 1% increase in deviation beyond the limit, and falls by 5% after 1 minute. Among them, the preset potential adjustment deviation of 5% and the associated parameter correction rule are based on the linear relationship between the fluctuation range of the ciliary muscle adjustment amount and the potential drift in the clinical data, and the corresponding relationship between the quantitative deviation and the parameter adjustment is realized to achieve accurate calibration of the mapping model. The initial proportion of red light in the multi-spectrum adjustment unit is set to 20%, and the proportion of red light is increased in steps when the deviation exceeds the limit, which not only avoids the stimulation of red light on the retina, but also dynamically enhances the intervention according to the degree of insufficient ion pump activity.

[0089] In the 250Hz interrupt service program, the training device calculates the first-order difference for each sample: if the absolute value of the slope is <0.5µV or the instantaneous slope exceeds ±40µV within 1s, it is determined that the electrode is off or abnormally drifted, and the invalid flag is immediately set, and a 0.5mm flat motor and a red LED are driven to emit a 100ms vibration-light prompt; this double-condition verification uses integer operations throughout, and the additional load on the CPU is <2%, with no risk of empty sampling at high sampling rates.

[0090] The above-mentioned flexible Ag / AgCl electrode is linked with the scheme generation unit, and when the proportion of red light is ≥30%, the scheme generation unit automatically adds 1 set of 5-minute brain training in the training plan for the day, and uses the synergistic effect of red light and brain training to strengthen the repair of ion pumps. At the same time, the electrode data collection and the mode switching unit of the dynamic training module cooperate, and when the potential drift does not change for 30 seconds, it is determined that the electrode contact is poor, triggering a shell vibration reminder (amplitude 0.5mm) to avoid data loss. The electrode data collection is converted into an active intervention strategy by linking with the scheme generation unit and the vibration reminder design, i.e. additional brain training when the proportion of red light is ≥30%, which utilizes the synergistic effect of the ion pump activation of red light and the neural regulation of brain training; no potential drift for 30 seconds triggers a vibration reminder, which can quickly identify electrode shedding and other data collection abnormalities, and avoid invalid training.

[0091] The ultra-thin base and special design improve the comfort of electrode wearing and prolong the tolerance time of children; the contact resistance control ensures the improvement of the signal-to-noise ratio of potential signal, and the detection accuracy of 0.5 mV level weak drift reaches a high level.

[0092] The parameter correction rule reduces the adjustment error of the mapping model to ±0.15Δ / D when the potential deviation exceeds the limit, compared with the fixed parameter model, which improves the accuracy; the red light step adjustment speeds up the recovery speed of ion pump activity and shortens the tear film break-up time.

[0093] The linkage with the scheme generation unit improves the training efficiency of children with severe adjustment function abnormalities, and the vibration reminder reduces the proportion of invalid training caused by data missing, realizing the whole-chain closed-loop optimization of signal acquisition, model calibration and training intervention.

[0094] Specifically, the data fusion unit performs four-dimensional correlation analysis on the corneal surface potential change data, intraocular pressure data, tear film break-up time and pressure deformation data, and establishes a dynamic coupling model combining potential, intraocular pressure, tear film and pressure.

[0095] When the linear deviation of potential drift amplitude and adjustment amount exceeds the preset potential adjustment deviation, and the intraocular pressure exceeds the fluctuation range or the tear film break-up time is shorter than the preset break-up threshold, the mapping model correlation parameter weight is corrected, the red and green light proportions of the multi-spectral adjustment unit are adjusted, and the contrast control unit operation is suspended.

[0096] Among them, the training device continuously captures the micro-deformation of the outer wall of the eye with a 5Hz flexible pressure sensor, and simultaneously outputs a 5-25mmHg near-end pressure from a micro-strain bridge chip packaged in the same area; after Goldman calibration, a linear mapping of "outer wall stress + corneal vertex micro-displacement" is established, and individual three-point anchoring (8 / 15 / 22mmHg) is used to correct zero drift every 30s in the tear film stable period, IOP_est is calculated in real time and used as the acquisition result of the intraocular pressure value, refreshed every 1s, with an error of ≤±0.5mmHg, and when the estimated value is >18mmHg or <1mmHg, the training intensity is automatically reduced to ensure training safety under high intraocular pressure.

[0097] The device captures the blinking behavior in real time through an infrared light supplementing front lens camera (30fps), takes the last blinking frame as the starting point, identifies the first ≥3 pixel connected dry spot on the cornea frame by frame, and calculates the tear film break-up time (TBUT) by dividing the frame number difference by 30, which is updated automatically every 30s, ≥10s is normal, ≤5s is dry eye, which is used for intraocular pressure zero drift correction and training parameter linkage.

[0098] More specifically, the dynamic coupling model adopts a multi-layer perceptron algorithm, the input layer contains 4-dimensional data, i.e. potential drift amplitude, intraocular pressure value, tear film break-up time and pressure deformation peak value, the hidden layer is set to 2 neurons, and the output layer is the data reliability score (0-100 points). The preset potential adjustment deviation is 5%, the intraocular pressure fluctuation range is 12-18 mmHg, and the tear film break-up threshold is 5 seconds. The multi-layer perceptron algorithm can efficiently process multi-dimensional nonlinear data correlation, balance the calculation complexity and fitting accuracy through 2 hidden layer neurons, and is suitable for real-time operation in the embedded mode of the present scheme. The output data reliability score (0-100 points) can quantify the data quality, providing clear judgment basis for subsequent intervention. The operation delay of the dynamic coupling model is controlled within 50 ms, the data correlation analysis can achieve high accuracy, compared with the traditional linear model, the accuracy can be greatly improved, and the reliability score can quickly identify invalid data, avoiding intervention based on incorrect data.

[0099] Among them, the preset parameters, i.e. potential adjustment deviation 5%, intraocular pressure 12-18 mmHg, and tear film threshold 5 seconds, refer to the statistical values of children's eye physiological data, which almost cover all normal fluctuation ranges and can timely capture abnormal states. The effect is to reduce most of the false triggering interventions, while the identification sensitivity for real abnormal states reaches a high level, ensuring that the present scheme can respond in time when the child's eye state is slightly abnormal.

[0100] When the triple regulation is triggered, the correction rule is: the weight of the associated parameter, the weight of the extraocular muscle pressure, the deviation of the actual value and the reference value of the extraocular muscle pressure changes by 1%, the weight is adjusted by ±0.03 synchronously, the response coefficient of the lens curvature is ±0.15Δ / D / kPa; the proportion of red light (620-660 nm) in the multi-spectrum regulation unit increases from 20% to 22%-24%, and the proportion of green light (500-550 nm) increases from 30% to 50%, and the green light falls to 40% after 2 minutes; during the suspension period of the contrast regulation unit, the contrast of the training picture is maintained at the current value.

[0101] In the above-mentioned triple regulation rule, the adjustment amplitude of the associated parameter weight (i.e. the weight adjustment ±0.03 corresponding to the change of 1% of the deviation) is based on the fitting of the previous clinical data, to ensure the balance between correction speed and stability; the stepwise regulation of the proportion of red light and green light avoids the stimulation of light wave mutation to vision, and can accurately intervene according to the pathological degree (red light enhances ion pump, green light improves tear film). After the correction of the associated parameters, the measurement error of the regulation amount of the mapping model is reduced to ±0.1Δ / D, which improves the ion pump activity of light wave regulation, prolongs the tear film break-up time, and reduces the discomfort of children to light wave changes.

[0102] The above specific mode and mode switching unit linkage, if dynamic coupling model score < 60 points and lasts 10 seconds, mode switching unit forced switching to brain training, while the scheme generation unit records this abnormal data, for optimization next day training scheme. In addition, the built-in data self-calibration module in this scheme, the first use of daily automatically collect 3 groups of basic data (including resting intraocular pressure, tear film break-up time and corneal surface potential baseline), as the daily model parameter reference value, reduce the individual day physiological fluctuations.

[0103] Linkage with mode switching unit set score < 60 points and lasts 10 seconds trigger switching, to prevent transient data fluctuations caused by misoperation, 10 seconds delay takes into account the response speed and accuracy; scheme generation unit records abnormal data, can be optimized through machine learning next day scheme, realize personalized training iteration. Reduce the mode mis-switching rate, improve the adaptability of next day training scheme, make the training more in line with the children's eye state.

[0104] Data self-calibration module daily collection of 3 groups of basic data, due to the existence of children's eye physiological indicators of day-to-day fluctuations (such as the morning intraocular pressure is high), with the daily reference value instead of fixed standard, can reduce the influence of individual physiological rhythm on the model accuracy, make the parameter correction more in line with the real-time state. Thus reduce the model error caused by day-to-day physiological fluctuations, improve the fitting degree of parameter correction, further guarantee the accuracy of training.

[0105] Specifically, the dynamic coupling model introduces a dynamic baseline calibration coefficient K, K is calculated based on the initial drift value of corneal surface potential, intraocular pressure baseline value, tear film break-up baseline time according to the formula K = 0.3 × potential baseline + 0.2 × eye pressure baseline + 0.5 × tear film baseline, updated every 30 seconds; K value embedded in the reliability score calculation of dynamic coupling model, when K fluctuation ≤ 5%, the score weight is allocated according to the original proportion; K fluctuation more than 5%, the score weight of intraocular pressure and tear film data is increased by 20%; at the same time, K value and multispectral adjustment unit linkage, when K fluctuation more than 5%, in addition to red and green light adjustment, the intensity of supplementary light increases linearly with the deviation of K value, and trigger the data fusion unit to shrink the error threshold to ± 0.2Δ / D.

[0106] More specifically, the initial value of the dynamic baseline calibration coefficient K is calculated by 10 groups of basic data collected continuously for 30 seconds before training. The potential baseline is taken as the average value in the interval of 0.5-1.2 mV, the intraocular pressure baseline is taken as the average value in the interval of 12-18 mmHg, and the tear film baseline is taken as the first detection of the break-up time. When the K fluctuation is ≤5%, the four-dimensional data weight in the reliability score is 25% for potential, 20% for intraocular pressure, 30% for tear film, and 25% for pressure; when the fluctuation is >5%, the weight of intraocular pressure and tear film is each increased by 20% (i.e., 24% for intraocular pressure and 36% for tear film). The light intensity adjustment step is 1% light power for each 0.1 unit K value deviation, and after the error threshold is contracted to ±0.2Δ / D, the image acquisition unit is triggered to take an additional 2 frames of corneal reflection images every 0.3 seconds.

[0107] The weight distribution of K value (i.e., 0.5 for tear film > 0.2 for intraocular pressure > 0.3 for potential) is based on the degree of influence of the ocular surface state on data acquisition, and the interference of tear film stability on image quality is the most significant, so it is given the highest weight. The 30-second update cycle balances real-time and data smoothness, avoiding frequent adjustments caused by transient fluctuations. In the linkage of dynamic baseline calibration coefficient K and double verification (i.e., comparison verification between the first data and the second data), the error threshold is dynamically contracted to realize secondary verification of data quality, forming a closed loop of calibration, verification, and re-sampling.

[0108] The dynamic adjustment of K value reduces invalid data caused by environmental interference during home training. In cooperation with the mode switching unit, when K value is continuously out of limit for 3 times, it automatically switches to low intensity mode, improving training safety; the additional image frames reduce the calculation error of tear film break-up time, and cooperate with the light enhancement of multi-spectral adjustment to improve image clarity in dark environments, realizing the combination of environmental self-adaptation, data precision, and training safety, achieving triple optimization.

[0109] In this specific embodiment, a ciliary muscle regulation-based children's visual function training method using a ciliary muscle regulation-based children's visual function training system is also included.

[0110] The above only describes the embodiments of the present application, and the specific structure and characteristics of the scheme are not described in detail. It should be noted that for those skilled in the art, without departing from the structure of the present application, a number of modifications and improvements can be made, which should also be considered as the protection scope of the present application, and these will not affect the effect and practicality of the present application. The protection scope claimed in this application should be subject to the content of its claims, and the specific embodiments and the like in the specification can be used to explain the content of the claims.

Claims

1. A child visual function training system based on ciliary muscle accommodation, characterized in that, The application relates to a dynamic ciliary muscle training system. The system comprises a cooperative acquisition module, an image acquisition unit, a pressure acquisition unit and a data fusion unit. The image acquisition unit acquires real-time eyeball anterior segment images and identifies eye feature parameters in the eyeball anterior segment images. The pressure acquisition unit comprises pressure sensors arranged on the skin surface corresponding to the attachment points of the extraocular muscles on the lateral side of the eye orbit, and can acquire real-time pressure deformation data generated by the extraocular muscles due to the convergence movement of the eyeball. The data fusion unit acquires the eye feature parameters and the pressure deformation data collected at the same time, and performs correlation analysis, establishes a mapping model through the pressure deformation data change and the lens curvature change, indirectly calculates the lens power change data as first data according to the eye feature parameters through a thin lens power conversion algorithm, inputs the pressure deformation data into the mapping model to calculate the real-time state of the ciliary muscle, and calculates the lens power change data in the eye feature parameters as second data according to the real-time state. The data fusion unit compares the first data with the second data, if the comparison result falls within a preset error threshold, the real-time state is processed into the real-time adjustment amount of the ciliary muscle according to the acquisition time, if the comparison result exceeds the error threshold range, the image acquisition unit and the pressure acquisition unit acquire the eye feature parameters and the pressure deformation data again, and the first data and the second data are calculated and compared. The dynamic training module acquires individual data of children and historical data of the cooperative acquisition module, formulates training parameters according to a preset training period, and dynamically adjusts the training parameters according to the real-time adjustment amount of the ciliary muscle. The real-time state comprises a fatigue state and an active state, and the training parameters comprise a training duration and a light wave combination. The training mode in the real-time state control parameters is adjusted according to the continuous real-time state. The data fusion unit is also used for verifying the accuracy of the AC / A constant according to the synchronous acquisition of the pressure deformation data and the eye feature parameters.

2. The ciliary muscle accommodation based vision function training system for children according to claim 1, characterized in that: The data fusion unit is also used for verifying the accuracy of the AC / A constant according to the synchronous acquisition of the pressure deformation data and the eye feature parameters. The dynamic training module comprises a dynamic adjustment unit and a multi-spectrum adjustment unit. When the dynamic adjustment unit detects that the real-time state of the ciliary muscle is a fatigue state, the target switching frequency is reduced to 8-12 times per minute, and the proportion of green light at 500-550 nm is increased to 60% to promote relaxation; when the active state is detected, the target switching frequency is increased to 15-20 times per minute, and the proportion of red light at 620-660 nm is increased to 50% to strengthen the training; The multispectral adjustment unit is used to emit specific light waves of 430-680 nm, wherein the green light adjusts the pupil diameter by alternating light and dark brightness to train the ciliary muscle and the pupil coordination, and the red light is used for light nutrition supplement to stimulate the acceleration of choroid blood circulation. This unit not only realizes training enhancement through light wave characteristics, but also enhances the linkage of the visual system through pupil adjustment.

3. The cyclo- accommodation based child vision function training system according to claim 2, wherein: The dynamic training module further comprises a contrast regulation unit, a mode switching unit and a scheme generation unit; The contrast regulation unit is used to dynamically reduce the contrast of the training picture in the training parameters to strengthen the regulation stability of the ciliary muscle in complex visual environment; The mode switching unit supports automatic switching between material training and brain training. When the data fusion unit determines that the ciliary muscle is in a fatigue state for three times in succession, the brain training is switched for 1-2 minutes; The scheme generation unit outputs the daily training time and light wave combination mode according to the age of the child, the degree of amblyopia and the historical data of the coordination acquisition module, realizes precise regulation of the training and recovery effect, and is linked with the dynamic adjustment unit, so that the generated scheme can be updated in real time according to the state of the ciliary muscle. The degree of amblyopia is obtained by matching the eye feature parameters with the stored amblyopia features.

4. The cyclo- accommodation based child vision function training system according to claim 3, wherein: It also includes a training device, which comprises a shell, a display component and a fastening component. The display component is arranged inside the shell and fixedly connected with the shell. The fastening component is fixedly connected with the outside of the shell and used to fix the shell of the training device around the eye; The shell is provided with an orbital hole matched with the orbit. The outside of the shell around the orbital hole is provided with a pressure detection area corresponding to the skin surface of the lateral rectus muscle attachment point of the orbit. A plurality of pressure sensors of the pressure acquisition unit are fixedly connected with the pressure detection area. The image acquisition unit is a binocular miniature camera integrated above the orbital hole inside the training device. The binocular miniature camera is equipped with an infrared fill light. The inside of the shell is also fixedly connected with the display component. The display component is used to display the training picture according to the training parameters provided by the dynamic training module.

5. The cyclo- accommodation based child vision function training system according to claim 4, wherein: The pressure acquisition unit is integrated with a contact sensing assembly for synchronous acquisition of intraocular pressure data. The image acquisition unit analyzes the corneal reflection image to obtain the tear film break-up time, forming a cooperative verification method of intraocular pressure and tear film state. The cooperative verification method takes the intraocular pressure fluctuation threshold and the tear film break-up time threshold as auxiliary verification conditions, and performs correlation analysis on the pressure deformation data collected by the pressure sensor and the eye feature parameters obtained by the image acquisition unit; when the intraocular pressure exceeds the preset fluctuation range, the data fusion unit corrects the weight of the correlation parameter in the mapping model, the correlation parameter being a weight coefficient in the mapping model for correlating the lens curvature and the extraocular muscle pressure; when the tear film break-up time is shorter than the preset break-up threshold, the multi-spectrum adjustment unit adjusts the light wave proportion to improve the ocular surface state.

6. The cyclo- accommodation based child vision function training system according to claim 5, wherein: The flexible Ag / AgCl electrode is fixedly connected to the tragus corresponding position outside the shell of the training device and the corneal limbus corresponding to the edge of the orbital hole, and has a sampling frequency of 250 Hz, and is used to collect the potential change of the corneal surface; When the ciliary muscle contracts, the epithelial ion pump activity is enhanced, and the potential of the corneal surface drifts to the negative value by 0.5-1.2 mV; The data fusion unit performs correlation analysis on the potential change data of the corneal surface, the intraocular pressure data, the tear film break-up time and the pressure deformation data, and when the linear deviation of the potential drift amplitude and the adjustment amount exceeds the preset potential adjustment deviation, the correlation parameter weight of the mapping model is corrected synchronously, the correlation parameter being a weight coefficient in the mapping model for correlating the lens curvature and the extraocular muscle pressure, and the multi-spectrum adjustment unit is controlled by the dynamic training module to adjust the red light proportion to enhance the ion pump activity.

7. The cyclo- accommodation based child vision function training system according to claim 6, wherein: The data fusion unit performs four-dimensional correlation analysis on the potential change data of the corneal surface, the intraocular pressure data, the tear film break-up time and the pressure deformation data, and establishes a dynamic coupling model combined with the potential, the intraocular pressure, the tear film and the pressure; When the linear deviation of the potential drift amplitude and the adjustment amount exceeds the preset potential adjustment deviation, and the intraocular pressure exceeds the fluctuation range or the tear film break-up time is shorter than the preset break-up threshold, the correlation parameter weight of the mapping model is corrected synchronously, the red light and green light proportions of the multi-spectrum adjustment unit are adjusted, and the operation of the contrast control unit is suspended.

8. The cyclo- accommodation based child vision function training system according to claim 7, wherein: The dynamic coupling model introduces a dynamic baseline calibration coefficient K, which is calculated based on the initial drift value of the potential of the corneal surface, the basic value of the intraocular pressure and the break-up reference time of the tear film according to the formula K=0.3×potential baseline+0.2×intraocular pressure baseline+0.5×tear film baseline, and is automatically updated every 30 seconds; The K value is embedded in the reliability score calculation of the dynamic coupling model, when the K fluctuation is less than or equal to 5%, the score weight is distributed in the original proportion; when the K fluctuation exceeds 5%, the score weight of the intraocular pressure and the tear film data is increased by 20%; Meanwhile, the K value is linked with the multi-spectrum adjustment unit, when the K fluctuation exceeds 5%, in addition to the red light and green light adjustment, the light intensity is linearly increased with the deviation of the K value, and the data fusion unit is triggered to shrink the error threshold to ±0.2Δ / D.

Images