A method, system and device for monitoring intraocular pressure after refractive surgery

Abstract

The present invention discloses a method, system and device for monitoring intraocular pressure after refractive surgery, and belongs to the field of medical auxiliary technology. The present invention combines finite element simulation with a multi-layer perceptron neural network, and proposes a prediction model for intraocular pressure after refractive surgery. Based on the prediction model, the postoperative corneal geometric configuration parameters can be obtained through existing medical instruments and refractive surgery design parameters, and the corneal geometric configuration parameters and the intraocular pressure before the operation are used as input values ​​of the prediction model to obtain the relationship curve between the corneal central vertex height and the intraocular pressure. Based on the obtained relationship curve, the intraocular pressure after refractive surgery can be obtained by monitoring the patient's corneal vertex height, thereby solving the problem that the postoperative intraocular pressure is difficult to accurately monitor by a tonometer.

Description

Technical Field

[0001] The present invention belongs to the field of medical auxiliary systems, and in particular relates to a method, system and device for monitoring intraocular pressure after refractive surgery. Background Art

[0002] Intraocular pressure refers to the pressure exerted by the contents of the eyeball on the wall of the eyeball. Stable intraocular pressure depends on the dynamic balance of the amount of ocular contents, the rate of aqueous humor production, and the rate of aqueous humor excretion. Due to impaired aqueous humor circulation, the accumulation of fluid in the eye leads to an increase in intraocular pressure, which usually leads to glaucoma.

[0003] In recent years, high myopia has been considered a risk factor for primary open-angle glaucoma. Refractive surgery is the most commonly used method for treating myopia. The four commonly used types of refractive surgery are all based on cutting the corneal stroma. The surgical instrument platform will determine the cutting depth, cutting radius and curvature in the corneal stroma based on the analysis and calculation of the various visual parameters and optical refraction of myopic patients, so that the central cornea appears flat, thereby changing the refractive index. After refractive surgery, due to the lack of corneal stroma, the thickness of the central cornea becomes thinner, resulting in the Imbert-Fick principle commonly used in tonometers: Pt (intraocular pressure) = W (applanation corneal external force) / A (applanation area), which is affected by use, resulting in large errors in the intraocular pressure values ​​measured by applanation tonometers (GAT) and non-contact tonometers (NCT) after refractive surgery. Therefore, correcting postoperative intraocular pressure measurement has become an important issue for patients after refractive surgery to prevent glaucoma. When evaluating postoperative intraocular pressure, the influence of multiple factors such as preoperative intraocular pressure, central corneal thickness (CCT), cutting depth, and corneal curvature should be fully considered.

[0004] At present, the safety, effectiveness and predictability of corneal refractive surgery have been recognized by the majority of myopic people, but the correction formula for postoperative intraocular pressure has not yet been unified. How to achieve rapid monitoring of intraocular pressure after refractive surgery through existing medical instruments and refractive surgery design parameters is a technical problem that needs to be solved urgently. Summary of the invention

[0005] The purpose of the present invention is to solve the technical problem that it is difficult to quickly monitor intraocular pressure after refractive surgery in the prior art, and to provide a method, system and device for monitoring intraocular pressure after refractive surgery.

[0006] The specific technical solutions adopted by the present invention are as follows:

[0007] In a first aspect, the present invention provides a method for monitoring intraocular pressure after refractive surgery, comprising:

[0008] S1. Aiming at the postoperative goal of laser ablation refractive surgery, the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and preoperative intraocular pressure IOP measured by tonometer are obtained by optical instrument measurement, and used as inputs of the first neural network prediction model to obtain the corneal initial vertex height S0 when there is no intraocular pressure, the corneal initial thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the corneal initial vertex;

[0009] S2. Sample a series of intraocular pressure sampling values ​​IOP from the value range of intraocular pressure IOP i , respectively, different intraocular pressure sampling values ​​IOP i The initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the corneal radius R0 are combined as the input of the second neural network prediction model to obtain different intraocular pressure sampling values ​​IOP. i The corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se are used to establish a mapping relationship between the postoperative target intraocular pressure IOP and the postoperative corneal vertex height S;

[0010] S3. For the postoperative target after laser ablation refractive surgery, obtain the postoperative corneal vertex height S measured by an optical instrument at the time to be monitored, and call the mapping relationship to obtain the postoperative intraocular pressure IOP corresponding to the time to be monitored.

[0011] As a preferred embodiment of the above-mentioned first aspect, the first neural network prediction model is formed by using an MLP neural network trained with a first training set.

[0012] As a preferred embodiment of the first aspect, the method for generating training samples in the first training set is: different simulation input parameters are obtained by Latin hypercube sampling within the parameter value ranges of the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the central corneal fovea and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the intraocular pressure IOP. Each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se, and then the sample input and sample labels required for training the first neural network prediction model are constructed according to the finite element simulation results.

[0013] As a preferred embodiment of the above-mentioned first aspect, the second neural network prediction model is formed by using an MLP neural network trained with a second training set.

[0014] As a preferred embodiment of the first aspect, the method for generating training samples in the second training set is: different simulation input parameters are obtained by Latin hypercube sampling within the parameter value ranges of the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the central corneal fovea and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the intraocular pressure IOP. Each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se, and then the sample input and sample labels required for training the second neural network prediction model are constructed according to the finite element simulation results.

[0015] As a preferred embodiment of the above-mentioned first aspect, when finite element simulation is performed on each set of simulation input parameters, based on the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the central corneal fovea and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex, the initial configuration of the cornea after surgery in the absence of intraocular pressure is constructed by finite element method, and then the intraocular pressure IOP is applied to perform finite element simulation to obtain the postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se corresponding to this set of simulation input parameters.

[0016] In a second aspect, the present invention provides a system for monitoring intraocular pressure after refractive surgery, comprising:

[0017] The corneal initial configuration parameter prediction module is used to obtain the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and intraocular pressure IOP measured by a tonometer measured by an optical instrument for the postoperative target after laser ablation refractive surgery, and use them as inputs of the first neural network prediction model to obtain the corneal initial vertex height S0 when there is no intraocular pressure, the corneal initial thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the corneal initial vertex;

[0018] The intraocular pressure and postoperative corneal vertex height mapping module is used to sample a series of intraocular pressure sampling values ​​IOP from the value range of intraocular pressure IOP i , respectively, different intraocular pressure sampling values ​​IOP iThe initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the corneal radius R0 are combined as the input of the second neural network prediction model to obtain different intraocular pressure sampling values ​​IOP. i The corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se are used to establish a mapping relationship between the postoperative target intraocular pressure IOP and the postoperative corneal vertex height S;

[0019] The real-time prediction module for postoperative intraocular pressure is used to obtain the postoperative corneal vertex height S measured by an optical instrument at the time to be monitored for the postoperative target after laser ablation refractive surgery, and call the mapping relationship to obtain the postoperative intraocular pressure IOP corresponding to the time to be monitored.

[0020] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for monitoring intraocular pressure after refractive surgery as described in any one of the schemes in the first aspect above is implemented.

[0021] In a fourth aspect, the present invention provides a computer electronic device comprising a memory and a processor;

[0022] The memory is used to store computer programs;

[0023] The processor is used to implement the method for monitoring intraocular pressure after refractive surgery as described in any one of the schemes in the first aspect above when executing the computer program.

[0024] In a fifth aspect, the present invention provides a device for monitoring intraocular pressure after refractive surgery, which includes an optical instrument, a detection module and a result display module;

[0025] The optical instrument is used to measure in real time the postoperative corneal vertex height S of the postoperative target after laser ablation refractive surgery;

[0026] The detection module is used to receive the postoperative corneal vertex height S measured by the optical instrument, and according to the method for monitoring intraocular pressure after refractive surgery described in any one of the schemes of the first aspect above, call the mapping relationship to obtain the corresponding postoperative intraocular pressure IOP;

[0027] The result display module is used to locally display or remotely upload and display the postoperative intraocular pressure IOP detected by the detection module.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The present invention uses the created prediction model (model A) to determine the initial configuration of the cornea when a set of corneal vertex height, thickness and IOP are measured with a tonometer. Combined with the prediction model for the vertex height and thickness of the cornea at a specific IOP (model B), the IOP can be determined as long as the vertex height of the cornea is measured.

[0030] 2. For the cornea before refractive surgery, once a set of vertex height, thickness and IOP are measured by medical instruments, IOP can be predicted by using the instrument to determine the corneal vertex height, avoiding damage to the cornea caused by repeated corneal compression by long-term use of GAT / NCT tonometer. In addition, by measuring the vertex height of the cornea, IOP can be continuously monitored.

[0031] 3. This study simplified the initial configuration of the cornea without IOP to an arched membrane with uniform thickness. The geometric parameters, vertex height and thickness of the cornea under IOP predicted by the prediction model are in good agreement with the geometric parameters of volunteers measured by medical optical instruments, verifying the assumptions of the initial configuration. In addition, the initial configuration of the cornea without IOP can be further used to study other issues, such as stress distribution.

[0032] 4. Based on the prediction model (model C) of the geometric parameters of the virtual initial configuration during surgery and the model (model D) of the geometric parameters of the cornea after surgery, an effective method for predicting corneal IOP after refractive surgery is provided, solving the problem that there is no effective means to measure IOP after refractive surgery.

[0033] 5. As long as there is an instrument that can detect the displacement of the corneal vertex, it is possible to predict the intraocular pressure of patients after refractive surgery, thereby preventing the risk of glaucoma in patients in the later stage.

[0034] The present invention combines finite element simulation with MLP neural network algorithm, and proposes a prediction model for intraocular pressure after refractive surgery. Based on the prediction model, the geometric parameters of cornea after surgery can be obtained through existing medical instruments and refractive surgery design parameters, and the corneal geometric configuration parameters and the intraocular pressure before surgery are used as input values ​​of the prediction model to obtain the relationship curve between the corneal central vertex height and the intraocular pressure. Using the relationship curve obtained by this method, the intraocular pressure after refractive surgery can be obtained by monitoring the patient's corneal vertex height in the future, thereby solving the problem that the intraocular pressure after surgery is difficult to accurately monitor by a tonometer. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 Schematic diagram of the initial corneal configuration before surgery without intraocular pressure;

[0036] Figure 2 Schematic diagram of the preoperative corneal configuration in the presence of intraocular pressure;

[0037] Figure 3 Schematic diagram of the initial corneal configuration after surgery without intraocular pressure;

[0038] Figure 4 Schematic diagram of the postoperative corneal configuration in the presence of intraocular pressure;

[0039] Figure 5 A flowchart of establishing a mapping relationship between the preoperative corneal vertex height S and IOP in an embodiment of the present invention;

[0040] Figure 6 A schematic diagram of a finite element model constructed for the initial configuration of the cornea before surgery in an embodiment of the present invention;

[0041] Figure 7 A flowchart of establishing a mapping relationship between postoperative corneal vertex height S and IOP in an embodiment of the present invention;

[0042] Figure 8 It is a schematic diagram of a finite element model constructed for the initial configuration of the cornea after surgery in an embodiment of the present invention;

[0043] Fig. 9 This is a result diagram of predicting the relationship between S and IOP of volunteers before and after refractive surgery in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purpose, features and advantages of the present invention more obvious and easy to understand, the specific implementation mode of the present invention is described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. The technical features in each embodiment of the present invention can be combined accordingly without conflicting with each other.

[0045] In the description of the present invention, it should be understood that the terms "first" and "second" are only used for the purpose of distinguishing descriptions, and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features.

[0046] For ease of description, the present invention defines the central corneal dent as a depression formed by the cornea portion removed after laser ablation refractive surgery. The cross section of the depression is an arc shape, and the deepest position of the depression is called the lowest point of the central corneal dent. In addition, the circular boundary line between the depression area and the original cornea is called the corneal cutting edge.

[0047] In the present invention, the initial configuration of the cornea before refractive surgery is simplified to a dome shape with uniform thickness, such as Figure 1 As shown in Figure 1, it can be defined by three geometric parameters: the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, and the corneal radius R0. Under the action of intraocular pressure IOP, the cornea will expand to Figure 2 The configuration shown in FIG. 1 shows that the initial vertex height S0 and thickness T0 are respectively changed to the corneal vertex height S and corneal thickness T when there is intraocular pressure. However, since the corneal edge is connected to the sclera, the corneal radius R0 will not change before and after the intraocular pressure increases. Therefore, the current configuration of the cornea in the presence of intraocular pressure IOP is determined by three geometric parameters S0, T0, R0 and IOP, which means that there is a quantitative relationship between the three geometric parameters and IOP. If this relationship is known, the IOP can be determined by the three geometric parameters measured by the medical instrument without the need for actual measurement by the instrument. To determine the relationship between the three geometric parameters and IOP, it is necessary to analyze the deformation process of the cornea from the initial state to the IOP state. If the parameters S0, T0 and R0 are known, the finite element method can be used to simulate the deformation process of the cornea under a specific IOP. However, due to the scarcity of human corneas, there are few reports on the anatomical parameters of the living human cornea, and it is almost impossible to know the initial configuration of the cornea. Therefore, the present invention can combine finite element simulation and machine learning methods, use the data obtained from finite element simulation to train the machine learning model, and use the machine learning model's ability to fit nonlinear relationships to construct a mapping relationship between IOP and geometric parameters.

[0048] Based on the above analysis of the initial configuration of the cornea before surgery, for patients who undergo laser ablation refractive surgery, the upper surface of the central area of ​​the cornea is cut into a flat shape during the refractive surgery. However, there is no relevant prior art record on how the initial configuration of the cornea after surgery should be set when there is no IOP. Since the center of the cornea after refractive surgery under IOP load is flat, the present invention deduce that the central curvature of the cornea after surgery should be concave downward after unloading the IOP. After repeatedly adjusting the concave shape and parameters and based on finite element analysis, the present invention assumes that the center of the initial configuration of the cornea after surgery is a regular arc-shaped structure that is concave downward, namely, a central concave mark, thereby satisfying the requirement that the central upper surface of the cornea is flat under IOP. Therefore, the present invention simplifies the initial configuration of the cornea after refractive surgery into an arched membrane with uniform thickness and a central arc-shaped downward concave, such as Figure 3As shown. The initial corneal configuration after the refractive surgery can be defined by five geometric parameters: T0, S0, R0, h2 and h3. Compared with the initial corneal configuration before surgery, h2 and h3 are two newly introduced hypothetical parameters that describe the postoperative central corneal dent. h2 is the height difference between the lowest point of the central corneal dent and the edge of the corneal cutting, and h3 is the height difference between the edge of the corneal cutting and the initial corneal vertex before surgery. S0, T0 and R0 are similar to the definitions in the initial corneal configuration before surgery. S0 is the height of the initial corneal vertex when there is no intraocular pressure, T0 is the initial corneal thickness when there is no intraocular pressure, and R0 is the corneal radius. When there is intraocular pressure IOP, the current configuration of the cornea after refractive surgery under IOP changes as follows: Figure 4 As shown, t is the postoperative corneal center thickness, S is the postoperative corneal vertex height, r is the laser cutting radius, and Se is the corneal cutting edge height. Similarly, since the five geometric parameters T0, S0, R0, h2 and h3 of the initial configuration of the postoperative cornea after refractive surgery can determine the postoperative corneal vertex height S under IOP, but since these five geometric parameters cannot be directly measured, the present invention may also need to combine finite element simulation and machine learning, use the data obtained by finite element simulation to train the machine learning model, and use the machine learning model's ability to fit nonlinear relationships to construct a mapping relationship between IOP and these five geometric parameters.

[0049] Based on the theoretical modeling of the initial corneal configuration before surgery without IOP and the initial corneal configuration after surgery without IOP, the present invention provides a method for monitoring intraocular pressure after refractive surgery, which comprises the following steps:

[0050] S1. Aiming at the postoperative goal of laser ablation refractive surgery, the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and preoperative intraocular pressure IOP measured by tonometer are obtained by optical instruments, and used as inputs of the first neural network prediction model to obtain the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central fovea and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex.

[0051] It should be noted that the acquisition of parameters R0, S, t, r, Se and IOP in this step should be understood as reading the data that have been measured, that is, reading from optical instruments or tonometers or from their data storage devices, but the acquisition process does not include the process of operating these instruments to measure corneal parameters.

[0052] It should be noted that among the six inputs of the first neural network prediction model, the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se can be measured by ophthalmic optical instruments after laser ablation refractive surgery, and the remaining input is the intraocular pressure IOP. In theory, the intraocular pressure IOP is also the best choice for postoperative intraocular pressure, but because the applanation tonometer (GAT) and non-contact tonometer (NCT) have large errors in measuring intraocular pressure values ​​after refractive surgery, the postoperative intraocular pressure cannot be accurately obtained. According to relevant ophthalmology theories, the intraocular pressure IOP is determined by the water circulation of the eyeball and will not change in a short period of time after refractive surgery. Therefore, the preoperative intraocular pressure IOP can be measured by a tonometer before refractive surgery, and it can replace the postoperative intraocular pressure as the input of the first neural network prediction model. In addition, the corneal radius R0 generally does not change before and after surgery, so it is possible to measure it before or after surgery.

[0053] It should also be noted that the above-mentioned first neural network prediction model is trained by the MLP neural network through the first training set. However, due to the fact that there are too few actual ophthalmic data samples and it is difficult to meet the problem of neural network training requirements, in the embodiment of the present invention, the first training set is generated based on finite element simulation. The above-mentioned initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex can constitute a hypothetical initial corneal configuration after surgery when there is no intraocular pressure, and the hypothetical initial corneal configuration after surgery is the basis of finite element simulation.

[0054] In an embodiment of the present invention, a method for generating training samples in the first training set is as follows:

[0055] For the five initial configuration parameters (respectively: initial corneal vertex height S0 when there is no intraocular pressure, initial corneal thickness T0 when there is no intraocular pressure, corneal radius R0, height difference h2 between the lowest point of the corneal central dent and the corneal cutting edge, and height difference h3 between the corneal cutting edge and the initial corneal vertex) and the intraocular pressure IOP, different simulation input parameters are obtained by Latin hypercube sampling within the parameter value range of each of the six parameters, and each group of simulation input parameters consists of the sampled values ​​of the six parameters. On the basis of each group of simulation input parameters, by constructing the aforementioned initial corneal configuration after surgery in the finite element, the corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se can be obtained through finite element simulation, and then the sample input and sample labels required for training the first neural network prediction model are constructed according to the finite element simulation results.

[0056] Specifically, in an embodiment of the present invention, when finite element simulation is performed on each set of simulation input parameters, the initial configuration of the cornea after surgery in the absence of intraocular pressure can be constructed through finite elements based on the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the central corneal fovea and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex. Then, finite element simulation is performed by simulating the application of intraocular pressure IOP to obtain the postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se corresponding to this set of simulation input parameters.

[0057] For the first training set, the sample inputs in the training samples are corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and intraocular pressure IOP, and the sample labels are corneal initial vertex height S0 when there is no intraocular pressure, corneal initial thickness T0 when there is no intraocular pressure, height difference h2 between the lowest point of the corneal central dent and the corneal cutting edge, and height difference h3 between the corneal cutting edge and the initial corneal vertex. Each set of simulation input parameters S0, T0, R0, h2, h3, IOP, and S, t, r, Se obtained by finite element simulation can construct a training sample.

[0058] After obtaining the first training set, a method of using the first training set to train an MLP neural network to form a first neural network prediction model belongs to the prior art. The MLP neural network is a multi-layer perceptron (Multi-Layer Perceptron, MLP), which consists of an input layer, a hidden layer and an output layer, and is a type of feedforward neural network. In the training process of the MLP neural network, there are two processes: forward calculation and back propagation. The output of the forward calculation is used to calculate the error function value, and the gradient is calculated through the back propagation algorithm. The error is back propagated from the output layer back to the input layer to update the network parameters. The specific structure and training method of the MLP neural network can be implemented with reference to the prior art, and will not be elaborated on.

[0059] S2. Sample a series of intraocular pressure sampling values ​​IOP from the value range of intraocular pressure IOP i , respectively, different intraocular pressure sampling values ​​IOP i The initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the corneal radius R0 are combined as the input of the second neural network prediction model to obtain different intraocular pressure sampling values ​​IOP. iThe corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se are used to establish a mapping relationship between the postoperative target intraocular pressure IOP and the postoperative corneal vertex height S.

[0060] Similarly, the second neural network prediction model is trained with the second training set using the MLP neural network. The method for generating the training samples in the second training set can refer to the generation of the first training set, specifically: different simulation input parameters are obtained by Latin hypercube sampling within the parameter value ranges of the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the corneal central dent and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the intraocular pressure IOP, and each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se, and then the sample input and sample labels required for training the second neural network prediction model are constructed according to the finite element simulation results. The sample inputs in the training samples of the above-mentioned second training set are the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the central corneal fovea and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, the corneal radius R0, and the intraocular pressure IOP. The sample labels are the postoperative corneal vertex height S, the postoperative corneal center thickness t, the laser cutting radius r, and the corneal cutting edge height Se.

[0061] Based on the generation process of the second training set mentioned above, compared with the first training set, it can be seen that the two have the same approach in the Latin hypercube sampling and finite element simulation process. Therefore, from the perspective of efficiency, the Latin hypercube sampling and finite element simulation processes of the two training sets can be combined. After obtaining different simulation input parameters S0, T0, R0, h2, h3, IOP, and S, t, r, Se obtained through finite element simulation, the first training set and the second training set can be constructed respectively by selecting different sample inputs and sample labels.

[0062] The functions of the first neural network prediction model and the second neural network prediction model obtained by the above training are different. Since the initial configuration of the cornea after surgery when there is no intraocular pressure is only assumed, but the intraocular pressure of the eye must exist in reality, the configuration parameters in the initial configuration of the cornea after surgery cannot be directly measured. Therefore, it is only based on the configuration parameters (R0, S, t, r, Se) that can be directly optically measured when there is intraocular pressure IOP combined with the first neural network prediction model to predict the relevant parameters (S0, T0, h2, h3) of the initial configuration of the cornea after surgery when there is no intraocular pressure. Then, based on the predicted relevant parameters of the initial configuration of the cornea after surgery (S0, T0, h2, h3) and the corneal radius R0 that remains unchanged before and after surgery, combined with the second neural network prediction model, the postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se under any intraocular pressure IOP can be predicted. Therefore, in the above step S2, a series of intraocular pressure sampling values ​​IOP can be sampled from the value range of the intraocular pressure IOP (which can be determined based on empirical values ​​or medical clinical data) i , keep the other 5 inputs (S0, T0, h2, h3, R0) unchanged, and continuously input different intraocular pressure sampling values ​​IOP i , you can read different intraocular pressure sampling values ​​IOP from the model output of the second neural network prediction model i The corresponding postoperative corneal apex height S can be used to establish a mapping relationship between the intraocular pressure IOP and the postoperative corneal apex height S according to the postoperative target.

[0063] After obtaining the mapping relationship between the intraocular pressure IOP and the postoperative corneal vertex height S, the intraocular pressure IOP can be directly determined by measuring the postoperative corneal vertex height S only, without the need for measurement through instruments, thus solving the problem of the difficulty in directly and accurately measuring the intraocular pressure IOP after refractive surgery.

[0064] It should be noted that the mapping relationship between the intraocular pressure IOP and the postoperative corneal vertex height S can be a mapping table or a relationship curve. If a mapping table is used, the sampled intraocular pressure sampling value IOP i The sampling interval should be as small as possible so that any intraocular pressure IOP can be queried by the table lookup method to obtain the corresponding postoperative corneal vertex height S. If a relationship curve is used, the postoperative corneal vertex height S corresponding to the queried intraocular pressure IOP can be directly read on the curve.

[0065] S3. For the postoperative target after laser ablation refractive surgery, obtain the postoperative corneal vertex height S measured by an optical instrument at the time to be monitored, and call the mapping relationship determined in the above step S2 to obtain the postoperative intraocular pressure IOP corresponding to the time to be monitored.

[0066] It should be noted that the acquisition of the postoperative corneal vertex height S in this step should be understood as the reading of the measured data, that is, reading from the optical instrument or from its data storage device, but the acquisition process does not include the process of operating the optical instrument to measure the corneal parameters.

[0067] Therefore, based on the same inventive concept, the present invention also provides a system for monitoring intraocular pressure after refractive surgery corresponding to the method for monitoring intraocular pressure after refractive surgery provided in the above embodiment, comprising:

[0068] The corneal initial configuration parameter prediction module is used to obtain the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and intraocular pressure IOP measured by a tonometer measured by an optical instrument for the postoperative target after laser ablation refractive surgery, and use them as inputs of the first neural network prediction model to obtain the corneal initial vertex height S0 when there is no intraocular pressure, the corneal initial thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the corneal initial vertex;

[0069] The intraocular pressure and postoperative corneal vertex height mapping module is used to sample a series of intraocular pressure sampling values ​​IOP from the value range of intraocular pressure IOP i , respectively, different intraocular pressure sampling values ​​IOP i The initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the corneal radius R0 are combined as the input of the second neural network prediction model to obtain different intraocular pressure sampling values ​​IOP. i The corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se are used to establish a mapping relationship between the postoperative target intraocular pressure IOP and the postoperative corneal vertex height S;

[0070] The real-time prediction module for postoperative intraocular pressure is used to obtain the postoperative corneal vertex height S measured by an optical instrument at the time to be monitored for the postoperative target after laser ablation refractive surgery, and call the mapping relationship to obtain the postoperative intraocular pressure IOP corresponding to the time to be monitored.

[0071] Similarly, based on the same inventive concept, the present invention also provides a computer electronic device corresponding to the method for monitoring intraocular pressure after refractive surgery provided in the above embodiment, which includes a memory and a processor;

[0072] The memory is used to store computer programs;

[0073] The processor is used to implement the method for monitoring intraocular pressure after refractive surgery as described above when executing the computer program;

[0074] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0075] Therefore, based on the same inventive concept, the present invention provides a computer-readable storage medium corresponding to a method for monitoring intraocular pressure after refractive surgery, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it can implement the method for monitoring intraocular pressure after refractive surgery as described above.

[0076] It is understandable that the above storage medium may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. The storage medium may also be a U disk, a mobile hard disk, a magnetic disk or an optical disk, etc., which can store program codes.

[0077] It can be understood that the above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be 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, discrete hardware components.

[0078] It should also be noted that those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the various embodiments provided in this application, the division of steps or modules in the system and method is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules or steps can be combined or integrated together, and a module or step can also be split.

[0079] Similarly, based on the same inventive concept, the present invention also provides a device for monitoring intraocular pressure after refractive surgery corresponding to the method for monitoring intraocular pressure after refractive surgery provided in the above embodiment, which includes an optical instrument, a detection module and a result display module;

[0080] The optical instrument is used to measure in real time the postoperative corneal vertex height S of the postoperative target after laser ablation refractive surgery;

[0081] The detection module is used to receive the postoperative corneal vertex height S measured by the optical instrument, and to call the mapping relationship to obtain the corresponding postoperative intraocular pressure IOP according to the method for monitoring intraocular pressure after refractive surgery as described in the above embodiment;

[0082] The result display module is used to locally display or remotely upload and display the postoperative intraocular pressure IOP detected by the detection module.

[0083] It should be noted that in the device for monitoring intraocular pressure after refractive surgery, the optical instrument can be a general ophthalmic optical instrument (such as optometry KR-800, etc.), as long as the corneal vertex height S can be accurately measured. In addition, in the detection module, according to the method shown in S1~S3 in the aforementioned embodiment, the mapping relationship between the intraocular pressure IOP of the postoperative target and the postoperative corneal vertex height S is first constructed according to S1~S2, and then the postoperative corneal vertex height S measured in real time by the optical instrument is read in, and the postoperative intraocular pressure IOP is converted according to step S3. The result display module can display the postoperative intraocular pressure IOP detected by the detection module. This display can be directly displayed on the local screen, or by uploading the data to the server or cloud platform, and querying and displaying through the mobile terminal, web terminal, etc., which is not limited.

[0084] The method for monitoring intraocular pressure after refractive surgery shown in S1 to S3 in the above embodiments is applied to a specific embodiment to demonstrate its technical effect.

[0085] Example

[0086] In this embodiment, in order to achieve the goal of monitoring intraocular pressure after refractive surgery, an intraocular pressure prediction model before refractive surgery and an intraocular pressure prediction model after refractive surgery are constructed at the same time. The intraocular pressure prediction model before refractive surgery is first used to verify the rationality of the initial corneal configuration proposed in the present invention, and then the intraocular pressure prediction model after refractive surgery is further designed and constructed, and the intraocular pressure after refractive surgery is monitored through this postoperative model. A detailed description is given below.

[0087] 1. Prediction model of intraocular pressure before refractive surgery

[0088] like Figure 1 As shown in FIG. 1 , the initial configuration of the cornea is simplified to a uniform thickness arch shape, which can be defined by three geometric parameters: corneal vertex height S0, corneal vertex height T0, and corneal radius R0. Under the action of intraocular pressure IOP, the cornea will expand to the current configuration as shown in FIG. Figure 2 As shown in FIG. 1 , the initial vertex height S0 and thickness T0 are changed to S and T respectively. Since the corneal edge is connected to the sclera, the corneal radius R0 will not change before and after the intraocular pressure increases. The current configuration of the cornea is determined by three geometric parameters and IOP. Therefore, in this embodiment, finite element simulation and machine learning methods are combined to learn the relationship between IOP and geometric parameters. The process is as follows: Figure 5 As shown, the details are as follows:

[0089] (1.1) In reality, it is difficult to accurately determine the corneal parameter values ​​S0, T0, and R0 of a single individual, but the ranges of these parameters can be estimated based on clinical samples. In addition, the range of intraocular pressure IOP can also be estimated based on samples. Therefore, a statistical analysis is first performed based on clinical sample data to determine the ranges of the parameters S0, T0, R0, and IOP.

[0090] (1.2) A program was written in ABAQUS to establish finite element models with different geometric parameters, such as Figure 6 As shown in the figure, as long as different S0, T0 and R0 are input, the initial configuration model of the cornea with the corresponding parameters can be obtained. Due to individual differences in the cornea, the Latin hypercube sampling method is used to extract 10,000 sets of simulation input parameter samples within the parameter value range determined by statistical analysis, and the finite element method is used to simulate the deformation of the finite element model under different simulation input parameters, thereby obtaining a data set for training the neural network model. Each sample data in the data set consists of the initial configuration parameters of the cornea S0, T0, R0 and IOP, as well as the parameters S and T obtained by FE simulation. This data set can be used to train two different MLP neural network models by combining different inputs and labels.

[0091] (3) Using S, T, R0 and IOP in the data set as the input values ​​of the target model and S0 and T0 as the output values, an MLP neural network model A (hereinafter referred to as model A) is trained. Model A can output the initial configuration parameters S0 and T0 of the cornea by inputting the S, T, R0 and IOP measured by any cornea, so as to obtain the parameters of the initial configuration of the cornea before surgery.

[0092] (4) Using the same data set, an MLP neural network model B (hereinafter referred to as model B) is trained with S0, T0, R0 and IOP as the input values ​​of the target model and S and T as the output values. The model B can obtain S corresponding to different IOP inputs by changing the value of IOP, and then obtain the relationship between IOP and S.

[0093] (5) For any cornea under any IOP state, the parameters S, T, R0 and IOP can be measured on existing medical instruments (optical instruments and tonometers), and then the initial configuration parameters S0 and T0 of the cornea can be predicted by model A. Since the geometric parameters S0, T0 and R0 of the initial state of the cornea without IOP are determined, the relationship between IOP and corneal vertex height S can be established using model B, represented by a curve of IOP and S. Therefore, for any corneal state, after measuring the corneal vertex height S using an optical instrument, the IOP can be easily obtained from the relationship curve.

[0094] Since the accuracy of the initial configuration of the cornea before surgery is the basis for the subsequent prediction of postoperative intraocular pressure, in order to verify the accuracy of the configuration, this embodiment was experimentally verified by recruiting 6 volunteers. Due to the physiological fluctuations of the human body, the aqueous humor circulation will show obvious periodic changes within 24 hours, so there is a significant difference in IOP between the morning and afternoon. In the verification experiment, the IOP of each volunteer was measured using the OCT SS-1000 NCT tonometer in the morning and afternoon, and the corresponding corneal geometric parameters R0, T, and S were measured using a computer refractometer KR-800 as an optical instrument. The parameters R0, T, S, and IOP measured by each volunteer in the morning and afternoon were used as inputs of model A, respectively, to determine the corresponding S0 and T0. It was found that the corneal parameters S0 and T0 of each patient predicted by model A using the data measured in the morning and afternoon, respectively, had an error of less than 5%. Therefore, it can be considered that it is reasonable to assume that the initial configuration of the cornea before surgery is a uniformly thick arched membrane configuration in the present invention. In addition, various organs or trunks of the human body do not show a perfectly uniform geometric configuration. In the present invention, it is assumed that the initial corneal configuration is a structure of uniform thickness, which also reduces geometric complexity and improves calculation efficiency.

[0095] 2. Prediction model of intraocular pressure after refractive surgery

[0096] For patients undergoing refractive surgery, the initial configuration of the cornea during refractive surgery can be simplified to a vaulted membrane of uniform thickness and a central arc-shaped downward incision, e.g. Figure 3 As shown in Figure 1, it can be defined by five geometric parameters: T0, S0, R0, h2, and h3. It is worth noting that h2 and h3 are two assumed parameters describing the central corneal fovea, and S0, T0, and R0 are the same as the configuration definitions before refractive surgery. The current configuration change of the cornea after refractive surgery under IOP is as follows: Figure 4 As shown, t is the corneal center thickness, S is the corneal vertex height, r is the laser cutting radius, and Se is the cutting radius edge height.

[0097] According to the method similar to the above-mentioned method for measuring corneal IOP before refractive surgery, a flow chart of the relationship between corneal vertex height S and IOP after refractive surgery was created as shown in Figure 7 Based on the clinical data, a large number of samples with different initial geometric parameters were determined, and a program was written in ABAQUS to establish finite element models with different geometric parameters. Figure 8 As shown, the deformation of these samples under different IOP conditions was simulated by finite element method. Figure 8 In the finite element model, the shape of the central fovea of ​​the cornea is jointly determined by the parameters h1, h2, and h3, where h1 is the central thickness of the initial configuration of the cornea after surgery, h2 is the height difference between the lowest point of the central fovea of ​​the cornea and the edge of the corneal cutting, and h3 is the height difference between the edge of the corneal cutting and the initial vertex of the cornea before surgery, so it meets h1=T0-h2-h3. Using the data set obtained by numerical simulation and MLP neural network, the prediction model C of the initial geometric parameters and the prediction model D of the current geometric parameters are trained. The specific steps are as follows:

[0098] (2.1) Based on the statistical analysis results of clinical sample data, 10,000 sets of simulation input parameters were obtained by Latin hypercube sampling within the range of the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the corneal radius R0, the height difference h2 between the lowest point of the corneal central dent and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the intraocular pressure IOP. Each set of simulation input parameters was simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se, and then a data set required for training the MLP neural network prediction model was constructed. Each sample data in the data set consists of S0, T0, R0, h2, h3, IOP, S, t, r, and Se. The data set can be used to train two different MLP neural network models by combining different inputs and labels.

[0099] (2.2) Using R0, S, t, r, Se and IOP in the data set as the input values ​​of the target model and S0, T0, h2, h3 as the output values, an MLP neural network model C (hereinafter referred to as model C) is trained. The model C can output the initial configuration parameters S0, T0, h2, h3 of the cornea by inputting the R0, S, t, r, Se and IOP measured by any cornea, so as to obtain the parameters of the initial configuration of the cornea after surgery.

[0100] (2.3) Using the same data set, S0, T0, h2, h3, R0 and IOP are used as the input values ​​of the target model, and S, T, r, Se are used as the output values ​​to train an MLP neural network model D (hereinafter referred to as model D). The model D can obtain S corresponding to different IOP inputs by changing the value of IOP, and then obtain the relationship between IOP and S.

[0101] Similar to the case without surgery, once the parameters S, t, r, Se, R0 and IOP are measured, model C can predict the parameters S0, T0, h2 and h3. Since IOP is determined by the water circulation of the eyeball, it will not change in a short period of time after surgery, so IOP is the preoperative measurement value, and the parameters S, t, r, Se, and R0 can be determined by surgical parameters and instrument measurement. Therefore, the prediction model C can be used to obtain the initial corneal configuration assumed after surgery through the instrument, and then the relationship between S and IOP can be obtained through model D.

[0102] In this embodiment, for any patient, a mapping relationship between the personalized intraocular pressure IOP and the postoperative corneal vertex height S is established, and then the process of realizing IOP prediction is as follows:

[0103] a) For patients undergoing laser ablation refractive surgery, the corneal radius R0, postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, corneal cutting edge height Se and preoperative intraocular pressure IOP measured by tonometer were obtained by optical instrument measurement and used as inputs of the MLP neural network prediction model C to obtain the initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central dent and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex.

[0104] b) Sample a series of intraocular pressure sampling values ​​IOP from the value range of intraocular pressure IOP i , respectively, different intraocular pressure sampling values ​​IOP iThe initial corneal vertex height S0 when there is no intraocular pressure, the initial corneal thickness T0 when there is no intraocular pressure, the height difference h2 between the lowest point of the corneal central concave and the corneal cutting edge, the height difference h3 between the corneal cutting edge and the initial corneal vertex, and the corneal radius R0 are combined as the input of the MLP neural network prediction model D to obtain different intraocular pressure sampling values ​​IOP. i The corresponding postoperative corneal vertex height S, postoperative corneal center thickness t, laser cutting radius r, and corneal cutting edge height Se are used to establish a mapping relationship between the patient's intraocular pressure IOP and the postoperative corneal vertex height S.

[0105] c) For patients who have undergone laser ablation refractive surgery, the postoperative corneal vertex height S measured by an optical instrument at the time to be monitored is obtained, and the mapping relationship is called to obtain the postoperative intraocular pressure IOP corresponding to the time to be monitored.

[0106] In order to better understand the implementation process of the present invention and demonstrate the accuracy of the monitoring results, the following is an example of a volunteer's measurement process. The process of using the four models constructed above to monitor intraocular pressure before and after surgery is as follows:

[0107] Step 1: The IOP measured by the NCT tonometer on the morning before the volunteer's refractive surgery was 15.7 mmHg, and the corresponding parameters R0, S and T measured by the optical instrument were 5.477 mm, 3.033 mm and 0.566 mm, respectively. Using these parameters as the input of prediction model A, the initial corneal configuration parameters T0 and S0 before surgery were calculated by model A to be 0.667 mm and 2.282 mm, respectively. Using T0 and S0 as the input of model B, the curve of corneal S and IOP was obtained, as shown in the figure. Fig. 9 Shown by blue lines.

[0108] Step 2: According to the degree of myopia, the patient underwent laser ablation refractive surgery with a depth of 0.16mm and a radius of 1.60mm. The postoperative parameters of refractive surgery measured by optical instruments were: R0 = 5.477mm, S = 3.056mm, t = 0.406mm, r = 1.600mm and Se = 3.452mm. Since the intraocular aqueous humor circulation did not change in a short period of time after refractive surgery, the IOP was maintained at 15.7mmHg after surgery.

[0109] Step 3: The above parameters are used as inputs to model C to determine the hypothetical initial geometric parameters of the cornea after refractive surgery as T0 = 0.667 mm, S0 = 2.281 mm, h2 = 2.281 mm, and h3 = 0.130 mm. It is worth mentioning that the initial geometric parameters T0 and S0 predicted by model A are 0.667 mm and 2.282 mm, respectively, and the initial geometric parameters predicted by model C are 0.667 mm and 2.281 mm, respectively, reflecting that the predicted values ​​are almost the same.

[0110] Step 4: Using S0, T0, h2, h3, R0 and continuous variable IOP as inputs of model D, the relationship curve between S and IOP after refractive surgery was obtained, as shown in Fig. 9 As shown by the red curve, due to the thinning of the central corneal thickness after refractive surgery, under the same IOP conditions, the postoperative corneal vertex height is higher than the preoperative corneal vertex height.

[0111] Step 5: After obtaining the relationship curve between corneal vertex height S and intraocular pressure IOP, you only need to measure the value of corneal vertex height S through an optical instrument, and then you can directly read the corresponding height of intraocular pressure IOP from the curve.

[0112] The above-described embodiment is only a preferred solution of the present invention, but it is not intended to limit the present invention. A person skilled in the relevant technical field may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A method for monitoring intraocular pressure after refractive surgery, characterized in that: include: S1. For the postoperative goal of laser ablation refractive surgery, the corneal radius, postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, corneal cutting edge height and preoperative intraocular pressure measured by tonometer are obtained by measuring optical instruments, and used as inputs of the first neural network prediction model to obtain the initial corneal vertex height without intraocular pressure, the initial corneal thickness without intraocular pressure, the height difference between the lowest point of the corneal central concave and the corneal cutting edge, and the height difference h3 between the corneal cutting edge and the initial corneal vertex; S2. Sampling a series of intraocular pressure sampling values ​​from the value range of intraocular pressure, respectively combining different intraocular pressure sampling values ​​with the initial corneal vertex height when there is no intraocular pressure, the initial corneal thickness when there is no intraocular pressure, the height difference between the lowest point of the central corneal dent and the corneal cutting edge, the height difference between the corneal cutting edge and the initial corneal vertex, and the corneal radius as inputs of the second neural network prediction model, and obtaining the postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height corresponding to different intraocular pressure sampling values, thereby establishing a mapping relationship between the postoperative target intraocular pressure and the postoperative corneal vertex height; S3, for the postoperative target after laser ablation refractive surgery, obtaining the postoperative corneal vertex height measured by an optical instrument at the time to be monitored, and calling the mapping relationship to obtain the postoperative intraocular pressure corresponding to the time to be monitored; The first neural network prediction model is formed by using an MLP neural network trained with a first training set; The method for generating training samples in the first training set is: different simulation input parameters are obtained by Latin hypercube sampling within the range of respective parameter values ​​of corneal initial vertex height without intraocular pressure, corneal initial thickness without intraocular pressure, corneal radius, height difference between the lowest point of corneal central fovea and corneal cutting edge, height difference between corneal cutting edge and corneal initial vertex, and intraocular pressure; each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height; and then the sample input and sample labels required for training the first neural network prediction model are constructed according to the finite element simulation results.

2. The method for monitoring intraocular pressure after refractive surgery according to claim 1, characterized in that: The second neural network prediction model is formed by using an MLP neural network trained with a second training set.

3. The method for monitoring intraocular pressure after refractive surgery according to claim 2, characterized in that: The method for generating training samples in the second training set is: different simulation input parameters are obtained by Latin hypercube sampling within the range of respective parameter values ​​of the initial corneal vertex height when there is no intraocular pressure, the initial corneal thickness when there is no intraocular pressure, the corneal radius, the height difference between the lowest point of the central corneal fovea and the corneal cutting edge, the height difference between the corneal cutting edge and the initial corneal vertex, and the intraocular pressure; each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height; and then the sample input and sample labels required for training the second neural network prediction model are constructed according to the finite element simulation results.

4. The method for monitoring intraocular pressure after refractive surgery according to claim 3, characterized in that: When finite element simulation is performed on each set of simulation input parameters, the initial configuration of the cornea after surgery in the absence of intraocular pressure is constructed through finite elements based on the initial corneal vertex height when there is no intraocular pressure, the initial corneal thickness when there is no intraocular pressure, the corneal radius, the height difference between the lowest point of the central corneal fovea and the corneal cutting edge, and the height difference between the corneal cutting edge and the initial corneal vertex. Then intraocular pressure is applied for finite element simulation to obtain the postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height corresponding to this set of simulation input parameters.

5. A system for monitoring intraocular pressure after refractive surgery, characterized in that: include: The corneal initial configuration parameter prediction module is used to obtain the corneal radius, postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, corneal cutting edge height and intraocular pressure measured by a tonometer measured by an optical instrument for the postoperative target after laser ablation refractive surgery, and use them as inputs of the first neural network prediction model to obtain the corneal initial vertex height when there is no intraocular pressure, the corneal initial thickness when there is no intraocular pressure, the height difference between the lowest point of the corneal central concave mark and the corneal cutting edge, and the height difference between the corneal cutting edge and the corneal initial vertex; An intraocular pressure and postoperative corneal vertex height mapping module is used to sample a series of intraocular pressure sampling values ​​from the value range of intraocular pressure, and respectively use different intraocular pressure sampling values ​​and the initial corneal vertex height when there is no intraocular pressure, the initial corneal thickness when there is no intraocular pressure, the height difference between the lowest point of the corneal central dent and the corneal cutting edge, the height difference between the corneal cutting edge and the initial corneal vertex, and the corneal radius as inputs of the second neural network prediction model to obtain the postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height corresponding to different intraocular pressure sampling values, thereby establishing a mapping relationship between the intraocular pressure of the postoperative target and the postoperative corneal vertex height; A postoperative intraocular pressure real-time prediction module is used to obtain the postoperative corneal vertex height measured by an optical instrument at the time to be monitored for the postoperative target after laser ablation refractive surgery, and to call the mapping relationship to obtain the postoperative intraocular pressure corresponding to the time to be monitored; The first neural network prediction model is formed by using an MLP neural network trained with a first training set; The method for generating training samples in the first training set is: different simulation input parameters are obtained by Latin hypercube sampling within the range of respective parameter values ​​of corneal initial vertex height without intraocular pressure, corneal initial thickness without intraocular pressure, corneal radius, height difference between the lowest point of corneal central fovea and corneal cutting edge, height difference between corneal cutting edge and corneal initial vertex, and intraocular pressure; each group of simulation input parameters is simulated by finite element simulation to obtain the corresponding postoperative corneal vertex height, postoperative corneal center thickness, laser cutting radius, and corneal cutting edge height; and then the sample input and sample labels required for training the first neural network prediction model are constructed according to the finite element simulation results.

6. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the method for monitoring intraocular pressure after refractive surgery as described in any one of claims 1 to 4 is implemented.

7. A computer electronic device, characterized in that: including memory and processor; The memory is used to store computer programs; The processor is used to implement the method for monitoring intraocular pressure after refractive surgery as described in any one of claims 1 to 4 when executing the computer program.

8. A device for monitoring intraocular pressure after refractive surgery, characterized in that: It includes an optical instrument, a detection module and a result display module; The optical instrument is used to measure in real time the postoperative corneal vertex height of the postoperative target after laser ablation refractive surgery; The detection module is used to receive the postoperative corneal vertex height measured by the optical instrument, and to call the mapping relationship to obtain the corresponding postoperative intraocular pressure according to the method for monitoring intraocular pressure after refractive surgery according to any one of claims 1 to 4; The result display module is used to locally display or remotely upload and display the postoperative intraocular pressure detected by the detection module.

Images