Intraocular pressure change-based contact lens pressure developing system for sports eye protection

Abstract

The invention discloses an intraocular pressure change-based contact lens pressure developing system for sports eye protection, and relates to the technical field of medical instruments, a data collection unit and a sending unit are arranged in a contact lens, the data collection unit collects eye related parameters, and the eye related parameters are sent to a central control unit through the sending unit; the central control unit transmits the eye related parameters to the intraocular pressure processing unit, intraocular pressure prediction is carried out based on a preset intraocular pressure prediction model, and an intraocular pressure prediction value is obtained; the central control unit receives the intraocular pressure predicted value and outputs a corresponding color development signal based on a preset range threshold value, and the color development unit performs related color development based on the color development signal. The intraocular pressure detection system provided by the invention is intelligent, high in accuracy and high in reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical devices, more particularly to a contact lens pressure coloration system for sports eye protection based on intraocular pressure changes. BACKGROUND

[0002] Glaucoma is an eye disease that damages the optic nerve due to excessive intraocular fluid, resulting in continuous or intermittent elevated eye pressure beyond the tolerance of the eyeball. Long-term continuous measurement of eye pressure is of great significance for the diagnosis and treatment of glaucoma patients.

[0003] By implanting a micro-sensor into the eye and forming a real-time monitoring system, it is possible to achieve long-term continuous monitoring of eye pressure. However, the implantable intraocular pressure monitoring devices and methods reported in the literature often use only one or one type of sensor, and the type of eye pressure data measured is relatively single. Patients in daily work and life have various complex situations such as stillness, movement, acceleration and deceleration, body temperature changes, and body position changes, which can affect the accurate measurement of eye pressure. Therefore, it is difficult to obtain accurate and reliable eye pressure data and information. Traditional eye pressure measurement methods such as Goldmann pressure flattening tonometer require professional equipment and anesthesia, and air pulse tonometer has insufficient precision, which cannot meet the needs of real-time and non-invasive monitoring in sports. In the sports scene, intraocular pressure monitoring faces the following special challenges:

[0004] Dynamic interference: vibration, sweat, and changes in eye blood flow during exercise can interfere with sensor signals;

[0005] Data scarcity: direct acquisition of eye pressure-physiological signal calibration data in sports has high cost and small sample size;

[0006] Wearing comfort: traditional monitoring equipment is difficult to balance biocompatibility and sports protection needs.

[0007] In summary, the implantable intraocular pressure monitoring devices and methods in the prior art cannot meet the needs of high accuracy, high reliability, and intelligent real-time monitoring of patient eye pressure.

[0008] Therefore, how to provide an intelligent, high-accuracy, and high-reliability intraocular pressure detection system is a problem that needs to be solved by those skilled in the art. SUMMARY

[0009] Therefore, the present application provides a contact lens pressure coloration system for sports eye protection based on intraocular pressure changes, and an intelligent, high-accuracy, and high-reliability intraocular pressure detection system.

[0010] To achieve the above purpose, the present application adopts the following technical solutions:

[0011] A contact lens pressure color display system for sports eye protection based on intraocular pressure changes includes: a data collection unit, a sending unit, a central control unit, an intraocular pressure processing unit, and a color display unit. The data collection unit and the sending unit are arranged inside the contact lens. The data collection unit collects eye-related parameters and sends them to the central control unit via the sending unit; the central control unit transmits the eye-related parameters to the intraocular pressure processing unit, predicts the intraocular pressure based on a preset intraocular pressure prediction model, and obtains an intraocular pressure prediction value; the central control unit receives the intraocular pressure prediction value and outputs a corresponding color display signal based on a preset range threshold; the color display unit performs relevant color display based on the color display signal.

[0012] Preferably, the data collection unit includes a carbon nanotube composite film strain sensor for collecting deformation data, a flexible FET biosensor for collecting ion concentration data, and a polyimide thermocouple for collecting eye surface temperature data.

[0013] Preferably, in the intraocular pressure processing unit, the preset intraocular pressure prediction model training includes: recording the existing data under different motion conditions as the source domain D S ={X S , Y S}; where X S The model input includes deformation data, ion concentration data and ocular surface temperature data, Y S is the model output, i.e. the predicted intraocular pressure of the corresponding patient;

[0014] The eye-related parameters of the motion to be predicted are recorded as the target domain, and the target domain data D is obtained. T ={X T}, X T is the model input to be predicted; obtain the corresponding intraocular pressure value under some of the motion conditions to be predicted, recorded as D G ={X G , Y G}; D S With D G Fusion, get training data D train ={X SG , Y SG}, the eye-related parameters that do not have corresponding intraocular pressure values ​​in the target domain are recorded as D test ={X TG};

[0015] To X SG and X TG Perform feature analysis mapping and parameter optimization to obtain the optimal feature analysis mapping; use the training data corresponding to the optimal feature analysis mapping to train the neural network model to obtain the intraocular pressure prediction value.

[0016] Preferably, the pair X SG and XTG Feature analysis and mapping specifically include:

[0017] Solve the feature map φ and use the neural tangent kernel method to measure the distance between the source domain and the target domain:

[0018]

[0019] Among them, n s 、n t 、n g are the number of samples in the source domain, target domain, and target domain containing IOP prediction values, respectively. ω is the weight parameter, ▽ is the Jacobian matrix; ||·|| represents the reproducing kernel Hilbert space norm;

[0020] Using the kernel function matrix K and the coefficient matrix L to simplify the operation, we get NTK = tr (KL), and further get the feature optimization target as:

[0021]

[0022] Solve to get the mapping matrix W, multiply it with the kernel function matrix W T K, get the model input of the mapped training data and the model input of the mapped predicted data Among them, λ is the regularization parameter.

[0023] Preferably, the obtaining of the optimal feature analysis mapping specifically includes: using K-means clustering to respectively and Perform clustering and obtain their respective cluster centers c s and c t ; Calculate c s and c t The distance CD=||c s -c t || 2 ; Compare the CD values ​​after feature analysis mapping with different hyperparameter values, and search for the λ value that minimizes the CD value within a given range through the hyperparameter optimization algorithm.

[0024] Preferably, the hyperparameter optimization algorithm adopts a grid search method.

[0025] Preferably, the color development unit is integrated or separate: the integrated type embeds a pH-responsive fluorescent hydrogel in the contact lens, and controls the amount of dye released by the color development signal output to achieve color change; the separate type transmits the predicted color development value to a portable wearable device via NFC, and an indicator light corresponding to the color development signal is set inside the wearable device, and the corresponding indicator light is controlled to light up based on the color development signal.

[0026] Through the above technical solutions, it can be seen that compared with the prior art, the present invention discloses a contact lens pressure color display system for sports eye protection based on intraocular pressure changes. The data collection unit includes a carbon nanotube composite film strain sensor, a flexible FET biosensor and a polyimide thermocouple, which can respectively collect corneal deformation data, ion concentration data and eye surface temperature data. The synchronous collection of multi-dimensional parameters can more comprehensively reflect the physiological state of the eye. The intraocular pressure processing unit uses the source domain and target domain data fusion method to train the prediction model, and uses the existing different motion situation data as the source domain and the eye parameters of the motion situation to be predicted as the target domain. By obtaining the intraocular pressure values ​​of some target domains and fusing them with the source domain data, the amount of training data is expanded, and the problem of scarcity of intraocular pressure calibration data during exercise is solved; the neural tangent kernel method is used to measure the distance between domains, and the kernel function matrix and coefficient matrix are simplified Operation, combined with K-means clustering and grid search method for feature optimization and hyperparameter optimization, to find the optimal feature analysis mapping, so that the model can better adapt to the differences in motion scenes, improve the accuracy of intraocular pressure prediction, and better handle nonlinear relationships in complex motion states than traditional models; the central control unit outputs a color signal according to the preset range threshold. When the intraocular pressure prediction value exceeds the normal range, the color unit promptly realizes real-time warning through color changes or light prompts, helping users or medical personnel to quickly detect abnormal intraocular pressure, and provide early intervention for eye diseases such as glaucoma. Eye parameters are collected non-invasively through sensors on contact lenses, transmitted to the central control unit in real time for processing and prediction of intraocular pressure, and then intuitively displayed through the color unit, realizing real-time and non-invasive monitoring of intraocular pressure during exercise, providing an effective means for glaucoma patients to manage intraocular pressure in daily activities such as exercise. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0028] Figure 1 This is a structural diagram provided by the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0030] The embodiment of the present invention discloses a pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes, such as Figure 1 As shown, it includes: a data collection unit, a sending unit, a central control unit, an intraocular pressure processing unit, and a color display unit. The data collection unit and the sending unit are arranged inside the contact lens. The data collection unit collects eye-related parameters and sends them to the central control unit through the sending unit; the central control unit transmits the eye-related parameters to the intraocular pressure processing unit, and predicts the intraocular pressure based on a preset intraocular pressure prediction model to obtain an intraocular pressure prediction value; the central control unit receives the intraocular pressure prediction value, and outputs a corresponding color display signal based on a preset range threshold, and the color display unit performs relevant color display based on the color display signal.

[0031] In a specific embodiment, the data collection unit includes a carbon nanotube composite film strain sensor for collecting deformation data, a flexible FET biosensor for collecting ion concentration data, and a polyimide thermocouple for collecting eye surface temperature data.

[0032] In the data collection unit, the sensor works as follows:

[0033] Working principle of carbon nanotube composite film strain sensor:

[0034] Sensitivity mechanism: Utilizes the piezoresistive effect of carbon nanotubes (CNTs) in a PDMS matrix. When corneal deformation (blinking, changes in intraocular pressure) acts on the film, the CNT network stretches or compresses, causing the carrier transport path to change and the resistance value to change accordingly (strain-resistance linear response).

[0035] Structural design:

[0036] The substrate is an ultra-thin PDMS film (thickness <50μm), ensuring flexibility and biocompatibility to fit the corneal surface. The conductive layer is made by evenly dispersing CNTs in PDMS through spin coating, spray coating, or transfer methods to form a conductive network. Gold or silver electrodes are evaporated at both ends of the electrode to extract the resistance signal.

[0037] Working principle of flexible FET biosensor for detecting Na+ concentration:

[0038] Ion-sensitive mechanism: Based on the principle of ion-sensitive field-effect transistor, the gate area is modified with a sodium ion selective membrane (such as valinomycin coating). After Na+ in the tear fluid combines with the membrane layer, it changes the charge density on the gate surface, thereby modulating the channel current of the FET.

[0039] Structural design:

[0040] The substrate is a flexible polyimide or PDMS film; the device layer uses deposited silicon nanowires or graphene as the channel material, and the source / drain electrodes are made of metal (Cr or Au); the gate region of the sensitive layer is coated with a Na+ selective ion exchange membrane, which is in direct contact with tear fluid;

[0041] Sensor working principle of polyimide thermocouple

[0042] Thermoelectric effect: based on the Seebeck effect, deposit two different metals (such as copper-constantan, nickel-nichrome) on the polyimide substrate to form a thermocouple junction. When the ocular surface temperature changes, the junction generates a thermoelectric potential proportional to the temperature difference (typical Seebeck coefficient 40-60 μV / ℃).

[0043] Structural design:

[0044] The substrate is ultra-thin polyimide (<10 μm), which has flexibility and high temperature resistance;

[0045] Thermoelectric electrode: form a metal thin film (thickness <1 μm) on PI by magnetron sputtering or evaporation technology, the intersection of the two electrodes is the temperature measurement point;

[0046] Insulating layer: PI covers the electrode, only the temperature measurement point is in contact with the ocular surface.

[0047] In specific embodiment 1, the central control unit maps the historical intraocular pressure data Y={y1, y2,..., y n} and the corresponding sensor features X through feature component analysis (TCA) to obtain feature vectors Z={z1, z2,..., z n};

[0048] K-cluster Z to obtain cluster centers C={c1, c2,..., c k}, and calculate the mean intraocular pressure of each cluster: Take the midpoint of the mean intraocular pressure of adjacent clusters as the boundary of the range threshold.

[0049] In specific embodiment 2, the central control unit constructs a KD-tree based on historical intraocular pressure data and corresponding sensor features for fast retrieval of adjacent data points. Let the intraocular pressure data set be T={(t i ,x i ,y i ,z i )}, where t i is the intraocular pressure value, x i , y i , z i are the deformation feature vector, ion concentration feature vector and ocular surface temperature feature vector respectively. K-nearest neighbor search is realized through KD-tree.

[0050] Based on KD-tree, calculate the Euclidean distance between each intraocular pressure sample and its nearest neighbor sample, and obtain the global average distance:

[0051] Single sample distance: for each intraocular pressure sample t i, calculate the Euclidean distance between it and the nearest neighbor sample where t j is the nearest neighbor intraocular pressure value, x j is the corresponding sensor feature;

[0052] Global average spacing:

[0053] Where N is the total number of samples, Characterize the overall dispersion of intraocular pressure data;

[0054] Divide the intraocular pressure-sensor feature space into a cubic grid for voxelization and calculate the point cloud density

[0055] The intraocular pressure sensor feature space is divided into a cubic grid with a grid side length of Δt×Δx×Δy×Δz. The number of grids in each direction is:

[0056]

[0057] Among them, t max , t min The intraocular pressure range.

[0058] Density calculation:

[0059] Where s(l) is the number of grids containing data, n i is the number of samples in each grid, and ρ represents the density of intraocular pressure data in a unit space;

[0060] Determine the compensation factor e and the initial threshold d i

[0061] Compensation Factor

[0062] Where, L is the number of sensor channels;

[0063] Select the intraocular pressure sample pair (t1, t2) in the high-density area and calculate the initial segmentation threshold d i =|t1-t2|;

[0064] Assume that the distance from the sensor feature of the current IOP to be predicted to the training data center is D (which can be calculated by Euclidean distance), then the dynamic threshold is: d t =e·D+d i ;

[0065] Threshold level classification:

[0066] Normal / mildly abnormal boundary:

[0067] Mild / moderate abnormality boundary:

[0068] Moderate / severe abnormality boundary:

[0069] Among them, e1, e2, and e3 are compensation factors corresponding to each level, which are obtained by training the boundary samples of each level in historical data.

[0070] In a specific embodiment, in the intraocular pressure processing unit, the preset intraocular pressure prediction model training includes: recording the existing data under different motion conditions as the source domain D S ={X S , Y S}; where X S The model input includes deformation data, ion concentration data and ocular surface temperature data, Y S is the model output, i.e. the predicted intraocular pressure of the corresponding patient;

[0071] The eye-related parameters of the motion to be predicted are recorded as the target domain, and the target domain data D is obtained. T ={X T}, X T is the model input to be predicted; obtain the corresponding intraocular pressure value under some of the motion conditions to be predicted, recorded as D G ={X G , Y G}; D S With D G Fusion, get training data D train ={X SG , Y SG}, the eye-related parameters that do not have corresponding intraocular pressure values ​​in the target domain are recorded as D test ={X TG};

[0072] To X SG and X TG Perform feature analysis mapping and parameter optimization to obtain the optimal feature analysis mapping; use the training data corresponding to the optimal feature analysis mapping to train the neural network model to obtain the intraocular pressure prediction value.

[0073] In a specific embodiment, X SG and X TG Feature analysis and mapping specifically include:

[0074] Solve the feature map φ and use the neural tangent kernel method to measure the distance between the source domain and the target domain:

[0075]

[0076] Among them, n s 、n t 、n gare the number of samples in the source domain, target domain, and target domain containing IOP prediction values, respectively. ω is the weight parameter, is the Jacobian matrix; ||·|| represents the reproducing kernel Hilbert space norm;

[0077] Using the kernel function matrix K and the coefficient matrix L to simplify the operation, we get NTK = tr(KL):

[0078]

[0079] The feature optimization objective is further obtained as:

[0080]

[0081] Among them, W is the feature mapping matrix, K is the kernel function matrix, K SS is the similarity between source domain data, K TS and K ST is the similarity between the source domain and the target domain data, K TT is the similarity between the target domain data, tr() represents the trace of the matrix, L is the label matrix, which is used to calculate the distance between domains; H is the centralization matrix, which is used for data standardization; I m is the m-dimensional identity matrix.

[0082] Solve to get the mapping matrix W, multiply it with the kernel function matrix W T K, get the model input of the mapped training data and the model input of the mapped predicted data The model inputs of the mapped training data and prediction data are the first n s +n g Row and back n t -n g rows; where λ is the regularization parameter.

[0083] In a specific embodiment, obtaining the optimal feature analysis mapping specifically includes: using K-means clustering to and Perform clustering and obtain their respective cluster centers c s and c t ; Calculate c s and c t The distance CD=||c s -c t || 2 ; Compare the CD values ​​after feature analysis mapping with different hyperparameter values, and search for the λ value that minimizes the CD value within a given range through the hyperparameter optimization algorithm.

[0084] In a specific embodiment, the hyperparameter optimization algorithm adopts a grid search method.

[0085] In a specific embodiment, the color display unit is integrated or separate: the integrated type embeds a pH-responsive fluorescent hydrogel in the contact lens, and controls the amount of dye released by the color display signal output to achieve color change; the separate type transmits the predicted color display value to a portable wearable device via NFC, and an indicator light corresponding to the color display signal is set inside the wearable device, and the corresponding indicator light is controlled to light up based on the color display signal.

[0086] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0087] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A contact lens pressure color display system for sports eye protection based on intraocular pressure changes, characterized in that: include: A data collection unit, a sending unit, a central control unit, an intraocular pressure processing unit, and a color display unit. The data collection unit and the sending unit are arranged inside the contact lens. The data collection unit collects eye-related parameters and sends them to the central control unit through the sending unit; the central control unit transmits the eye-related parameters to the intraocular pressure processing unit, predicts the intraocular pressure based on a preset intraocular pressure prediction model, and obtains an intraocular pressure prediction value; the central control unit receives the intraocular pressure prediction value, and outputs a corresponding color display signal based on a preset range threshold, and the color display unit performs relevant color display based on the color display signal.

2. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 1, characterized in that: The data collection unit includes a carbon nanotube composite film strain sensor for collecting deformation data, a flexible FET biosensor for collecting ion concentration data, and a polyimide thermocouple for collecting eye surface temperature data.

3. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 2, characterized in that: In the intraocular pressure processing unit, the preset intraocular pressure prediction model training includes: recording the existing data under different motion conditions as the source domain D S ={X S , Y S }; where X S The model input includes deformation data, ion concentration data and ocular surface temperature data, Y S is the model output, i.e. the predicted intraocular pressure of the corresponding patient; The eye-related parameters of the motion to be predicted are recorded as the target domain, and the target domain data D is obtained. T ={X T }, X T is the model input to be predicted; obtain the corresponding intraocular pressure value under some of the motion conditions to be predicted, recorded as D G ={X G , Y G }; D S With D G Fusion, get training data D train ={X SG , Y SG }, the eye-related parameters that do not have corresponding intraocular pressure values ​​in the target domain are recorded as D test ={X TG }; To X SG and X TG Perform feature analysis mapping and parameter optimization to obtain the optimal feature analysis mapping; use the training data corresponding to the optimal feature analysis mapping to train the neural network model to obtain the intraocular pressure prediction value.

4. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 3, characterized in that: The pair X SG and X TG Feature analysis and mapping specifically include: Solve the feature map φ and use the neural tangent kernel method to measure the distance between the source domain and the target domain: Among them, n s 、n t 、n g are the number of samples in the source domain, target domain, and target domain containing IOP prediction values, respectively. ω is the weight parameter, is the Jacobian matrix; ||·|| represents the reproducing kernel Hilbert space norm; Using the kernel function matrix K and the coefficient matrix L to simplify the operation, we get NTK = tr (KL), and further get the feature optimization target as: Solve to get the mapping matrix W, multiply it with the kernel function matrix W T K, get the model input of the mapped training data and the model input of the mapped predicted data Among them, λ is the regularization parameter.

5. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 4, characterized in that: The method of obtaining the optimal feature analysis mapping specifically includes: using K-means clustering to respectively and Perform clustering and obtain their respective cluster centers c s and c t ; Calculate c s and c t The distance CD=||c s -c t || 2 ; Compare the CD values ​​after feature analysis mapping with different hyperparameter values, and search for the λ value that minimizes the CD value within a given range through the hyperparameter optimization algorithm.

6. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 5, characterized in that: The hyperparameter optimization algorithm adopts a grid search method.

7. The pressure color display system for contact lenses for sports eye protection based on intraocular pressure changes according to claim 1, characterized in that: The color rendering unit can be integrated or separated: the integrated type embeds a pH-responsive fluorescent hydrogel in the contact lens, and controls the amount of dye released by the color rendering signal output to achieve color change; the separated type transmits the predicted color rendering value to a portable wearable device via NFC, and an indicator light corresponding to the color rendering signal is set inside the wearable device, which controls the corresponding indicator light to light up based on the color rendering signal.

Images