Personalized sclera lens intelligent fitting system based on biomechanics and application thereof

Abstract

The invention discloses a personalized sclera lens intelligent fitting system based on biomechanics. The personalized sclera lens intelligent fitting system comprises the following steps that 1, information of anterior segments of eyes is collected through an information collection module; 2, recording the change process of anterior segment information through a micro excitation mode, extracting local strain characteristics of cornea and sclera by combining a DIC technology, and analyzing biomechanical response under micro deformation; 3, based on biomechanical response analysis under tiny deformation, a mechanical inversion method, an optical coherence elastography (OCE) technology or a finite element iteration method is adopted to carry out backstepping on mechanical parameters to calculate material attributes of cornea and sclera; 4, establishing an SL finite element model database, establishing a personalized whole eyeball finite element model, establishing contact simulation of the sclera lens and the eyeballs of the human eyes and an overall simulation wearing process of the sclera lens and the eyeballs of the human eyes to obtain an SL finite element simulation model, and finally outputting an SL adaptation result obtained through simulation; and 5, based on the condition of the SL adaptation result, constructing a personalized SL fitting optimization strategy, and recommending the most suitable SL model and parameter.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical treatment, and in particular to a biomechanically based personalized scleral lens intelligent fitting system and application thereof. Background Art

[0002] In ophthalmology clinics, visual correction of complex corneal and ocular surface diseases has always been a difficult problem that needs to be solved urgently. Irregular astigmatism, difficult-to-heal ocular surface wounds, and severe dry eyes seriously affect patients' vision and quality of life. Traditional correction methods are difficult to meet the needs of such patients: frame glasses cannot correct irregular astigmatism, conventional contact lenses may further irritate the cornea, and surgical treatment has high risks and limited applicability. Therefore, how to provide more accurate and efficient correction methods for patients with complex corneal and ocular surface diseases has become a clinical problem that needs to be solved urgently.

[0003] In recent years, scleral lenses (SL) have become a new diagnostic and treatment method, providing a breakthrough solution for such patients. SL covers the cornea with a large diameter design to avoid direct contact with it, and forms a fluid reservoir (FR) between the lens and the cornea, which can effectively correct irregular astigmatism, relieve dry eye symptoms, and promote ocular surface healing. However, although the clinical efficacy of SL has been widely recognized, its fitting process is relatively complicated and highly dependent on the doctor's experience, lacking scientific and standardized guidance. At present, the fitting of SL is mainly based on the trial fitting and adjustment of corneal morphology and scleral contour. The Regional Variations in Postlens Tear Layer Thickness During Scleral Lens Wear[J]. Eye Contact Lens, 2020, 46(6): 368-74 published by VINCENT SJ et al. shows that the biomechanical properties of the cornea and sclera are also key factors affecting the fitting effect of SL. Due to the significant differences in the biomechanical properties of the cornea and sclera between individuals, fitting based only on morphological parameters may not accurately predict the actual wearing effect of SL. This fitting method is not only inefficient, increasing the number of trial wear times and discomfort for patients, but may also lead to complications such as corneal hypoxia, abnormal tear dynamics, and corneal epithelial damage due to poor fitting, further leading to a decline in visual quality and affecting the long-term wearing safety of SL.

[0004] In addition, the settling effect of SL lenses (i.e. the lens gradually adheres to the ocular surface after wearing for a period of time) takes 2-4 hours to reach a stable state, making it difficult to directly predict the fit after stable wearing during the initial fitting, further prolonging the fitting process, increasing the burden on patients, and increasing the learning curve for clinicians. These factors have, to a certain extent, hindered the large-scale promotion of SL in clinical practice, and there is an urgent need to develop a more scientific and efficient personalized fitting method to optimize the fitting process, improve fitting accuracy, and reduce patient discomfort and complication rates during trial wear.

[0005] In order to solve the above problems, the current scleral lens fitting mainly selects the initial trial lens based on the central sagittal height, without fully considering the force conditions in the contact area between SL and sclera, FR thickness changes and long-term sedimentation effects. Adjustments after trial fitting still rely on the physician's experience, and adjustments are made by observing the fitting state with a slit lamp. There is a lack of quantitative analysis methods, and it is difficult to scientifically and accurately optimize the fitting parameters. In addition, the biomechanical properties of the eyeball may affect the final fitting state, and existing research methods for in vivo measurement of corneal and scleral material properties still need to be further optimized. In the clinical fitting process, there is a lack of time-correlation analysis of the SL fitting situation, and it is impossible to accurately predict the fitting trend over time, resulting in a long fitting process and a complicated adjustment process.

[0006] The market needs an intelligent scleral lens fitting system that can solve the above problems, accurately predict the fitting effects of different types of SL, output FR thickness distribution and contact area force conditions, and achieve scientific, efficient and personalized SL fitting plan recommendations. The present invention solves such problems. Summary of the invention

[0007] In order to address the deficiencies of the prior art, the purpose of the present invention is to provide a personalized scleral lens intelligent fitting system based on biomechanics and its application, to solve the problems of the prior art such as complicated fitting process, low fitting accuracy, low patient wearing comfort and complications, and to predict the most suitable scleral lens model, thereby optimizing the wearing effect.

[0008] In order to achieve the above object, the present invention adopts the following technical solution: A personalized scleral lens intelligent fitting system based on biomechanics, including the following contents: Step 1, using an information collection module to collect information about the anterior segment of the eye; Step 2: Record the changes of the anterior segment information by micro-stimulation, extract the local strain characteristics of the cornea and sclera by combining DIC technology, and analyze the biomechanical response under micro-deformation; Step 3: Based on the biomechanical response analysis under small deformation, mechanical inversion method, optical coherence elastography (OCE) technology or finite element iteration method are used to reversely calculate the mechanical parameters and material properties of the cornea and sclera; Step 4: Establish an SL finite element model database, establish a personalized full-eye finite element model, establish a contact simulation between the scleral lens and the human eye, simulate the wearing process of the scleral lens and the human eye as a whole, obtain an SL finite element simulation model, and finally output the simulated SL adaptation results; the SL adaptation results include: FR thickness map characterizing the SL settlement, outputting the contact pressure and strain between the SL and the cornea and sclera to characterize its force condition; Step 5: Based on the SL fitting results, build a personalized SL fitting optimization strategy and recommend the most suitable SL model and parameters.

[0009] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics, the information acquisition module includes: an OCT optical coherence tomography scanner, a pressure sensor or a laser interferometer. The selection of the information acquisition module is not restricted. As long as it is a tool that can collect corneal and scleral images with high resolution, a pressure sensor that measures the pressure changes at different parts of the eyeball surface in real time or a laser interferometer that measures tiny deformations of the eyeball surface is suitable for the present invention.

[0010] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics, further, the content of step two includes: collecting a continuous image sequence of the cornea and sclera before and after micro-stimulation through the information acquisition module to obtain the displacement field information before and after the deformation; then using the AL-DIC algorithm to perform registration analysis on the OCT image, matching the image feature points through cross-correlation calculation, extracting the displacement field at the pixel level, calculating the strain, and constructing the strain distribution map.

[0011] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics, further, the content of step three includes: using the mechanical inversion method combined with the finite element method to simulate the force response of the tissue under a small excitation, and adjusting the material parameters through iterative optimization to minimize the error between the simulated displacement field and the experimentally measured displacement field; selecting a nonlinear elastic constitutive model suitable for soft tissue, calculating the elastic modulus, shear modulus and Poisson's ratio of the cornea and sclera, and providing accurate biomechanical parameter support for personalized SL fitting.

[0012] In the aforementioned personalized scleral lens intelligent fitting system based on biomechanics, as an embodiment, the nonlinear elastic constitutive model applicable to soft tissue is the Ogden model. The Ogden model is only a preferred embodiment and is not exhaustive.

[0013] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics, further, the content of step 4 includes: (1) Establishing SL finite element model database Simulate the shape and material properties of each SL model: Based on the scleral lens OCT image, the three-dimensional coordinates of each model of SL are obtained, and the function expression of the three-dimensional morphology of the SL surface is obtained by fitting the 10th-order Zernike polynomial; then the material parameters of the SL bending performance are obtained to describe the material properties of each model of SL, and finally the SL model grid is divided by C3D15H unit, and then the three-dimensional finite element model of each model of SL is established to form an SL finite element model database; the material parameter of the SL bending performance is the bending modulus; (2) Establish a personalized finite element model of the entire eye Firstly, based on the classic Gullstrand eye model data, the C3D15H unit was used to divide the mesh and establish an idealized whole eye model. Then, on the basis of the idealized model, the three-dimensional coordinates of the cornea and sclera of each human eye were obtained according to the large-scale anterior segment OCT images, and the coordinates of some nodes of the cornea and sclera in the model were modified to obtain a morphologically personalized whole eye finite element model.

[0014] Secondly, the boundary conditions of the model are set to limit the displacement of the scleral equatorial part in the Z direction and the displacement of the central node of the entire eyeball in the XY plane; the intraocular pressure is simulated by controlling the pressure freedom of the fluid cavity surrounded by the inner surface of the model. Each personalized model uses the biomechanically corrected intraocular pressure measurement value provided by Corvis ST, and the Ogden material model is used to simulate the material properties of the cornea and sclera. The equation expression of the Ogden material model is expressed as follows: , Where U is the strain potential energy, is the main stretch amount, μ and α represent the material parameters of the tissue, and μ and α of the personalized human eye are set according to the biomechanical parameters measured in step 3 for each subject; (3) Establish contact simulation between scleral lens and human eyeball Set the mechanical contact between the SL model and the whole eyeball model, set the friction contact mode between the SL and the human eye, the friction coefficient FRICTION is 0.05, the damping coefficient STABILIZE of the contact between the SL and the human eye is 1.0e-5, set the gravity according to the material and density of the SL, use the non-homogeneous pressure method to simulate the tear tension between the SL and the front surface of the cornea, and use the non-homogeneous pressure on the front surface of the SL to simulate the pressure of the eyelid on the SL; (4) Simulation of wearing process of scleral lens and human eyeball The three-dimensional coordinates of the anterior segment collected clinically are the shape of the eyeball under the action of intraocular pressure. The geometric shape of the eyeball under the condition of no intraocular pressure, that is, no stress, is obtained through reverse iteration. The solution is as follows: first, the three-dimensional coordinates are obtained according to the anterior segment image measured by OCT, that is, the shape under the action of intraocular pressure IOP, and a finite element model is established. The shape is recorded as X0. IOP is applied to the finite element model, and the node displacement vector u1 = X0 – x0 caused by IOP is calculated. Then the shape of the first estimated stress-free form can be expressed as X1 = X0 – u1; then in the second step, IOP is reloaded on the stress-free form X1, and the displacement error u2 = X0 – x1 between the loaded shape x1 and the target shape X0 is calculated, and then the shape X2 = X1 – u2 for subsequent iterative analysis is calculated. Repeat this process, while monitoring the node error value and distribution uk = X0 – xk-1, Xk = Xk-1 – uk relative to the target shape, until the error is less than 10-5, it is considered to have converged to the stress-free state; When the whole eyeball is in a stress-free state, the load and contact settings of the SL during the contact process between the SL and the human eye are taken as input conditions, and the simulated SL adaptation results are finally output; the loads include: the positive loading of the whole eyeball by the fluid cavity simulating the intraocular pressure, the tear tension between the SL and the cornea, and the pressure of the eyelid on the SL; the contact settings include: the friction, damping, and adhesive contact between the SL and the eyeball; the SL adaptation results include: the FR thickness map characterizing the SL sedimentation, and the output of the contact pressure and strain between the SL and the cornea to characterize its stress condition.

[0015] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics, further, the content of step five includes: Finite element analysis was used to simulate the effects of various SL parameters on FR thickness and force distribution, extract FR thickness gradient and contact pressure, and establish a standardized database; Optimize SL adaptation parameters by setting the optimal FR thickness range and force threshold; Based on the simulation data, a multivariate regression analysis was used to establish a FR thickness-force distribution regression model, and the relationship between SL parameters and FR thickness and force distribution was analyzed to provide a quantitative basis for parameter optimization. The FR thickness-force distribution regression model predicts the FR thickness and corneoscleral force under various SL parameters; AI optimization algorithm is used for multi-objective parameter optimization, which comprehensively balances FR thickness adaptation, contact area force, and wearing stability factors, and automatically recommends the optimal SL parameter combination; Output personalized FR thickness map and force assessment report to provide doctors with a scientific fitting basis.

[0016] Further, the SL parameters include: diameter, base arc, and sag of each zone; further, the AI ​​optimization algorithm includes: genetic algorithm GA or Bayesian optimization BO. The enumeration here is only an example of an embodiment and is not exhaustive.

[0017] The aforementioned personalized scleral lens intelligent fitting system based on biomechanics also includes: 6. Clinical data verification prediction plan.

[0018] An application of a personalized scleral lens intelligent fitting system based on biomechanics is applied to a computer-readable storage medium equipped with the personalized scleral lens intelligent fitting system as described above, and a device provided with the computer-readable storage medium.

[0019] The present invention is beneficial in that: This invention applies DIC technology (digital image correlation analysis technology) to SL (scleral lens) fitting analysis for the first time, filling the gap in existing scleral material property measurement; the strain characteristics extracted by DIC are applied to SL fitting, expanding the mechanical information of SL fitting and reducing the complications of SL wearing; The present invention combines the strain information extracted by DIC with the inversion algorithm to reversely infer the material properties of eye tissues (such as cornea, sclera, and corneosclera); by establishing a finite element model of the eyeball and a finite element model database of the scleral lens, and combining individualized parameters for mechanical simulation and prediction, it ensures that the scleral lens fitting plan for each patient can meet the specific needs of their eyes to the greatest extent; The personalized SL finite element model established by the present invention simulates the contact, load action and deformation process between the scleral lens and the eyeball; by simulating the wearing stability of different types of scleral lenses, analyzing the tear thickness under the lens, the force and deformation of the cornea and sclera, the most suitable scleral lens model can be predicted based on biomechanical changes, thereby optimizing the wearing effect; The scleral lens fitting and evaluation method established by the present invention helps to improve the fitting success rate, reduce the patient's waiting time and lens replacement rate, which is of far-reaching significance to doctors and patients, and is also of great benefit to the promotion of SL in my country, and has far-reaching social and economic benefits.

[0020] Explanation of special terms: Augmented Lagrangian-DIC (AL-DIC) is a hybrid DIC algorithm that combines the advantages of Local Subset DIC and Global DIC. It uses the Augmented Lagrangian multiplier method to ensure global kinematic consistency while maintaining the fast computational speed and parallel computing capability of Local Subset DIC.

[0021] Scleral lenses, whose full name is hard gas permeable scleral contact lenses, have large diameter lenses that are arched like a bridge, covering the cornea and the limbus of the cornea, spanning over the cornea, and finally fitting to the conjunctiva and sclera.

[0022] Optical Coherence Tomography (OCT) uses the principle of low-coherence interference to perform high-resolution imaging of biological tissues using single scattered light and weak scattered light. When the optical path difference between the single scattered light and the forward scattered light returned in the scattering medium is within the coherence length of the light source, they interfere with the reference light to form a signal that can be used to calculate the tomographic image. However, since the optical path difference of multiple scattered light (diffuse light) exceeds the coherence length, it cannot form an interference signal and is therefore not recorded by OCT. In this way, OCT can achieve non-invasive, high-resolution tomographic imaging of biological tissues and is widely used in ophthalmology (such as retinal and corneal imaging), cardiovascular, neuroscience, dermatology, and dentistry.

[0023] The anterior segment includes the cornea, iris, ciliary body, anterior chamber, posterior chamber, angle, zonule, and some definitions also include the anterior lens. In some cases, the anterior vitreous, conjunctiva, anterior sclera, and extraocular muscle attachments may also be involved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a schematic diagram of OCT images when first wearing glasses; Figure 2 This is a schematic diagram of the image after wearing the glasses for 4 hours; Figure 3 is a schematic diagram of the image after segmentation; Figure 4 is a schematic diagram of the image after segmentation; Figure 5 is a schematic diagram of displacement in the x direction; Figure 6 is a schematic diagram of displacement in the y direction; Figure 7 is to construct the full-field strain distribution map; Figure 8 Schematic diagram of scleral lens coordinates based on swept frequency OCT images (left) and schematic diagram of scleral lens three-dimensional finite element model (right) of the present invention; Fig. 9 Schematic diagram of the cornea and sclera coordinates based on the swept frequency OCT image of the present invention (left); Schematic diagram of the finite element model of the whole eyeball (right); Fig.10 is a schematic diagram of a simulation of the contact between the scleral lens of the present invention and the eyeball of a human eye; Fig.11 is the FR thickness diagram of the present invention; Fig.12is the simulated contact pressure and eyeball strain result of the contact area between the scleral lens of the present invention and the eyeball; Fig.13 The clinical FR thickness diagram when the glasses are just worn in the verification test of the present invention (left); the model FR thickness diagram when the glasses are just worn (right); Fig.14 This is the von Mises strain field in the horizontal direction from 0 to 4 hours of wearing the glasses in the verification test of the present invention (left); and the von Mises strain field in the vertical direction from 0 to 4 hours of wearing the glasses (right). DETAILED DESCRIPTION

[0025] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] A personalized scleral lens intelligent fitting system based on biomechanics, including the following contents: Step one, using an information acquisition module to collect information of the eyeball and sclera; as a preferred embodiment, the information acquisition module is an OCT optical coherence tomography scanner; the choice of the information acquisition module is not limited, and it can also be a large-scale anterior segment OCT, CLARUSM 500, a high-resolution fundus camera, Pentacam, etc. that can collect high-resolution images of the cornea and sclera are all suitable for the present invention.

[0027] Step 2: Use optical coherence tomography (OCT) to obtain high-resolution images of the cornea and sclera under tiny excitations (external piezoelectric loading or physiological regulatory response), and combine Augmented Lagrangian digital image correlation AL-DIC or DIC algorithm technology and mechanical inversion methods to calculate individualized corneal and scleral material properties. First, use a wide range of anterior segment OCT to collect continuous image sequences of the cornea and sclera before and after tiny excitation to obtain displacement field information before and after deformation. Then, use the AL-DIC algorithm to perform registration analysis on the OCT images, match image feature points through cross-correlation calculation, extract pixel-level displacement fields, calculate strains, and construct strain distribution maps.

[0028] Figure 1 This is the OCT image when I first put on the glasses. Figure 2 This is the image after wearing glasses for 4 hours. First, the cornea and sclera are segmented ( Figure 3 , 4 is the segmented image), zero-mean normalized cross-correlation calculation is performed through fast Fourier transform, and pixel-level matching is performed in the subset area based on multi-scale search of the image pyramid to obtain the initial displacement estimation before and after deformation ( Figure 5 , 6are the displacements in the x-direction and y-direction, respectively). Subsequently, in the local inverse combined Gauss-Newton iterative optimization process, the alternating direction multiplier method is used for global optimization. The strain calculation can be performed using the finite difference or finite element method to perform gradient calculations on the displacement field, obtain the Green-Lagrange strain tensor, and construct the full-field strain distribution map ( Figure 7 ).

[0029] It should be noted that: the information acquisition module includes: an OCT optical coherence tomography scanner, a pressure sensor or a laser interferometer. The information acquisition module can be an OCT optical coherence tomography scanner, and the image correlation method extracts the strain data of the surface of the eyeball by tracking the feature points in the image. Sensor technologies such as pressure sensors can also be used as an alternative. By installing miniature pressure sensors on the scleral lens and the surface of the eyeball, the pressure changes in different parts of the surface of the eyeball can be measured in real time, thereby indirectly calculating the strain distribution, which is suitable for situations requiring high-frequency sampling and real-time monitoring. In addition, the laser interferometer can also be used as an alternative technology to measure the tiny deformation of the surface of the eyeball and calculate the corresponding strain conditions, which is suitable for research scenarios with higher precision requirements. All alternative solutions that adopt a design concept of the present invention are inspired by the present invention and are within the scope of protection of the present invention.

[0030] Step three, the mechanical inversion method is combined with the finite element method to simulate the force response of the tissue under the action of micro-stimulation, and the material parameters are adjusted through iterative optimization to minimize the error between the simulated displacement field and the experimentally measured displacement field; the nonlinear elastic constitutive model (Ogden model) suitable for soft tissue is selected to calculate the elastic modulus, shear modulus and Poisson's ratio of the cornea and sclera, providing accurate biomechanical parameter support for personalized SL fitting. In the mechanical inversion process, the finite element model of the cornea and sclera is first established based on the experimentally measured AL-DIC displacement field data, and the Ogden hyperelastic constitutive model is used to describe the nonlinear mechanical behavior of the tissue. The initial parameters set the material parameters 𝜇 and 𝛼 of the cornea and sclera according to age. Subsequently, the same load as the experiment is applied to the finite element model, and the displacement field obtained by numerical simulation is calculated. The error between the displacement field measured experimentally and the simulated displacement field is calculated , the Levenberg-Marquardt optimization method is used to adjust the material parameters so that the error converges gradually. During the optimization process, the shear modulus and exponential parameters are updated in each iteration, and the step size is =0.1 for correction, and the update method is , ,After 5-8 iterations, when the error meets the convergence condition, the optimized cornea and sclera material parameters are obtained.

[0031] The mechanical inversion method can also be replaced by optical coherence elastography (OCE) technology, which performs a deep scan of the high-resolution image of the surface of the eyeball, observes the geometric deformation of the eyeball in real time, and indirectly infers the mechanical properties. In addition, the mechanical parameters can be inferred by the finite element iteration method. This method inputs the known geometric shape, load, and boundary conditions into the finite element model, combines multiple iterative calculations, and gradually optimizes the material constitutive parameters until the simulation results are consistent with the experimental data. It is suitable for situations where material parameters are not completely known, and has good predictive ability under certain conditions. All alternative solutions that adopt a design thinking of the present invention are inspired by the present invention and are within the scope of protection of the present invention.

[0032] Step 4: Construct SL-personalized full-eye finite element model and simulate the wearing process (1) Establishing SL finite element model database Establishing the SL finite element model database requires simulating the morphology and material properties of different types of SL. First, based on the scleral lens OCT image, the three-dimensional coordinates of different types of SL are obtained (Figure 8 left), and the functional expression of the three-dimensional morphology of the SL surface is obtained by fitting the 10th-order Zernike polynomial. Secondly, SL generally uses a hard material with high oxygen permeability, and the material parameter of its bending resistance, the bending modulus, is obtained to describe the material properties of different types of SL. Finally, the C3D15H unit is used to divide the SL model grid, and then the three-dimensional finite element models of different types of SL are established (Figure 8 right), forming an SL finite element model database.

[0033] (2) Establish a personalized finite element model of the entire eye First, based on the classic Gullstrand eyeball model data, the C3D15H unit was used to divide the mesh and establish an idealized whole eyeball model. On the basis of the idealized model, the three-dimensional coordinates of the cornea and sclera of different human eyes were obtained according to the large-scale anterior segment OCT images (Figure 9 left), and the coordinates of some nodes of the cornea and sclera in the model were modified to obtain a morphologically personalized whole eyeball finite element model (Figure 9 right).

[0034] Secondly, the boundary conditions of the model are set to limit the displacement of the scleral equatorial part in the Z direction and the displacement of the central node of the entire eyeball in the XY plane; the intraocular pressure is simulated by controlling the pressure freedom of the fluid cavity surrounded by the inner surface of the model. Each personalized model uses the biomechanically corrected intraocular pressure measurement value provided by Corvis ST, and the Ogden material model is used to simulate the material properties of the cornea and sclera. The equation expression of the Ogden material model is expressed as follows: , Where U is the strain potential energy, is the main stretch amount, μ and α represent the material parameters of the tissue, and μ and α of the personalized human eye are set according to the biomechanical parameters measured in step 3 for each subject; (3) Establish contact simulation between scleral lens and human eyeball Set up the mechanical contact between the SL model and the whole eyeball model, such as Fig.10 As shown in the figure, a friction contact mode (Cohesive Contact) is set between SL and the human eye, the friction coefficient FRICTION is 0.05, and the damping coefficient STABILIZE of the contact between the two is set to 1.0e-5. In addition, the gravity is set according to the material and density of SL. There is tear tension between SL and the anterior surface of the cornea. This study intends to simulate it using a non-homogeneous pressure method, that is, the pressure value takes into account the influence of tear gravity. Similarly, the pressure of the eyelid on SL is simulated as a non-homogeneous pressure on the anterior surface of SL, taking into account the difference in tension between the upper and lower eyelids.

[0035] (4) Simulation of wearing process of scleral lens and human eyeball The three-dimensional coordinates of the anterior segment collected clinically are the shape of the eyeball under the action of intraocular pressure. The geometric shape of the eyeball under the condition of no intraocular pressure, that is, no stress, is obtained through reverse iteration. The solution is as follows: first, the three-dimensional coordinates are obtained according to the anterior segment image measured by OCT, that is, the shape under the action of intraocular pressure IOP, and a finite element model is established. The shape is recorded as X0. IOP is applied to the finite element model, and the node displacement vector u1 = X0 – x0 caused by IOP is calculated. Then the shape of the first estimated stress-free form can be expressed as X1 = X0 – u1; then in the second step, IOP is reloaded on the stress-free form X1, and the displacement error u2 = X0 – x1 between the loaded shape x1 and the target shape X0 is calculated, and then the shape X2 = X1 – u2 for subsequent iterative analysis is calculated. Repeat this process, while monitoring the node error value and distribution uk = X0 – xk-1, Xk = Xk-1 – uk relative to the target shape, until the error is less than 10-5, it is considered to have converged to the stress-free state; In the stress-free state of the whole eyeball, the load and contact settings of the SL and the human eye during the contact process are used as input conditions, and the simulated SL adaptation results are finally output; the load includes: the positive load of the whole eyeball simulated by the fluid cavity intraocular pressure, the tear tension between the SL and the cornea, and the pressure of the eyelid on the SL; the contact settings include: the friction, damping, and adhesive contact between the SL and the eyeball; the SL adaptation results include: the FR thickness map represents the SL sedimentation ( Fig.11 ), and the contact pressure and strain between the output SL and the cornea and sclera to characterize the stress conditions ( Fig.12 ). Step 5: Based on the SL fitting results, build a personalized SL fitting optimization strategy and recommend the most suitable SL model and parameters: Finite element analysis was used to simulate the effects of various SL parameters on FR thickness and force distribution, extract FR thickness gradient and contact pressure, and establish a standardized database; Optimize SL fitting parameters by setting the optimal FR thickness range and force threshold: First, based on finite element analysis, the FR thickness distribution under different SL parameters, anterior segment morphology, and material parameters was simulated, and the thickness gradients of the central and peripheral zones were extracted. The optimal range was set in combination with clinical experience, with the recommended thickness of 150-300 μm in the central zone and 50-100 μm in the peripheral zone to maintain good optical quality and tear exchange. Subsequently, the 360° contact stress distribution between SL and sclera was calculated, and a personalized force threshold was established through parametric learning, and the stress intervals of different adaptation states were defined in a data-driven manner. First, the stress distribution under different SL parameters and eye conditions was collected, and the mapping relationship between SL parameters and maximum contact stress was constructed using regression modeling (such as XGBoost and random forest). The reasonable force range was delineated based on clinical practice and statistical analysis, where the comfortable wearing range corresponds to the 25%-50% quantile of contact stress, the alert range falls in the 50%-90% quantile, and the area exceeding the 90% quantile may cause corneal hypoxia or scleral tissue stress concentration, which is not suitable for long-term wearing. Based on this optimization scheme, when the FR thickness exceeds the optimal range and affects the optical quality, the SL diameter or sagittal height can be adjusted. When the contact stress exceeds the warning or critical range, the SL edge design can be optimized to ensure that the personalized SL fit achieves a balance between the optimal FR thickness and biomechanical stability.

[0036] Based on the simulation data, a multivariate regression analysis was used to establish a FR thickness-force distribution regression model, and the relationship between SL parameters and FR thickness and force distribution was analyzed to provide a quantitative basis for parameter optimization. The FR thickness-force distribution regression model predicts the FR thickness and corneoscleral force under various SL parameters; AI optimization algorithm is used for multi-objective parameter optimization, comprehensively balancing FR thickness adaptation, contact area force, and wearing stability factors, and automatically recommending the optimal SL parameter combination; SL parameters include: diameter, base curve, and sagittal height of each area; AI optimization algorithms include: genetic algorithm GA or Bayesian optimization BO; As an embodiment, Bayesian optimization is used to optimize the multi-objective parameters of scleral lens SL, a proxy model is constructed through Gaussian process regression (GPR), and the expectation improvement (EI) strategy is used to optimize the SL parameter combination so that the FR thickness adaptation, contact stress balance and wearing stability are optimally balanced. First, the optimization objective function is defined, including the FR thickness adaptation error in the central and peripheral zones, the maximum stress value deviation in the contact area, and the wearing stability deviation, and the SL diameter, base curve and sagittal height of each zone are used as optimization variables. During the optimization process, BO randomly selects the initial SL parameter combination for finite element calculation, trains the GPR proxy model to predict the objective function value under different parameters, and selects the parameter combination that is most likely to improve the result through the EI strategy for iterative update. After each round of iteration, new experimental data is added to the data set, and the optimization model is continuously adjusted and gradually converges to the optimal solution. Finally, after about 30 iterations, BO was able to find the optimal SL parameter combination, keeping the FR thickness in the central zone at 150-300 μm and the thickness in the peripheral zone at 50-100 μm, while keeping the contact stress within the comfortable wearing range, ensuring long-term wearing comfort and biomechanical stability.

[0037] Output personalized FR thickness map and force assessment report to provide doctors with scientific fitting basis; Step six: Use clinical data to verify the prediction scheme, optimize the fitting strategy, improve the feasibility of SL in clinical applications, and achieve accurate fitting and intelligent promotion.

[0038] An application of a personalized scleral lens intelligent fitting system based on biomechanics is applied to a computer-readable storage medium equipped with the personalized scleral lens intelligent fitting system as described above, and a device provided with the computer-readable storage medium.

[0039] The device includes: memory, processor, input and output devices, and communication interface.

[0040] The memory includes: high-speed random access memory (RAM), non-volatile memory, etc. Any memory with a computer-readable storage medium can be applied to the present invention. The communication interface can be the Internet, a wide area network, a local area network, a metropolitan area network, etc.

[0041] The processor includes a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0042] Accuracy experimental verification: The preliminary scleral lens images and patient database of the SL fitting model have been established, and the SL-whole eyeball finite element modeling has been basically completed. The FR central thickness and gradient distribution trend output by the finite element model when the glasses are just worn are highly consistent with the clinical observation results (Figure 13). The contact pressure of the contact area simulated by the model and the strain distribution of the eyeball are as follows Fig.12 As shown in the figure, the force in the contact area is unevenly distributed. The AL-DIC method has been developed to analyze clinical OCT images. It was found that from 0 hours to 4 hours after wearing the glasses, the strain in the nasal area was higher than that in the temporal area, and the upper strain was higher than the lower strain. This trend is consistent with the force distribution results output by the finite element model, which verifies the accuracy and applicability of the model in predicting the strain characteristics of the eye during SL wearing (Figure 14). The SL fitting simulation model has taken shape, laying a model foundation for the development of the project.

[0043] In summary, the present invention proposes a personalized scleral lens intelligent fitting method based on biomechanics. The present invention integrates digital image correlation (DIC), mechanical inversion method and finite element analysis, and accurately predicts the fitting effect of different models of SL from the perspective of personalized biomechanics of patients, outputs FR thickness distribution and contact area force conditions, and realizes scientific, efficient and personalized SL fitting scheme recommendation. The present invention promotes the development of SL fitting from experience-driven to data-driven and precise, and improves the objectivity, accuracy and scientificity of fitting.

[0044] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A personalized scleral lens intelligent fitting system based on biomechanics, characterized in that: It includes the following: Step 1, using an information collection module to collect information about the anterior segment of the eye; Step 2: Record the changes of the anterior segment information by micro-stimulation, extract the local strain characteristics of the cornea and sclera by combining digital image correlation (DIC) technology, and analyze the biomechanical response under micro-deformation; Step 3: Based on the biomechanical response analysis under small deformation, mechanical inversion method, optical coherence elastography (OCE) technology or finite element iteration method are used to reversely calculate the mechanical parameters and material properties of the cornea and sclera; Step 4: Establish a SL finite element model database for scleral lenses, establish a personalized full-eye finite element model, establish a contact simulation between the scleral lens and the human eye, simulate the wearing process of the scleral lens and the human eye as a whole, obtain the SL finite element simulation model, and finally output the SL adaptation results obtained by simulation; the SL adaptation results include: FR thickness map characterizing the SL settlement, outputting the contact pressure and strain between the SL and the cornea and sclera to characterize its stress condition; Step 5: Based on the SL fitting results, build a personalized SL fitting optimization strategy and recommend the most suitable SL model and parameters.

2. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: The information acquisition module includes: an OCT optical coherence tomography scanner, a pressure sensor or a laser interferometer.

3. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: The content of step 2 includes: collecting a continuous image sequence of the cornea and sclera before and after micro-excitation through the information acquisition module to obtain the displacement field information before and after deformation; then using the AL-DIC algorithm to perform registration analysis on the OCT image, matching image feature points through cross-correlation calculation, extracting the displacement field at the pixel level, calculating the strain, and constructing a strain distribution map.

4. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: The contents of step three include: using the mechanical inversion method combined with the finite element method to simulate the force response of the tissue under a small excitation, and adjusting the material parameters through iterative optimization to minimize the error between the simulated displacement field and the experimentally measured displacement field; selecting a nonlinear elastic constitutive model suitable for soft tissue, calculating the elastic modulus, shear modulus and Poisson's ratio of the cornea and sclera, and providing accurate biomechanical parameter support for personalized SL fitting.

5. The biomechanically based personalized scleral lens intelligent fitting system according to claim 4, characterized in that: The nonlinear elastic constitutive model suitable for soft tissue is the Ogden model.

6. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: The content of step 4 includes: (1) Establishing SL finite element model database Simulate the morphology and material properties of each SL model: Based on the scleral lens OCT image, the three-dimensional coordinates of each model of SL are obtained, and the function expression of the three-dimensional morphology of the SL surface is obtained by fitting the 10th-order Zernike polynomial; then the material parameters of the SL bending performance are obtained to describe the material properties of each model of SL, and finally the SL model grid is divided by C3D15H unit, and then the three-dimensional finite element model of each model of SL is established to form an SL finite element model database; the material parameter of the SL bending performance is the bending modulus; (2) Establish a personalized finite element model of the entire eye Firstly, based on the classic Gullstrand eyeball model data, the C3D15H unit was used to divide the mesh and establish an idealized whole eyeball model. Then, based on the idealized model, the three-dimensional coordinates of the cornea and sclera of each human eye were obtained according to the anterior segment OCT image, and the coordinates of some nodes of the cornea and sclera in the model were modified to obtain a morphologically personalized whole eyeball finite element model. Secondly, the boundary conditions of the model are set to limit the displacement of the scleral equatorial part in the Z direction and the displacement of the central node of the entire eyeball in the XY plane; the intraocular pressure is simulated by controlling the pressure freedom of the fluid cavity surrounded by the inner surface of the model. Each personalized model uses the biomechanically corrected intraocular pressure measurement value provided by Corvis ST, and the Ogden material model is used to simulate the material properties of the cornea and sclera. The equation expression of the Ogden material model is expressed as follows: ， in U is the strain potential energy, The main stretch amount, μ and α Represents the material parameters of tissue, personalizing the human eye μ and α Set according to the biomechanical parameters measured in step three for each subject; (3) Establish contact simulation between scleral lens and human eyeball Set the mechanical contact between the SL model and the whole eyeball model, set the friction contact mode between the SL and the human eye, the friction coefficient FRICTION is 0.05, the damping coefficient STABILIZE of the contact between the SL and the human eye is 1.0e-5, set the gravity according to the material and density of the SL, use the non-homogeneous pressure method to simulate the tear tension between the SL and the front surface of the cornea, and use the non-homogeneous pressure on the front surface of the SL to simulate the pressure of the eyelid on the SL; (4) Simulation of wearing process of scleral lens and human eyeball The three-dimensional coordinates of the anterior segment collected clinically are the shape of the eyeball under the action of intraocular pressure. The geometric shape of the eyeball under the condition of no intraocular pressure, that is, no stress, is obtained through reverse iteration. The solution is as follows: first, the three-dimensional coordinates, that is, the shape under the action of intraocular pressure IOP, are obtained based on the anterior segment image measured by OCT, and a finite element model is established. The shape is recorded as X0, and IOP is applied to the finite element model. The node displacement vector u1 = X0 – x0 caused by IOP is calculated, and the shape of the stress-free form estimated for the first time can be expressed as X1 = X0 – u1; then in the second step, IOP is reloaded on the stress-free form X1, and the displacement error u2 = X0 – x1 between the loaded shape x1 and the target shape X0 is calculated, and then the shape X2 = X1– u2 for subsequent iterative analysis is calculated; Repeat this process while monitoring the node error value and distribution u relative to the target shape k = X0 – x k-1 , X k =X k-1 – u k , until the error is less than 10 -5 , it is considered to have converged to a stress-free state; When the whole eyeball is in a stress-free state, the load and contact settings of the SL during the contact process between the SL and the human eye are used as input conditions, and the simulated SL adaptation results are finally output; the load includes: the positive loading of the whole eyeball by the fluid cavity simulating the intraocular pressure, the tear tension between the SL and the cornea, and the pressure of the eyelid on the SL; the contact settings include: friction, damping, and adhesive contact between the SL and the eyeball; the SL adaptation results include: the FR thickness map characterizing the SL sedimentation, and the output of the contact pressure and strain between the SL and the cornea to characterize its stress condition.

7. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: The content of step five includes: Finite element analysis was used to simulate the effects of various SL parameters on FR thickness and force distribution, extract FR thickness gradient and contact pressure, and establish a standardized database; Optimize SL adaptation parameters by setting the optimal FR thickness range and force threshold; Based on the simulation data, a multivariate regression analysis was used to establish a FR thickness-force distribution regression model, and the relationship between SL parameters and FR thickness and force distribution was analyzed to provide a quantitative basis for parameter optimization. The FR thickness-force distribution regression model predicts the FR thickness and corneoscleral force under various SL parameters; AI optimization algorithm is used for multi-objective parameter optimization, which comprehensively balances FR thickness adaptation, contact area force, and wearing stability factors, and automatically recommends the optimal SL parameter combination; Output personalized FR thickness map and force assessment report to provide doctors with a scientific fitting basis.

8. The biomechanically based personalized scleral lens intelligent fitting system according to claim 7, characterized in that: The SL parameters include: diameter, base arc, and sagittal height of each zone; the AI ​​optimization algorithm includes: genetic algorithm GA or Bayesian optimization BO.

9. The biomechanically based personalized scleral lens intelligent fitting system according to claim 1, characterized in that: Also includes: Step six, clinical data verification of the prediction scheme.

10. An application of a personalized scleral lens intelligent fitting system based on biomechanics, characterized in that: A computer-readable storage medium used for carrying the personalized scleral lens intelligent fitting system as claimed in claim 1, and a device provided with the computer-readable storage medium.

Images