TWI834987B - Method performed by computing device - Google Patents

Abstract

The present invention relates to a method for recommending a vision correction surgery, and the method according to one aspect of the present invention comprises: obtaining an examination data of a subject; predicting whether the vision correction surgery is suitable for the subject from the examination data; when the vision correction surgery is suitable for the subject, predicting whether the vision correction surgery using a laser is available for the subject from the examination data; when the vision correction surgery using the laser is available for the subject, calculating corneal shape factor prediction values of the subject after a standard vision correction surgery and a custom vision correction surgery from the examination data; and when the vision correction surgery using the laser is available for the subject, suggesting a vision correction surgery corresponding to the subject from the examination data.

Description

Method performed by a computing device

The present invention relates to a method and device for recommending vision correction surgery, and more specifically, to a method and device for recommending vision correction surgery for a subject or the like using artificial intelligence.

Vision correction surgeries such as laser-assisted in situ keratomileusis (LASIK) and laser-assisted sub-epithelial keratectomy (LASEK) attract young and old alike. There is a lot of concern from people with poor vision. Attention to vision correction surgery is increasing day by day, and according to statistics, the number of people undergoing vision correction surgery has reached 100,000.

However, it is difficult for subjects who want to undergo vision correction surgery to determine which vision correction surgery is suitable for them. The examinee should basically make a choice between LASIK, LASEK, small incision lenticule extraction (SMILE), lens implantation and other types of surgery. In addition, the person being examined should use surgical equipment such as iLASIK, DaVinci LASIK, Crystal LASIK, Z-LASIK, VISU LASIK, OPTI LASIK, or other procedures such as Contoura Vision, additional ( Choose a type of surgery such as extra) LASIK, wavefront (wavefront) LASIK and other surgical methods. In addition, since the recommended amount of corneal ablation varies between hospitals and doctors, the reality is that it is more difficult for the examinee to choose one type of surgery, and it is inevitable that the quality or side effects of the vision after surgery can only rely on the doctor or consultation. Teacher's words.

[Technical Issue]

One purpose is to assist doctors in making judgments or to provide objective information related to vision correction surgery to doctors, consultants, and examinees.

Another purpose is to recommend vision correction surgery to doctors, consultants and patients.

Yet another purpose is to provide physicians, counselors, and subjects with reasons for recommending vision correction surgery.

The purposes provided are not limited to the above-mentioned purposes, and those with ordinary knowledge in the art can clearly understand the unmentioned purposes based on this description and the drawings. [Technical solutions]

According to one aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the subject, and the examination data includes consultation data and eyes. Characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being derived from the examination of the subject Obtained from the data; when the vision correction surgery is suitable for the subject, by inputting a second set of data into a second prediction model to determine whether the subject can undergo vision correction surgery using laser It is predicted that the second set of data is obtained from the examination data of the subject; when the vision correction surgery using the laser is possible for the subject, by Input the third set of data into the third prediction model to calculate the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted value of corneal shape factor after customized vision correction surgery for determining whether Customized vision correction surgery is required, the third set of data is obtained from the examination data of the subject; and the vision of the subject using the laser is capable of being performed During correction surgery, a vision correction surgery corresponding to the subject is recommended by inputting a fourth set of data from the examination data of the subject to a fourth prediction model. Obtained, wherein the fourth prediction model is trained based on at least one selected from the group consisting of: examination data of a plurality of subjects undergoing vision correction surgery, and the plurality of subjects being treated The corresponding vision correction surgery and the vision of the plurality of treated subjects after the vision correction surgery.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the subject, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being derived from the subject's Obtained from examination data; when the vision correction surgery is suitable for the subject, by inputting the second set of data into the second prediction model, whether the subject can undergo vision correction surgery using laser Making a prediction, the second set of data is obtained from the examination data of the subject; when the vision correction surgery using the laser is possible for the subject, by Predicting whether the subject requires customized vision correction surgery by inputting a third set of data into a third prediction model, the third set of data obtained from the examination data of the subject; and When the vision correction surgery using the laser is feasible for the subject, a vision correction surgery corresponding to the subject is recommended by inputting a fourth set of data into a fourth prediction model , the fourth set of data is obtained from the examination data of the subject, wherein the step of predicting whether the customized vision correction surgery is needed is based on the subject undergoing standard vision correction surgery. The predicted value of corneal shape factor after and the predicted value of corneal shape factor after customized vision correction surgery are used to predict whether the customized vision correction surgery is needed, and wherein the fourth prediction model is based on a group selected from the following At least one of the following: examination data of multiple subjects undergoing vision correction surgeries, vision correction surgeries corresponding to the multiple subjects, and post-vision correction surgeries of the multiple subjects vision.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the examinee, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being derived from the subject's Obtained from examination data; when the vision correction surgery is suitable for the subject, by inputting the second set of data into the second prediction model, whether the subject can undergo vision correction surgery using laser Making a prediction, the second set of data is obtained from the examination data of the subject; and when the vision correction surgery using the laser is feasible for the subject, Recommending vision correction surgery corresponding to the subject by inputting a third set of data obtained from the examination data of the subject into a third prediction model, wherein The step of recommending the vision correction surgery is to recommend the vision correction surgery based on the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted corneal shape factor after customized vision correction surgery. , and wherein the third prediction model is trained based on at least one selected from the group consisting of: examination data of a plurality of subjects undergoing vision correction surgery, and the plurality of subjects The corresponding vision correction surgery and the vision of the plurality of treated subjects after the vision correction surgery.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the examinee, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being derived from the subject's obtained from examination data; and when the vision correction surgery is suitable for the subject, recommending a vision correction surgery corresponding to the subject by inputting a second set of data into a second prediction model, the A second set of data is obtained from the examination data of the subject, wherein the step of recommending the vision correction surgery is based on the corneal shape factor prediction value of the subject after standard vision correction surgery. and customizing corneal shape factor prediction values after vision correction surgery to recommend the vision correction surgery, and the second prediction model is trained based on at least one selected from the group consisting of: undergoing vision correction Examination data of a plurality of subjects undergoing surgery, vision correction surgeries corresponding to the plurality of subjects being treated, and visual acuity of the plurality of subjects being treated after vision correction surgeries.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the examinee, and the examination data includes consultation data and Eye characteristic data measurements; and recommending vision correction surgery corresponding to the subject by inputting a set of data obtained from the examination data of the subject into a predictive model , wherein the step of recommending the vision correction surgery is to recommend the vision based on the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted corneal shape factor after customized vision correction surgery. Correction surgery, and wherein the prediction model is trained based on at least one selected from the group consisting of: examination data of a plurality of subjects undergoing vision correction surgery, and the plurality of subjects The corresponding vision correction surgery and the vision of the plurality of treated subjects after the vision correction surgery.

The solutions to be provided are not limited to the above-mentioned solutions, and those with ordinary knowledge in the art can clearly understand the solutions not mentioned based on this description and the accompanying drawings. [Beneficial effect]

According to one embodiment, a method for assisting a doctor in making a judgment or providing objective information related to vision correction surgery for doctors, consultants, and subjects, and a device for performing the method can be provided.

According to another embodiment, artificial intelligence may be used to recommend vision correction surgery to doctors, consultants, subjects, etc.

According to yet another embodiment, artificial intelligence may be used to provide doctors, consultants, subjects, etc. with reasons for recommending vision correction surgery.

The effects to be provided are not limited to the above-mentioned effects, and those with ordinary knowledge in the art can clearly understand the effects not mentioned based on this description and the drawings.

Since the embodiments set forth in this specification are intended to clearly explain the ideas of the present invention to those skilled in the art, the present invention is not limited to the embodiments set forth in this specification, and the scope of the invention should be construed to include without departing from modifications or variations within the spirit of the present invention.

Although the terms used herein are selected from currently widely used common terms in consideration of the functions in the embodiments of the present invention, the terms may change according to the intentions of those skilled in the art, precedents, or the emergence of new technologies, or the like. However, in certain circumstances, some terms may be chosen at the discretion of the applicant. In this case, its meaning will be elaborated in the corresponding description of the embodiment of the present invention. Therefore, terms used in this specification should be defined based on the meanings of the terms and the entire content of this specification, rather than solely on the basis of the terms themselves.

The drawings attached to this specification are for the purpose of easily explaining the present invention, and the shapes shown in the drawings can be exaggerated and displayed as needed to help understand the present invention, and therefore the present invention is not limited by the drawings.

In this specification, when it is considered that the detailed description of related well-known technology will obscure the gist of the present invention, the detailed description may be omitted. In addition, the terms "first" and "second" used in the description of this specification may refer to different configurations, but may also correspond to the same configuration.

The following describes methods and devices for recommending vision correction surgery to subjects based on examination data. Specifically, a method and device for generating a model that uses artificial intelligence to recommend vision correction surgery and using the generated model to recommend vision correction surgery for a subject are described.

In addition, methods and devices for providing visual vision correction surgery-related information to subjects based on examination data are described. Specifically, for generating models that provide visualization information for vision correction surgery to subjects (for example, using artificial intelligence to provide images of expected visual acuity after vision correction surgery, predicting corneal topography images, or analyzing the causes of predicted results) and using the generated The model describes the method and device for providing visual information for vision correction surgery to the subject.

For the purposes of this specification, vision correction surgery should be understood broadly to include corneal ablation through the use of lasers (eg, laser in situ keratomileusis (LASIK), laser epithelial keratomileusis (LASEK), and minor keratomileusis). Incisional lenticule extraction (SMILE, etc.) is a procedure that corrects the vision of the person being treated, and should be interpreted broadly to include vision-correcting procedures that do not use lasers, such as lens implants.

In addition, in this specification, visual acuity includes visual acuity that can be measured based on the judgment of the subject and visual acuity that can be measured by eye examination or the like. For example, visual acuity can be measured using an eye chart. Alternatively, vision may include lower-order aberrations (eg, myopia, hyperopia, and astigmatism) that are fundamental refractive errors, as well as spherical aberration, coma aberration, etc. ), trefoil aberration and other higher-order aberrations. Additionally, visual acuity may include uncorrected visual acuity and corrected visual acuity.

According to one aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the subject, and the examination data includes consultation data and eyes. Characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being obtained from the examination data of the subject ; When the vision correction surgery is suitable for the subject, predicting whether the subject can undergo vision correction surgery using laser by inputting the second set of data into the second prediction model, the The second set of data is obtained from the examination data of the subject; when vision correction surgery using laser is possible for the subject, by inputting the third set of data into the third prediction The model is used to calculate the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted corneal shape factor value after customized vision correction surgery, so as to determine whether customized vision correction surgery is needed. Three sets of data are obtained from examination data of the subject; and when vision correction surgery using laser is possible for the subject, by inputting a fourth set of data into a fourth prediction The model recommends vision correction surgery corresponding to the subject, the fourth set of data is obtained from the examination data of the subject, wherein the fourth prediction model is based on a group selected from the following At least one of the groups is trained: examination data of a plurality of subjects undergoing vision correction surgeries, vision correction surgeries corresponding to the plurality of subjects being treated, and vision correction surgeries performed on the plurality of subjects being treated vision afterward.

Herein, the step of recommending the vision correction surgery may be to recommend the vision correction surgery based on the predicted vision value of the subject after the vision correction surgery.

Herein, the step of recommending the vision correction surgery may be to recommend the vision correction surgery by calculating the predicted vision value of the subject after the vision correction surgery respectively corresponding to a plurality of vision correction surgeries.

Herein, the step of recommending vision correction surgery may recommend vision correction surgery by calculating a plurality of vision prediction values corresponding to a plurality of different time points.

In this article, the method may further include the step of: based on the calculated corneal shape factor predicted value after standard vision correction surgery and the calculated corneal shape factor predicted value after customized vision correction surgery, calculating the examined corneal shape factor. Predict whether a patient will need customized vision correction surgery.

Herein, the step of recommending the vision correction surgery may be to recommend the vision correction surgery based on the calculated corneal shape factor prediction value after the standard vision correction surgery and the calculated corneal shape factor prediction value after the customized vision correction surgery. .

Herein, when the fourth prediction model is trained by considering the plurality of subjects' preference for vision correction surgery, the fourth set of data may include the subject's preference for vision correction surgery. Preferences.

Herein, at least one selected from the group consisting of the first prediction model, the second prediction model, the third prediction model and the fourth prediction model may include a plurality of sub-models, and is based on The results of the multiple sub-models are calculated.

Herein, at least one selected from the group consisting of the first set of data, the second set of data, the third set of data and the fourth set of data may include the inspected At least part of the examination data of the subject, or a new type of data calculated based on at least part of the examination data of the subject.

In this article, the corneal shape factor prediction value may include a prediction value selected from an index of height decentralization (IHD) prediction value, an index of surface variance (ISV) prediction value, and an index of vertical asymmetry (index of At least one of the groups consisting of vertical asymmetry (IVA) predicted values.

Herein, the examination data may further include genetic information.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the subject, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being obtained from the examination data of the subject When the vision correction surgery is suitable for the subject, predicting whether the subject can undergo vision correction surgery using laser by inputting the second set of data into the second prediction model, so The second set of data is obtained from the examination data of the subject; when vision correction surgery using laser is possible for the subject, by inputting the third set of data into the third Predicting the model to predict whether the subject needs customized vision correction surgery, the third set of data is obtained from the examination data of the subject; and can be used for the subject During laser vision correction surgery, a vision correction surgery corresponding to the subject is recommended by inputting a fourth set of data to a fourth prediction model, the fourth set of data being derived from the examination of the subject Obtained from the data, wherein the step of predicting whether customized vision correction surgery is required is based on the predicted corneal shape factor value of the subject after standard vision correction surgery and the corneal shape after customized vision correction surgery. Factor predictive values are used to predict whether customized vision correction surgery is needed, and wherein the fourth prediction model is trained based on at least one selected from the group consisting of: a plurality of subjects undergoing vision correction surgery. The examination data, the vision correction surgeries corresponding to the plurality of treated subjects, and the vision of the plurality of treated subjects after vision correction surgeries.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the examinee, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being derived from examination data of the subject Obtained from; when the vision correction surgery is suitable for the subject, predicting whether the subject can undergo vision correction surgery using laser by inputting a second set of data into a second prediction model , the second set of data is obtained from the examination data of the subject; and when vision correction surgery using laser is possible for the subject, by inputting the third set of data The third prediction model is used to recommend vision correction surgery corresponding to the subject, and the third set of data is obtained from the examination data of the subject, wherein the step of recommending vision correction surgery is based on The predicted value of corneal shape factor of the subject after standard vision correction surgery and the predicted value of corneal shape factor after customized vision correction surgery are used to recommend vision correction surgery, and wherein the third prediction model is based on one selected from the following At least one of the groups formed is trained: examination data of a plurality of subjects undergoing vision correction surgery, vision correction surgeries corresponding to the plurality of subjects, and procedures performed by the plurality of subjects Vision after vision correction surgery.

According to another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the examinee, and the examination data includes consultation data and Eye characteristic data measurements; predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data into a first prediction model, the first set of data being obtained from the examination data of the subject when the vision correction surgery is suitable for the subject, recommending a vision correction surgery corresponding to the subject by inputting a second set of data into a second prediction model, the second set of data being Obtained from the examination data of the subject, wherein the step of recommending vision correction surgery is based on the predicted corneal shape factor of the subject after standard vision correction surgery and the predicted value of corneal shape factor after customized vision correction surgery. corneal shape factor predictive values to recommend vision correction surgery, and wherein the second prediction model is trained based on at least one selected from the group consisting of: examination of a plurality of subjects undergoing vision correction surgery Data, vision correction surgeries corresponding to the plurality of treated subjects, and the vision of the plurality of treated subjects after vision correction surgeries.

According to yet another aspect, a method for recommending vision correction surgery can be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of the subject, and the examination data includes consultation data and eye characteristic data measurements; and recommending vision correction surgery corresponding to the subject by inputting a set of data obtained from the examination data of the subject into a predictive model, wherein the step of recommending vision correction surgery is to recommend vision correction surgery based on the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted corneal shape factor value after customized vision correction surgery, and wherein The prediction model is trained based on at least one selected from the group consisting of: examination data of a plurality of subjects undergoing vision correction surgeries, vision correction surgeries corresponding to the plurality of subjects and the vision of the plurality of subjects after vision correction surgery.

According to yet another embodiment, a method for providing expected vision images after vision correction surgery may be provided. The method uses artificial intelligence and is executed by a computing device. The method includes the following steps: obtaining examination data of a subject, The examination data includes consultation data and eye characteristic data measurements; by inputting a first set of data obtained from the examination data of the examinee into a first prediction model, the examination result of the examinee in vision correction surgery is calculated. The predicted values of the plurality of eye characteristic data include at least one of the visual acuity prediction value and the corneal shape factor prediction value; and the prediction based on the plurality of eye characteristic data Values are used to generate an expected vision image, wherein the first prediction model can be trained based on at least one of the following: measured values of a plurality of eye characteristic data of a plurality of subjects who have undergone vision correction surgery before the surgery, The surgical parameters of the vision correction surgeries performed by the plurality of subjects, and the measured values of the plurality of eye characteristic data of the plurality of subjects after the surgery.

Herein, the step of generating an expected vision image may include calculating or selecting a filter based on predicted values of a plurality of pieces of eye characteristic data, and applying the filter to the original image to generate the expected vision image.

Herein, the predicted values of the plurality of pieces of eye characteristic data may include a first predicted value of the eye characteristic data and a second predicted value of the eye characteristic data, and the expected vision image may include a third predicted value based on the eye characteristic data. A first expected vision image generated by a predicted value, and a second expected vision image generated based on a second predicted value of the eye characteristic data and different from the first expected vision image.

Herein, the first predicted value of the eye characteristic data and the second predicted value of the eye characteristic data may respectively correspond to the predicted value of the eye characteristic data after standard vision correction surgery and the eye after customized vision correction surgery. Predictive values of the characteristic data, and the first expected vision image and the second expected vision image may respectively correspond to the expected vision image after standard vision correction surgery and the expected vision image after customized vision correction surgery.

As used herein, the expected vision image may include information on at least one selected from the group consisting of: clarity of vision, light bleeding, and light bleeding expected after the subject undergoes vision correction surgery. , contrast sensitivity, night vision, glare, double vision, and afterimage.

Herein, the method of providing the expected vision image may further include the step of: predicting the subject by inputting a second set of data obtained from the subject's examination data into a second prediction model. Predicting corneal topography images after vision correction surgery, wherein the second prediction model can be trained based on at least one selected from the group consisting of: a plurality of subjects who have undergone vision correction surgery before the surgery The corneal topography image, the surgical parameters of the vision correction surgery performed on the plurality of subjects, and the corneal topography image after the surgery on the plurality of subjects.

In this article, the method of providing the expected vision image may further include the step of calculating the dependence of the predicted value of the eye characteristic data on the first set of data.

In this article, the predicted values of the plurality of pieces of eye characteristic data may include a first predicted value of the eye characteristic data and a second predicted value of the eye characteristic data, and the dependence may include at least one piece of the first set of data. The dependence coefficient corresponding to the data, and the dependence coefficient may include a first dependence coefficient corresponding to the first predicted value of the eye characteristic data and a second predicted value corresponding to the eye characteristic data and corresponding to the first dependence Second dependence coefficient with different coefficients.

Herein, the first predicted value of the eye characteristic data and the second predicted value of the eye characteristic data may be compared with the predicted value of the eye characteristic data after standard vision correction surgery and the predicted value of the eye characteristic after customized vision correction surgery, respectively. The predicted values of the data correspond to each other, and the first dependence coefficient and the second dependence coefficient may correspond to the dependence of the predicted value of the eye characteristic data after standard vision correction surgery on the first set of data and the customized vision correction surgery, respectively. The predicted value of the posterior eye characteristic data corresponds to the dependence of the first set of data.

In this article, the method of providing the expected vision image may further include the step of outputting a dependence coefficient greater than a predetermined value or outputting a predetermined number of dependence coefficients among the dependence coefficients.

Herein, the corneal shape factor prediction value may include at least one of an index of height decentration (IHD) prediction value, an index of surface variation (ISV) prediction value, and an index of vertical asymmetry (IVA) prediction value.

Herein, the visual acuity prediction value may include at least one of a lower-order aberrations prediction value and a higher-order aberrations prediction value.

The vision correction surgery assistance system according to the embodiment may include a training device and a prediction device. Here, the training device and the prediction device may be a computing device including at least one control unit. Examples of the computing device may include, but are not limited to, desktop computers, laptop computers, tablet personal computers (PCs), and smart phones.

FIG. 1 is a diagram showing a vision correction surgery assisting system 10 according to an embodiment. Referring to FIG. 1 , the training device 100 may train and/or generate a model in which information related to vision correction surgery is generated based on training data (hereinafter referred to as "vision correction surgery related model"). Here, the training data is the data required for training and/or generating vision correction surgery-related models, such as numbers, text, images, etc., and there is no restriction on the way in which it is expressed. For example, training data may include examination data and surgical parameters of subjects who have undergone vision correction surgery.

The prediction device 300 may calculate a prediction result as information related to the vision correction surgery based on the vision correction surgery related model generated by the training device 100 and the input data. Here, the input data is, for example, numbers, text, images, etc., which are used as the basis for calculating the prediction results, and there is no restriction on the expression method thereof.

Specific examples of training data, input data, prediction results, and vision correction surgery-related models are described below.

In FIG. 1 , the training device 100 and the prediction device 300 are shown as separate devices, but the training device 100 and the prediction device 300 may be the same device. For example, a model related to vision correction surgery may be trained and/or generated within the same device, and the model may be used to calculate predictions. As another option, at least some components of the training device 100 and at least some components of the prediction device 300 may be the same.

In addition, the vision correction surgery assistance system according to the embodiment may include a plurality of training devices and/or a plurality of prediction devices.

FIG. 2 is a diagram for explaining a training device/prediction device according to an embodiment. Referring to FIG. 2 , the training device/prediction device according to the embodiment may include a memory unit 5000 and a control unit 1000 .

The training device/prediction device according to the embodiment may include a control unit 1000 for controlling its operation. The control unit 1000 may include a central processing unit (Central Processing Unit, CPU), a random access memory (Random Access Memory, RAM), a graphics processing unit (Graphic Processing Unit, GPU), one or more microprocessors and can be configured according to One or more electronic components that predetermine logic to process input data.

The control unit 1000 can read system programs and various processing programs stored in the memory unit 5000. For example, the control unit 1000 may perform data processing and the like for performing training and prediction operations of a vision correction surgery-related model to be described below in the RAM, and perform various processes according to the performed program. For example, the control unit 1000 may perform training of a vision correction surgery-related model. As another example, the control unit 1000 may use a vision correction surgery related model to generate prediction results.

The training device/prediction device according to the embodiment may include a memory unit 5000. The memory unit 5000 can store data required for training, training models, and trained vision correction surgery-related models. The memory unit 5000 can store parameters, variables, etc. of models related to vision correction surgery.

The memory unit 5000 may be implemented as a non-volatile semiconductor memory, a hard disk, a flash memory, a RAM, a read-only memory (ROM), or an electrically erasable programmable read-only memory (ROM). erasable programmable read-only memory (EEPROM) or tangible (tangible) non-volatile recording media, etc.

The memory unit 5000 may store various processing programs, parameters for executing program processing, or result data of the processing, etc. For example, the memory unit 5000 may store: a data processing program for performing training and prediction operations of a vision correction surgery-related model described below, a diagnostic process program, parameters for executing each program, and results based on execution of the program. Obtained data (for example, processed data or predicted results).

The training device and/or prediction device according to the embodiment may further include a communication unit 9000. The communication unit 9000 can communicate with external devices. For example, the communication unit 9000 of the training device may communicate with the communication unit 9000 of the prediction device. The communication unit 9000 can perform wired or wireless communication. The communication unit 9000 can perform bi-directional or unidirectional communication.

The training device/prediction device shown in FIG. 2 is only illustrative, and the configuration of the training device and/or prediction device is not limited thereto.

The vision correction surgery assistance system according to embodiments may include a server device and a client device. FIG. 3 is a diagram illustrating a server device 500 and client devices 700a and 700b according to an embodiment.

The server device 500 according to the embodiment may correspond to the above-mentioned training device/prediction device. The server device 500 according to embodiments may train, store and/or execute vision correction surgery related models.

Client devices 700a and 700b according to embodiments may correspond to the training device/prediction device described above. Client devices 700a and 700b according to embodiments may train, store and/or execute vision correction surgery related models.

Client devices 700a and 700b according to embodiments may obtain the trained vision correction surgery-related model from the server device 500. For example, the client devices 700a and 700b may download vision correction surgery related models from the server device 500 via the network.

The server device 500 according to an embodiment may calculate prediction results based on input data obtained from the client devices 700a and 700b. For example, client devices 700a and 700b may receive information about the subject and transmit the information to server device 500, and server device 500 may use a vision correction surgery related model calculation based on the information about the subject. Predict the results. The server device 500 according to an embodiment can obtain input data from a plurality of client devices 700a and 700b.

The server device 500 according to an embodiment may transmit the calculated prediction results to the client devices 700a and 700b. For example, the client devices 700a and 700b may provide prediction results obtained from the server device 500 to doctors, consultants, and subjects. Server device 500 according to embodiments may obtain feedback from client devices 700a and 700b. The server device 500 according to an embodiment may transmit prediction results to the plurality of client devices 700a and 700b.

Client devices 700a and 700b according to embodiments may request prediction results from the server device 500.

Client devices 700a and 700b according to embodiments may transmit the input data to the server device 500. Client devices 700a and 700b according to embodiments may change multiple pieces of input data and transmit the changed data to the server device 500. The client devices 700a and 700b according to the embodiment can provide the prediction results obtained from the server device 500 to doctors, consultants and subjects. The client devices 700a and 700b according to embodiments may change at least part of the prediction results obtained from the server device 500 and provide the changed prediction results to doctors, consultants, and subjects.

In Figure 3, the relationship between one server device 500 and two client devices 700a and 700b is shown, but is not limited thereto, and the relationship can be applied to one or more server devices 500 and a or multiple client devices 700a and 700b.

FIG. 4 is a diagram illustrating the configuration of a server device 500 and a client device 700 according to an embodiment. Referring to FIG. 4 , the server device 500 and the client device 700 may include memory units 5000a and 5000b, control units 1000a and 1000b, and communication units 9000a and 9000b. The server device 500 and the client device 700 can transmit and obtain information through the communication units 9000a and 9000b respectively. For example, the client device 700 can obtain the trained vision correction surgery-related model from the communication unit 9000a of the server device 500 through its communication unit 9000b. As another example, the client device 700 can transmit the input data to the communication unit 9000a of the server device 500 through its communication unit 9000b, and the server device 500 can transmit the prediction results to the client device through its communication unit 9000a. 700 communication unit 9000b.

As described above, the vision correction surgery related model is a model in which various information that can be considered during, before, and after the vision correction surgery is generated.

FIG. 5 is a diagram showing the vision correction surgery related model M according to the embodiment. Referring to Figure 5, the vision correction surgery related model M may include a surgical suitability prediction model M10, a laser surgery availability prediction model M11, a corneal shape factor prediction model M12, a customized vision correction surgery necessity prediction model M13, and a vision correction surgery recommendation model. M14, surgical parameter recommendation model M15, visual acuity prediction model M16, expected visual acuity image generation model M17, corneal topography image prediction model M18 and prediction result calculation cause analysis model M19. Detailed descriptions of each model are given below.

At least part of the vision correction surgery related models M may be trained and/or generated in the same training device. For example, the surgical suitability prediction model M10 and the laser surgery availability prediction model M11 may be trained and/or generated in the same training device. As another option, at least some of the vision correction surgery related models M may be trained and/or generated in different training devices.

At least some of the vision correction surgery related models M may be implemented in the same prediction device. For example, the surgical suitability prediction model M10 and the laser surgery availability prediction model M11 can be implemented in the same prediction device. Furthermore, at least some of the vision correction surgery related models M may be implemented in different prediction devices.

The vision correction surgery-related model according to the embodiment can be trained and/or implemented using artificial intelligence models/algorithms, and the training and/or implementation methods are not limited. For example, vision correction surgery related models can be trained and/or implemented using various machine training models/algorithms and deep training models/algorithms as follows: classification algorithm, regression algorithm, supervised learning learning), unsupervised learning (unsupervised learning), reinforcement learning (reinforcement learning), support vector machine (support vector machine), decision tree (decision tree), random forest (random forest), minimum absolute value compression and selection operator ( least absolute shrinkage and selection operator (LASSO), adaptive boosting (Adaptive Boosting (AdaBoost)), extreme gradient boosting (eXtreme Gradient Boosting (XGBoost)), artificial neural network (artificial neural network), etc.

Vision correction surgery related models according to embodiments may be trained and/or generated through training operations. The vision correction surgery related model trained and/or generated by the training operation can calculate the prediction result by the prediction operation.

6 is a diagram for explaining the training operation S10 and the prediction operation S30 of the vision correction surgery related model according to the embodiment. Referring to FIG. 6 , the training operation S10 may include a training data obtaining operation S110 and a model training operation S150.

The training data obtaining operation S110 may be an operation of obtaining training data from the training device. The training data is data used for training and/or generating a model related to vision correction surgery.

The model training operation S150 may be an operation of training and/or generating a model related to vision correction surgery. In the model training operation S150, the training device may train and/or generate the model based on the training data obtained in the training data obtaining operation. For example, model parameters constituting a vision correction surgery related model may be changed in the model training operation S150. Depending on changes in the parameters, the accuracy of the model can be improved.

A vision correction surgery related model may be trained based on vision correction surgery related information of a subject who has undergone vision correction surgery. For example, the model may be trained based on at least one selected from the group consisting of: examination data of a plurality of subjects who have undergone vision correction surgery before the vision correction surgery, examination data of the plurality of subjects who have undergone vision correction surgery, Surgical parameters of the vision correction surgery performed and examination data after the vision correction surgery of the plurality of subjects.

Referring to FIG. 6 , the prediction operation S30 may include an input data obtaining operation S310 and a model execution operation S350.

The input data obtaining operation S310 may be an operation of obtaining input data from the prediction device, and the input data is data that can be used to calculate prediction results using a vision correction surgery-related model.

The model execution operation S350 may be an operation of calculating a prediction result based on the vision correction surgery-related model trained and/or generated in the training operation S10 and the input data. For example, the prediction device may output prediction results based on input data and model parameters obtained from the training device.

Input data to a vision correction surgery-related model may include all information obtained about the subject. Alternatively, the input data may include at least a portion of the information obtained about the subject. In addition, the input data may be the same or different according to the vision correction surgery related model.

The prediction results of models related to vision correction surgery can vary depending on the training data. Additionally, the accuracy of vision correction surgery-related models can vary based on training data.

Predictions may vary depending on the type of real values contained in the training data. For example, prediction results may differ depending on whether the true value takes into account the subject's subjective intentions. In this context, the subject's subjective intentions may include preferences (eg, whether the subject prefers a particular vision correction surgery) and affordability (eg, how much one can pay for vision correction surgery). The predicted result may be a medically recommended result without considering the subject's subjective intention. On the other hand, when taking into account the subject's subjective intentions, the predicted outcome may be the same or different from the medically recommended outcome.

When the first training data and the second training data are obtained from different hospitals or from different doctors or are training data from different periods, the first model trained and/or generated based on the first training data is the same as the first model trained and/or generated based on the first training data. The second models trained and/or generated by the second training data may be different from each other. Therefore, the first prediction result output from the first model and the second prediction result output from the second model may be different from each other. Additionally, the accuracy of the first model and the accuracy of the second model may differ from each other.

However, even when the pieces of training data are different, the same vision correction surgery-related model can be trained and/or generated. Alternatively, the same prediction results can be calculated even when a vision correction surgery-related model is trained and/or generated based on multiple different pieces of training data.

Training data, input data and predicted results (hereinafter referred to as "input/output data") may include examination data and variables such as surgical parameters. Examination data and surgical parameters can be expressed in various ways, such as numbers, text, images, etc., and there are no restrictions on the way they are expressed. In addition, the image is a two-dimensional (2D) image or a three-dimensional (3D) image, etc., and its dimensions are not limited.

Examination data may include consultation information obtained by questioning without being subjected to equipment or examinations, eye characteristics information obtained by means of equipment or examinations, and genetic information. Information about eyes.

Questionnaire data may include, for example, the following variables: demographic characteristics, such as gender, age, race; living environment, such as the area where the person being tested lives; occupation; income; academic qualifications; education; family size of the person being tested; medical history and family history, For example, hypertension, diabetes, etc.; and the examinee's preference for vision correction surgery.

Eye characteristic data may include all kinds of information related to the eye. For example, eye characteristic data may include variables such as visual acuity, intraocular pressure, and/or retinal examination results.

Eye characteristic data may include information about the physical shape of the eye. For example, eye characteristic data may include, for example, white-to-white (WTW) distance, angle-to-angle (ATA) distance, internal anterior chamber depth (ACD), sulcus Variables such as sulcus-to-sulcus (STS) distance and pupil size. As another example, the eye characteristic data may include information about the shape of the cornea, such as corneal shape factor, corneal topography image, and the like. Here, the corneal shape factor is a numerical value indicating the physical shape of the cornea, and may include, for example, surface variation index (ISV), vertical asymmetry index (IVA), keratoconus index (KI), central keratoconus index (central keratoconus index (CKI), minimum radius of curvature (Rmin), index of height asymmetry (IHA), height decentration index (IHD), central cornea thickness (central cornea thickness) and other variables. In addition, the corneal topography image is an image about the shape of the cornea, and may include a corneal topography map, an anterior corneal curvature image, a posterior corneal curvature image, a corneal thickness map, etc.

Eye characteristics data can be obtained through tomography, topography, optical coherence tomography, etc., such as anterior segment analyzer (Pentacam), CASIA2, AL-Scan, Optical Quality Analysis System (OQAS), etc. coherence tomography (OCT), ultrasound biomicroscopy (UBM) equipment, etc., but are not limited to these. Information obtainable by and derived from the devices described above and similar devices may be included in the eye characteristic data.

Genetic information can be obtained through genetic testing, etc. Genetic information can be used to determine whether vision correction surgery is appropriate for the person being tested. In addition, genetic information can be used to predict side effects after vision correction surgery. For example, genetic information can be used to predict the presence of corneal dystrophy.

Surgical parameters are variables related to the performance of vision correction surgery and may include variables that can be changed during vision correction surgery: such as type of vision correction surgery, e.g., LASIK, LASEK, SMILE, lens implant, standard vision correction surgery, custom Vision correction surgery, etc.; corneal flap thickness (flap thickness); corneal flap diameter (flap diameter); flap side cut angle (flap side cut angle); corneal cutting profile; eye suction ring operation time (suction time); optical area (optic zone); hinge structure, such as hinge position, hinge angle, hinge width, etc.

Standard vision correction surgery may refer to vision correction surgery that corrects low-order aberrations, while custom vision correction surgery may refer to vision correction surgery that corrects both low-order aberrations and high-order aberrations.

The types of surgical parameters that are considered may differ depending on the subject being treated for standard vision correction surgery and the subject being treated for custom vision correction surgery. For example, the types of surgical parameters considered by a subject undergoing custom vision correction surgery may include the types of surgical parameters considered by a subject undergoing standard vision correction surgery.

The number of surgical parameters that vary depending on the subject being treated for standard vision correction surgery may be different from the number of surgical parameters that vary depending on the subject being treated for custom vision correction surgery. Standard vision correction surgery and custom vision correction surgery can be distinguished by the number of surgical parameters that vary depending on the person being treated. The number of surgical parameters used to distinguish standard vision correction surgery from customized vision correction surgery (hereinafter referred to as "reference values") that vary according to the patient may be a predetermined value. For example, when the number of surgical parameters that vary according to the patient is greater than or equal to the reference value, a customized vision correction surgery may be performed, and when the number of surgical parameters is less than the reference value, a standard vision correction surgery may be performed.

Standard vision correction surgery and custom vision correction surgery may vary depending on the hospital, doctor and/or period in which the vision correction surgery is performed. For example, the reference value may vary based on hospital, doctor, and/or time period.

Customized vision correction surgeries can include Contoura Vision and wavefront LASIK.

The quality of vision when performing custom vision correction surgery can be improved compared to when performing standard vision correction surgery. The quality of vision is a term that generally expresses the quality of vision, and can be based not only on the visual acuity, low-order aberrations and high-order aberrations measured by an eye test chart, but also on the clarity of vision, light transmission, and contrast sensitivity. degree, night vision, glare, double vision, afterimages and other discomforts.

Input/output data may include not only measurement values that can be obtained/measured through consultation and examination, but also predictive values that can be calculated by vision correction surgery-related models. For example, the eye characteristic data may include not only measured values of the eye characteristic data obtained/measured by examination before the vision correction surgery, but also include predicted values of the eye characteristic data calculated by the vision correction surgery related model after the vision correction surgery.

The input/output data can be preprocessed and input into vision correction surgery related models. For example, the training device may preprocess the obtained input/output data and then use it for training of vision correction surgery related models. As another option, the prediction device may pre-process the obtained input/output data, and then input the obtained input/output data into the vision correction surgery related model to calculate the prediction result.

Preprocessing should be interpreted broadly to include all changes applied to input/output data and is not limited to the examples disclosed in this specification.

Preprocessing may include selection of at least a portion of the variables contained in the input/output data, such as feature selection. For example, preprocessing may include t-test, Gini index, information gain, relief, DistAUC, signal to noise, minimum redundancy and maximum Relevance (minimum-Redundancy-Maximum-Relevance, MRMR), Fisher score (Fisher score), Laplacian score (Laplacian score), spectral feature selection (spectral feature selection, SPEC), etc.

Preprocessing may include generating new variables from at least a portion of the variables included in the input/output data, such as feature extraction. For example, preprocessing may include principal component analysis, linear discriminant analysis, canonical correlation analysis, singular value decomposition, isometric feature mapping , ISOMAP), locally linear embedding (locally linear embedding), etc. As another example, preprocessing may include generating a spectrum from the values or generating an image from the values.

When the input/output data does not include variables required for vision correction surgery-related models (or training devices and/or prediction devices), preprocessing may include processing methods such as missing value processing. For example, preprocessing may include treating missing values as the mean of the corresponding variable or treating missing values as the mode.

Preprocessing can improve the accuracy of vision correction surgery-related models. For example, the accuracy of a vision correction surgery-related model that includes feature selection may be higher than the accuracy of a model that does not include feature selection. As another example, a vision correction surgery-related model in which images are generated from numerical values and predictions are calculated based on the images may be more accurate than a vision correction surgery-related model in which predictions are calculated based on numerical values.

FIG. 7 is a diagram illustrating preprocessing of input/output data according to an embodiment. Referring to FIG. 7 , the input/output data can be input into the vision correction surgery related model through the preprocessing operation S500. The vision correction surgery related model may calculate prediction results based on the preprocessed input/output data (S700).

A vision correction surgery related model may include multiple sub-models. The plurality of sub-models may calculate predictions based on input data.

Vision correction surgery related models may include models in which sub-models are connected in series and/or parallel, such as an ensemble model, etc.

A vision correction surgery related model may include multiple sub-models connected in series. Here, the sub-models connected in series may mean that the output of at least one sub-model is calculated based on the output of at least another sub-model, for example, the output of the first sub-model is the input of the second sub-model, etc. Alternatively, serially connected sub-models may mean that multiple sub-models should be executed sequentially to obtain prediction results from the input data by the vision correction surgery related model.

FIG. 8 is a diagram illustrating a vision correction surgery related model M including series-connected sub-models M1 and M2 according to an embodiment. Referring to FIG. 8 , the vision correction surgery related model M may include a first sub-model M1 and a second sub-model M2 connected in series. The first sub-model M1 can calculate the output based on the input/output data, and the second sub-model M2 can calculate the prediction result based on the output of the first sub-model M1.

A vision correction surgery related model may include multiple sub-models connected in parallel. Here, sub-models connected in parallel may mean that the output of one sub-model does not affect the output of another sub-model, for example, the output of the first sub-model does not depend on the output of the second sub-model, etc.

Multiple submodels connected in parallel can have the same input. As another option, the inputs of multiple sub-models connected in parallel may be different. For example, the first inspection data input to the first sub-model may be different from the second inspection data input to the second sub-model.

The first examination data may include at least some variables that are different from the variables of the second examination data. For example, the first examination profile may include a preference for vision correction surgery, but the second exam profile may not include a preference for vision correction surgery.

The first exam profile may have the same type of variables as the second exam profile, but the variables may have different values. For example, the first inspection data and the second inspection data may include corneal thickness, but the methods of obtaining the numerical value of corneal thickness are different (for example, using different devices to measure corneal thickness), so the numerical values may be different.

The vision correction surgery related model may include an output sub-model in which the output is calculated based on the output of a plurality of sub-models connected in parallel. For example, when the outputs of the multiple sub-models are the same, an output sub-model may provide the same output. As another example, when the outputs of the plurality of sub-models are different, the output sub-model may output a result considering the outputs of the plurality of sub-models at a predetermined ratio, or may provide a specific output among the plurality of outputs. As yet another example, an output sub-model may output results generated based on outputs of the plurality of sub-models.

FIG. 9 is a diagram illustrating a vision correction surgery related model M including parallel-connected sub-models M1 and M2 according to an embodiment. Referring to FIG. 9 , the vision correction surgery related model M may include a first sub-model M1 and a second sub-model M2 connected in parallel. In addition, the vision correction surgery related model M may include an output sub-model M3 in which the output is calculated based on the outputs of the first sub-model M1 and the second sub-model M2. The first sub-model M1 and the second sub-model M2 can respectively calculate the first output and the second output based on the input/output data, and the output sub-model M3 can calculate the prediction result based on the first output and the second output.

An example of a vision correction surgery-related model including sub-models connected in parallel may include an overall model, but is not limited thereto.

In the following, individual examples of models related to vision correction surgery will be described.

The surgical suitability prediction model predicts whether vision correction surgery is appropriate for the person being examined. Surgical suitability prediction models predict surgical suitability based on input data.

Suitability for vision correction surgery may refer to medical suitability. Therefore, the input data to the surgical suitability prediction model may not include the subject's preference for vision correction surgery. Alternatively, a surgical suitability prediction model may predict suitability for surgery without taking into account the subject's preference for vision correction surgery.

Suitability for surgery may include whether surgery is possible and whether surgery is needed. For example, a surgical suitability prediction model may determine that a subject with good vision who does not require surgery is not suitable for surgery. As another example, a surgical suitability prediction model may determine that a subject whose vision cannot be improved by vision correction surgery is not suitable for surgery.

Suitability for surgery can be output as suitable or unsuitable for surgery. As another option, surgical suitability may be output by quantifying or visualizing surgical suitability.

The laser surgery availability prediction model predicts whether vision correction surgery using lasers will be available for a subject. Here, vision correction surgery using laser may refer to surgery to correct vision by corneal ablation using laser. Laser surgery availability prediction models predict the availability of laser surgery based on input data.

The availability of vision correction surgery using lasers may refer to medical availability. Therefore, the input data to the laser surgery availability prediction model may not include the subject's preference for vision correction surgery. Alternatively, a laser surgery availability prediction model may predict laser surgery availability regardless of the subject's preference for vision correction surgery.

Laser surgery availability can be exported as laser surgery available/unavailable. As another option, laser surgery availability can be output by numericizing or visualizing the surgery availability.

The corneal shape factor prediction model can predict the corneal shape factor of subjects after vision correction surgery. The model can predict one or more corneal shape factors. The corneal shape factor prediction model predicts corneal shape factor based on input data.

Input data to the corneal shape factor prediction model may include surgical parameters. For example, the input data may include the type of surgery as a surgical parameter. The model predicts corneal shape factors corresponding to input surgical parameters. For example, when the input data includes LASIK, LASEK, and SMILE as surgical parameters, the output of the model may include the predicted corneal shape factor after LASIK, the predicted corneal shape factor after LASEK, and the predicted corneal shape factor after SMILE. Predictive value of corneal shape factor. As another example, when the input data includes standard vision correction surgery and customized vision correction surgery as surgical parameters, the output of the model may include predicted values of corneal shape factor after performing standard vision correction surgery and after performing customized vision correction. Predictive value of corneal shape factor after corrected surgery.

Regardless of whether the input data includes surgical parameters, the corneal shape factor prediction model can predict the corneal shape factor corresponding to the predetermined surgical parameters. For example, in the case where the model is trained to predict corneal shape factor after standard vision correction surgery and corneal shape factor after custom vision correction surgery, the model's input data does not include surgical parameters. The model can also output the predicted corneal shape factor after standard vision correction surgery and the predicted corneal shape factor after customized vision correction surgery.

Table 1 shows the output of the cornea shape factor prediction model according to an embodiment. Referring to Table 1, the corneal shape factor prediction model can output the IHD, ISV and IVA values of the subject. In addition, the corneal shape factor prediction model may include the subject's IHD, ISV, and IVA measurements obtained before vision correction surgery, the predicted IHD, ISV, and IVA values after standard vision correction surgery, and the predicted IHD, ISV, and IVA values after custom vision correction surgery. The subsequent predicted IHD, ISV and IVA values. [Table 1] Project current Predictions After Standard Vision Correction Surgery Predictions after Customized Vision Correction Surgery IHD 0.015 0.021 0.018 ISV 20.0 27.8 25.0 IVA 0.19 0.23 0.22

The prediction model for the necessity of customized vision correction surgery can predict whether the examinee needs customized vision correction surgery. The prediction model for the necessity of customized vision correction surgery can predict the necessity of customized vision correction surgery based on the input data.

The prediction model for the need for customized vision correction surgery can predict the need for customized vision correction surgery without taking into account the subject's preference for vision correction surgery. As another option, a customized vision correction surgery necessity prediction model may consider the subject's preference for vision correction surgery to predict the necessity of customized vision correction surgery. The output of the customized vision correction surgery necessity prediction model may vary depending on whether the subject's preference for vision correction surgery is taken into account.

The prediction model for the necessity of customized vision correction surgery can determine whether the subject needs customized vision correction surgery based on eye characteristics data such as corneal shape factor and corneal topography image.

The prediction model for the necessity of customized vision correction surgery can determine the necessity of customized vision correction surgery based on the absolute value of the corneal shape factor. For example, when the corneal shape factor exceeds a certain range, the prediction model for the necessity of customized vision correction surgery can predict that the subject needs customized vision correction surgery.

The prediction model for the necessity of customized vision correction surgery can determine the necessity of customized vision correction surgery based on the relative value of the corneal shape factor. For example, a prediction model for the need for customized vision correction surgery may compare the corneal shape factor after standard vision correction surgery with the corneal shape factor after custom vision correction surgery to determine the need for customized vision correction surgery.

The vision correction surgery recommendation model may recommend vision correction surgery corresponding to the subject. For example, the model may output a vision correction surgery. As another example, the model may output multiple vision correction surgeries. As yet another example, the model may output multiple vision correction surgeries and information regarding their priority.

The vision correction surgery recommendation model can recommend vision correction surgery based on the input data.

The vision correction surgery corresponding to the subject may refer to the vision correction surgery determined by the vision correction surgery recommendation model without considering the subject's preference for the vision correction surgery. Alternatively, the vision correction surgery corresponding to the subject may refer to the vision correction surgery determined by the vision correction surgery recommendation model taking into account the subject's preference for the vision correction surgery. The output of the vision correction surgery recommendation model may vary depending on whether the subject's preference for vision correction surgery is considered.

The output of the vision correction surgery recommendation model may be determined taking into account the necessity for customized vision correction surgery. For example, the output may be determined by considering the necessity of customized vision correction surgery (eg, standard LASIK, customized LASIK, standard LASEK, customized LASEK, standard SMILE, customized SMILE, etc.).

Alternatively, the output of the vision correction surgery recommendation model may be determined without considering the need for customized vision correction surgery. For example, output can be determined without considering the need for custom vision correction surgery (eg, LASIK, LASEK, SMILE, lens implants, etc.).

The vision correction surgery recommendation model can recommend vision correction surgery based on the quality of vision after the vision correction surgery. For example, the model can recommend vision correction surgery based on the predicted value of vision after the vision correction surgery, which corresponds to multiple vision correction surgeries.

Table 2 shows the output of the vision correction surgery recommendation model according to an embodiment. Referring to Table 2, the vision correction surgery recommendation model can output LASIK, LASEK, and SMILE as well as information about their priorities, such as suitability. In Table 2, the value corresponding to SMILE is greater than the value corresponding to LASIK and LASEK, which may mean that the vision correction surgery recommendation model preferentially recommends SMILE. In addition to the method in Table 2, information about the priority can be output in various ways to indicate the priority. [Table 2] Types of vision correction surgery Suitability LASIK 15.45% LASEK 13.30% SMILE 71.25%

The surgical parameter recommendation model may recommend surgical parameters. The model may recommend one or more surgical parameters. The surgical parameter recommendation model can recommend surgical parameters based on input data.

The input data of the surgical parameter recommendation model may include data in the form of images. Here, the image may be an image obtained by using a device for measurement and/or monitoring, such as a corneal topography map. Alternatively, the image may be an image obtained by interpolation, extrapolation, artificial intelligence, etc. based on measured values. For example, the image may be an image generated from a corneal shape factor.

Surgical outcomes following vision correction surgery performed based on surgical parameters suggested by the surgical parameter recommendation model based on imaging may be better than surgical outcomes following vision corrective surgery performed based on surgical parameters suggested based on numerical values. For example, surgical outcomes following vision correction surgery based on surgical parameters suggested by a model based on corneal topography images may be better than those after visual correction surgery based on surgical parameters suggested based on corneal shape factors (e.g., IHD, ISV, IVA, etc.) Surgery results. Here, surgical results may refer to the quality of vision after surgery. Alternatively, surgical results may refer to the shape of the cornea after surgery.

The vision prediction model predicts the subject's vision after vision correction surgery. The vision prediction model can output vision prediction values based on input data. The model can predict one or more visual acuities.

Input data to the vision prediction model may include surgical parameters. For example, the input data may include the type of surgery as a surgery parameter. The model predicts visual acuity corresponding to input surgical parameters. For example, when the input data includes LASIK, LASEK, and SMILE as surgical parameters, the output of the model may include the predicted value of visual acuity after LASIK, the predicted value of visual acuity after LASEK, and the predicted value of visual acuity after SMILE. As another example, when the input data includes standard vision correction surgery and customized vision correction surgery as surgical parameters, the output of the model may include the predicted value of vision after the standard vision correction surgery and the predicted value of vision after the customized vision correction surgery.

The vision prediction model predicts vision corresponding to predetermined surgical parameters regardless of whether the input data includes surgical parameters. For example, in a scenario where a model is trained to predict visual acuity after standard vision correction surgery and visual acuity after custom vision correction surgery, the model can output visual acuity after standard vision correction surgery even when the input data to the model does not include surgical parameters. The predicted value of visual acuity and the predicted value of visual acuity after customized vision correction surgery.

Vision prediction models predict vision corresponding to multiple different points in time. For example, the model may predict vision corresponding to a first time and a second time after vision correction surgery. Examples of multiple different time points may include, but are not limited to, one day, one week, one month, six months, one year, etc. after vision correction surgery.

The vision prediction model can predict the vision recovery rate of the subject based on vision prediction values corresponding to multiple different time points. For example, the model may predict the rate of vision recovery based on a first time and a second time after vision correction surgery.

Anticipated visual acuity image generation model predicts a subject's visual field after vision correction surgery. The model can generate an image obtained by visualizing the subject's vision quality after vision correction surgery (hereinafter referred to as the "expected vision image"). The model outputs an expected vision image, making it easier to explain vision correction surgery to the subject. The model visualizes and outputs the visual field after vision correction surgery, so that the examinee can more clearly understand the expected results after vision correction surgery, and thus can get help when choosing vision correction surgery.

Input data to the expected visual acuity image generation model may include surgical parameters. For example, the input data may include the type of surgery as a surgery parameter. The model predicts expected visual acuity images corresponding to input surgical parameters. For example, when the input data includes LASIK, LASEK, and SMILE as surgical parameters, the output of the model may include an expected visual acuity image after LASIK, an expected visual acuity image after LASEK, and an expected visual acuity image after SMILE. As another example, when the input data includes standard vision correction surgery and customized vision correction surgery as surgical parameters, the output of the model may include expected visual acuity images after standard vision correction surgery and expected visual acuity images after customized vision correction surgery. As yet another example, when the input data includes multiple view zones, the output of the model may include multiple expected visual acuity images corresponding to the multiple view zones.

The expected visual acuity image generation model can generate expected visual acuity images corresponding to predetermined surgical parameters, regardless of whether the input data includes surgical parameters. For example, in a scenario where a model is trained to generate images of expected visual acuity after standard vision correction surgery and images of expected visual acuity after custom vision correction surgery, the model can output standard images even when the input data to the model does not include surgical parameters. Expected visual acuity images after vision correction surgery and expected visual acuity images after customized vision correction surgery.

The expected vision image may include information about at least one selected from the group consisting of: clarity of vision, light transmission, contrast sensitivity, night vision, glare, diplopia, expected after the subject undergoes vision correction surgery. , and afterimages.

Figure 10 shows a view of an expected visual acuity image according to an embodiment. Referring to a of FIG. 10 , expected visual acuity images I1 and I2 can be expressed by visualizing information about visual acuity clarity. Referring to b of FIG. 10 , expected visual acuity images I3, I4, I5, and I6 can be expressed by visualizing information about light blur.

Multiple expected visual acuity images may correspond to different surgical parameters. Referring to a of FIG. 10 , the first expected vision image I1 may correspond to the vision clarity after customized vision correction surgery, and the second expected vision image I2 may correspond to the vision clarity after standard vision correction surgery. Referring to b of FIG. 10 , the third to sixth expected vision images I3 to I6 may be expected vision images corresponding to different viewing zones. For example, the viewing area of the third expected vision image I3 may be larger than the viewing areas of the fourth to sixth expected vision images I4 to I6.

Anticipated visual acuity images according to embodiments may be generated by filtering using a filter. Here, filtering is a concept used in the general image processing field, and may refer to generating a filtered image by convolution of an image and a filter. Examples of filters may include average filter, weighted average filter, low pass filter, Gaussian filter, median filter, bilateral filter, blur filter, high pass filter, unsharp mask , high-boost filter, sharpening filter, etc., but not limited to these.

FIG. 11 is a diagram illustrating an expected vision image generation model M17 using a filter according to an embodiment. Referring to FIG. 11 , the expected visual acuity image generation model M17 may include a first sub-model M171 and a second sub-model M172.

The first sub-model M171 may calculate and/or select filters based on the input data. For example, the input data may include predicted values of the subject's eye characteristic data after vision correction surgery, and the first sub-model M171 may calculate and/or select filters based on the predicted values of the eye characteristic data. As another example, the input data may include measured values of the subject's eye characteristic data and surgical parameters, and the first sub-model M171 may calculate and/or select the filter based on the measured values of the eye characteristic data and the surgical parameters.

The second sub-model M172 may generate an expected vision image based on the filters calculated and/or selected by the first sub-model M171. For example, the second sub-model M172 may generate an expected vision image by applying a filter to the original image. Here, the original image is an image used as a basis for generating the expected visual acuity image, and may be input from outside the expected visual acuity image generation model M17, or may be included in the model M17.

The corneal topography image prediction model can predict the corneal topography image of subjects after vision correction surgery. The model can predict one or more corneal topographic images. The corneal topography image prediction model can generate corneal topography images based on input data.

The input data of the corneal topography image prediction model may include surgical parameters. For example, the input data may include the type of surgery as a surgery parameter. The model predicts corneal topography images corresponding to input surgical parameters. For example, when the input data includes LASIK, LASEK, and SMILE as surgical parameters, the output of the model may include corneal topography images after LASIK, corneal topography images after LASEK, and corneal topography images after SMILE. As another example, when the input data includes standard vision correction surgery and customized vision correction surgery as surgical parameters, the output of the model may include corneal topography images after standard vision correction surgery and corneal topography images after custom vision correction surgery.

The corneal topography image prediction model can generate corneal topography images corresponding to predetermined surgical parameters, regardless of whether the input data includes surgical parameters. For example, in a scenario where a model is trained to generate corneal topography images after standard vision correction surgery and corneal topography images after custom vision correction surgery, the model can output standard corneal topography images even when the model's input data does not include surgical parameters. Corneal topography image after vision correction surgery and corneal topography image after customized vision correction surgery.

The input data of the corneal topography image prediction model may include the subject's corneal topography image measured before the vision correction surgery. 12 is a diagram illustrating a corneal topography image according to an embodiment. Referring to FIG. 12 , the corneal topography image prediction model M18 can predict the corneal topography image CI2 of the subject after the vision correction surgery based on the cornea topography image CI1 of the subject before the vision correction surgery. In FIG. 12 , only the corneal topography image CI1 is shown as input to the model M18 , but other pieces of input data may also be input.

The prediction result calculation cause analysis model can analyze the calculation cause of the prediction result generated by the vision correction surgery related model. The model calculates the dependence of vision correction surgery-related models on input data. Here, dependencies may include the impact of specific variables of the input data on the predicted results.

The prediction result calculation reason analysis model can output the calculation reason of the prediction result. The model may output one or more computational reasons for the predicted results. In the following, for the convenience of explanation, the calculation reason of the prediction result is expressed as a numerical value, such as a dependency coefficient. However, the calculation reason of the prediction result is not limited to this, and it can be represented by numerical values, images, texts, and combinations thereof. There are no restrictions.

The prediction result calculation cause analysis model may output at least some dependence coefficients. For example, the model may output all calculated dependence coefficients.

The prediction result calculation cause analysis model can output the dependence coefficients that fall within a specific range among the calculated dependence coefficients. For example, the model may output a dependence coefficient greater than a predetermined value among the calculated dependence coefficients. Alternatively, the model may output a dependence coefficient whose absolute value is greater than a predetermined value among the calculated dependence coefficients.

The prediction result calculation cause analysis model can output a certain number of dependency coefficients. For example, the model can output a predetermined number of dependency coefficients.

The prediction result calculation cause analysis model may include a surgical suitability prediction analysis model, a laser surgery availability prediction analysis model, a corneal shape factor prediction analysis model, a customized vision correction surgery necessity prediction analysis model, a vision correction surgery recommendation analysis model, and surgical parameters. Suggestion analysis model, visual acuity prediction analysis model, expected visual acuity image generation analysis model and corneal topography image prediction analysis model. For example, a vision prediction analytic model may calculate the impact of variables included in input data to the vision prediction model on predicting vision in the vision prediction model. Alternatively, the vision prediction analysis model may calculate the dependence of the vision prediction value calculated by the vision prediction model on the input data of the vision prediction model.

The prediction result calculation cause analysis model may include at least a part of a model related to vision correction surgery. For example, a vision prediction analysis model may include a vision prediction model.

FIG. 13 is a diagram illustrating the prediction result calculation cause analysis model M19 including the vision correction surgery related model M192 according to the embodiment. Referring to FIG. 13 , the prediction result calculation cause analysis model M19 may include an input data disturbance model M191, a vision correction surgery related model M192, and a prediction result analysis model M193.

The input data perturbation model M191 can output input data based on the input data perturbation. Model M191 can output one or more disturbance input data. Here, the perturbation of the input data may refer to a change in the input data, such as a change in at least part of the variables included in the input data. For example, when the variable is numeric, a perturbation can refer to an increase or decrease in value. Alternatively, when the variable is an image, the perturbation may refer to an increase or decrease in the pixel value of at least some of the pixels of the image.

The vision correction surgery related model M192 can output a first prediction result corresponding to the input data and a second prediction result corresponding to the perturbation input data based on the input data and the perturbation input data. For example, when the vision correction surgery related model M192 included in the prediction result calculation cause analysis model M19 is a vision prediction model, the first prediction result and the second prediction result may be different vision prediction values. In addition, when the disturbance input data is provided with multiple pieces of disturbance input data, the model can output multiple prediction results corresponding to the multiple pieces of disturbance input data.

The prediction result analysis model M193 can output the calculation reason of the prediction result based on the prediction result. For example, model M193 may calculate the calculated cause of the predicted result based on a first predicted result calculated from the unperturbed input data and a second predicted result calculated from the perturbed input data. Specifically, the dependence of the first prediction on the first variable may be calculated based on the difference between the first prediction and a second prediction calculated from the input data in which the first variable is perturbed.

14 is a view showing the calculation reason of the prediction result calculated according to the embodiment, and specifically, is a view showing the reason of the vision prediction. Referring to FIG. 14 , the calculation cause of the prediction result can be expressed as a numerical value and an image. Characters such as Cornea_Back_Rmin V1, Astigmatism V2, Mono V3, Nearsightedness V4, Op_flag V5 and Pupil_Dia V6 shown in Figure 14 may correspond to variables included in the input data. Expressed as corresponding characters, 5.95 N1, -0.5 N2, 0 N3, -1.37 N4, 2 N5, and 3.1 N6 may correspond to the numerical values of the variables included in the input data.

The dependence of prediction results on variables can be visualized and expressed. For example, the dependence can be expressed by the length, color and direction of the arrow. Referring to FIG. 14 , the length of the arrow may correspond to the absolute value of the dependence coefficient. Furthermore, the direction and color of the arrow can correspond to the sign of the dependence coefficient. In Figure 14, it can be explained that the predicted visual acuity value OV is 1.18, Cornea_Back_Rmin V1, astigmatism V2, monocular V3 and myopia V4 have a positive impact on the predicted visual acuity value, while Op_flag V5 and Pupil_Dia V6 have a negative impact on the predicted visual acuity value.

Examples of the prediction result calculation cause analysis model may include model-independent local interpretable explanation (local interpretable model-agnostic explanation, LIME), etc., but are not limited thereto.

Vision correction surgery related models can be combined with each other. The models may be combined in at least one of a series connection or a parallel connection.

Vision correction surgery-related models connected in series may mean that the output of at least one vision correction surgery-related model is calculated based on the output of at least one other vision correction surgery-related model.

Vision correction surgery related models connected in parallel may mean that the output of one vision correction surgery related model does not affect the output of another vision correction surgery related model.

In the following, examples of combinations of vision correction surgery related models will be illustrated.

15 is a diagram illustrating vision correction surgery-related models connected in series according to an embodiment. Referring to FIG. 15 , the first vision correction surgery related model Ma and the second vision correction surgery related model Mb may be combined in series. The first vision correction surgery related model Ma may calculate the first prediction result based on the first input data and the second prediction result output by the second vision correction surgery related model Mb receiving the second input data.

16 to 19 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of the corneal shape factor prediction model, according to an embodiment.

Referring to FIG. 16 , the customized vision correction surgery necessity prediction model M13 may predict the subject's customized vision correction surgery based on the corneal shape factor output by the corneal shape factor prediction model M12 that receives the second input data and the first input data. necessity.

Referring to FIG. 17 , the vision correction surgery recommendation model M14 may recommend a vision correction surgery corresponding to the subject based on the cornea shape factor output by the cornea shape factor prediction model M12 that receives the second input data and the first input data.

18 , the expected vision image generation model M17 may generate an expected vision image of the examinee after vision correction surgery based on the corneal shape factor output by the corneal shape factor prediction model M12 that receives the second input data and the first input data.

Referring to FIG. 19 , the corneal topography image prediction model M18 may predict the corneal topography image of the subject after the vision correction surgery based on the corneal shape factor output by the corneal shape factor prediction model M12 that receives the second input data and the first input data.

20 is a diagram illustrating calculation of the output of a vision correction surgery related model based on the output of a customized vision correction surgery necessity prediction model, according to an embodiment. Referring to FIG. 20 , the vision correction surgery recommendation model M14 may suggest the necessity of the customized vision correction surgery corresponding to the subject based on the customized vision correction surgery necessity prediction model M13 that receives the second input data and the first input data. vision correction surgery.

21 and 22 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of a vision correction surgery recommendation model according to an embodiment.

Referring to FIG. 21 , the corneal shape factor prediction model M12 may predict the corneal shape factor of the subject after the vision correction surgery based on the vision correction surgery output by the vision correction surgery recommendation model M14 that receives the second input data and the first input data.

Referring to FIG. 22 , the vision prediction model M16 may calculate the vision prediction value of the subject after the vision correction surgery based on the vision correction surgery output by the vision correction surgery recommendation model M14 that receives the second input data and the first input data.

The expected vision image generation model may generate an expected vision image of the subject after the vision correction surgery based on the vision correction surgery output by the vision correction surgery recommendation model that receives the second input data and the first input data.

The corneal topography image prediction model can predict the corneal topography image of the subject after the vision correction surgery based on the vision correction surgery output by the vision correction surgery recommendation model that receives the second input data and the first input data.

23 is a diagram illustrating calculation of output of a vision correction surgery-related model based on output of a surgical parameter recommendation model, according to an embodiment.

Referring to FIG. 23 , the vision prediction model M16 may calculate the vision prediction value of the subject after the vision correction surgery based on the surgical parameters output by the surgical parameter recommendation model M15 that receives the second input data and the first input data.

The corneal shape factor prediction model can predict the corneal shape factor of the subject after vision correction surgery based on the surgical parameters output by the surgical parameter recommendation model that receives the second input data and the first input data.

The expected visual acuity image generation model may generate an expected visual acuity image of the subject after the vision correction surgery based on the surgical parameters output by the surgical parameter recommendation model that receives the second input data and the first input data.

The corneal topography image prediction model can predict the corneal topography image of the subject after the vision correction surgery based on the surgical parameters output by the surgical parameter recommendation model that receives the second input data and the first input data.

24 and 25 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of a vision prediction model according to an embodiment.

Referring to FIG. 24 , the vision correction surgery recommendation model M14 may recommend a vision correction surgery corresponding to the subject based on the vision prediction value output by the vision prediction model M16 that receives the second input data and the first input data.

Referring to FIG. 25 , the expected vision image generation model M17 may generate an expected vision image of the subject after the vision correction surgery based on the vision prediction value output by the vision prediction model M16 that receives the second input data and the first input data.

26 and 27 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of the corneal topography image prediction model according to an embodiment.

Referring to FIG. 26 , the customized vision correction surgery necessity prediction model M13 can predict the subject's customized vision correction surgery based on the corneal topography image output by the corneal topography image prediction model M18 that receives the second input data and the first input data. necessity.

Referring to FIG. 27 , the vision correction surgery recommendation model M14 may recommend a vision correction surgery corresponding to the subject based on the cornea topography image output by the cornea topography image prediction model M18 that receives the second input data and the first input data.

The corneal shape factor prediction model can predict the corneal shape factor of the subject after the vision correction surgery based on the corneal topography image output by the corneal topography image prediction model that receives the second input data and the first input data.

The expected visual acuity image generation model may generate an expected visual acuity image of the subject after the vision correction surgery based on the corneal topography image output by the corneal topography image prediction model that receives the second input data and the first input data.

Three or more vision correction surgery related models can be combined in series and/or parallel. 28 is a diagram illustrating a combination of three or more vision correction surgery related models according to an embodiment. Referring to FIG. 28 , the first vision correction surgery-related model Ma can be configured by inputting the second prediction result output by the second vision correction surgery-related model Mb that receives the second input data, and the third vision correction surgery related model that receives the third input data. The third prediction result output by the relevant model Mc and the first input data are used to calculate the first prediction result. Here, the first vision correction surgery related model Ma may be regarded as being connected in series to the second vision correction surgery related model Mb and the third vision correction surgery related model Mc. Furthermore, the second vision correction surgery related model Mb and the third vision correction surgery related model Mc may be regarded as being connected in parallel with each other.

29 and 30 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of the corneal shape factor prediction model and the vision prediction model, according to an embodiment.

29 , the vision correction surgery recommendation model M14 may be based on the cornea shape factor output by the cornea shape factor prediction model M12 that receives the second input data, the vision prediction value output by the vision prediction model M16 that receives the third input data, and the first Data is entered to recommend vision correction surgery for the person being examined.

30 , the expected visual acuity image generation model M17 may be based on the corneal shape factor output by the corneal shape factor prediction model M12 that receives the second input data, the visual acuity prediction value output by the visual acuity prediction model M16 that receives the third input data, and the first Data is entered to generate an image of the subject's expected vision after vision correction surgery.

Vision correction surgery related models can be merged with each other. Multiple vision correction surgery related models may be combined into one model to perform at least some of the functions of the multiple individual models.

31 is a diagram for explaining the incorporation of vision correction surgery related models according to the embodiment. Referring to FIG. 31 , a first vision correction surgery related model and a second vision correction surgery related model may be merged to form one model Mab. The one model Mab can calculate prediction results based on input data. Here, the prediction result may include information corresponding to at least one of a first prediction result corresponding to the output of the first vision correction surgery-related model or a second prediction result corresponding to the output of the second vision correction surgery-related model. For example, the prediction result may include at least one of a first prediction result or a second prediction result. As another option, the prediction result may include information about at least one of the first prediction result or the second prediction result.

In FIG. 31 , the case of merging two models is illustrated, but it is not limited thereto, and three or more models may be merged.

32 to 34 are diagrams illustrating an implementation example of the merger of vision correction surgery-related models according to the embodiment.

Referring to Figure 32, the corneal shape factor prediction model and the customized vision correction surgery necessity prediction model can be combined. The combined model M25 calculates predictions based on the input data. The predicted results may include information corresponding to at least one of corneal shape factor or necessity for customized vision correction surgery. For example, the predicted results may include at least one of corneal shape factor or necessity for customized vision correction surgery.

Referring to Figure 33, a customized vision correction surgery necessity prediction model and a vision correction surgery recommendation model can be combined. The combined model M27 calculates predictions based on input data. The predicted results may include information corresponding to at least one of the need for customized vision correction surgery or vision correction surgery. For example, the predicted outcome may include at least one of customizing the need for vision correction surgery or vision correction surgery.

Referring to FIG. 34 , a laser surgery availability prediction model, a customized vision correction surgery necessity prediction model, and a vision correction surgery recommendation model can be combined. The combined model M38 calculates predictions based on input data. The predicted results may include information corresponding to at least one selected from the group consisting of availability of laser surgery, necessity of customized vision correction surgery, and vision correction surgery. For example, the predicted results may include at least one selected from the group consisting of availability of laser surgery, necessity of customized vision correction surgery, and vision correction surgery.

In the following, examples of recommended approaches to vision correction surgery will be described.

The recommended approach to vision correction surgery may be performed using one or more vision correction surgery related models. When the method is implemented using multiple vision correction surgery-related models, whether to perform at least one vision correction surgery-related model may depend on the predicted results of at least one other vision correction surgery-related model. For example, whether to execute the second vision correction surgery-related model may depend on the predicted results of the first vision correction surgery-related model.

The operations of the recommended method of vision correction surgery to be explained below can be performed by the prediction device.

35 is a diagram illustrating a first example of a recommended method of vision correction surgery according to the embodiment.

Referring to FIG. 35 , the method for recommending vision correction surgery according to the embodiment may include the operation of obtaining the examination data of the subject S1100, the operation of predicting whether the vision correction surgery is suitable for the subject S1200, and the operation of predicting whether the subject can use laser. The operation S1300 of the vision correction surgery, the operation S1400 of calculating the predicted corneal shape factor value of the subject, and the operation S1500 of recommending the vision correction surgery corresponding to the subject.

The operation S1100 of obtaining the examination data of the subject may include obtaining, from the computing device, examination data including measurement values of consultation data and eye characteristic data.

The operation S1200 of predicting whether the vision correction surgery is suitable for the subject may include predicting whether the vision correction surgery is suitable for the subject by inputting a first set of data obtained from the examination data of the subject into a first prediction model. The first prediction model may be a surgical suitability prediction model. The surgical suitability prediction model can predict whether vision correction surgery is suitable for the subject based on the first set of data.

The operation S1300 of predicting whether the subject can undergo vision correction surgery using laser may include predicting whether the subject can perform vision correction surgery by inputting a second set of data obtained from the examination data of the subject into a second prediction model. Laser vision correction surgery. Whether to perform operation S1300 may depend on whether vision correction surgery is suitable for the subject. For example, when vision correction surgery is suitable for the subject, operation S1300 may be performed. The second prediction model may be a laser surgery availability prediction model. The laser surgery availability prediction model can predict the examinee's laser surgery availability based on the second set of data.

The operation S1400 of calculating the cornea shape factor prediction value of the subject may include calculating the cornea of the subject after standard vision correction surgery by inputting a third set of data obtained from the examination data of the subject into a third prediction model. Predictive value of shape factor and predicted value of corneal shape factor in subjects after customized vision correction surgery. Whether to perform operation S1400 may depend on whether the examinee can undergo vision correction surgery using laser. For example, when the subject is capable of vision correction surgery using laser, operation S1400 may be performed. The third prediction model may be a cornea shape factor prediction model. The third prediction model may predict the corneal shape factor based on the third set of data. In this case, the need for customized vision correction surgery may be determined based on the corneal shape factor.

The operation S1500 of recommending a vision correction surgery corresponding to the subject may include recommending a vision correction surgery corresponding to the subject by inputting a fourth set of data obtained from the examination data of the subject into a fourth prediction model. Whether to perform operation S1500 may depend on whether the examinee can undergo vision correction surgery using laser. For example, when the subject is capable of vision correction surgery using laser, operation S1500 may be performed. The fourth prediction model may be trained based on at least one selected from the group consisting of: examination data of a plurality of subjects who have undergone vision correction surgeries, vision correction surgeries corresponding to the plurality of subjects, and The vision of the plurality of subjects after vision correction surgery. A fourth predictive model may suggest a model for vision correction surgery. The fourth prediction model may recommend vision correction surgery corresponding to the subject based on the fourth set of data.

Referring to FIG. 35 , vision correction surgery may be a surgery determined without considering the predicted value of corneal shape factor and/or the need for customized vision correction surgery. For example, vision correction surgeries may include LASIK, LASEK, and SMILE. In addition, since lens implantation is not impossible even when laser vision correction surgery is possible, the vision correction surgery may include lens implantation, which is the same in other examples and implementation examples of this specification. In this case, vision correction surgery may be determined by the physician and/or consultant that takes into account the predictive value of the corneal shape factor and/or the need for customized vision correction surgery. For example, the physician and/or consultant may consider the need for customized vision correction surgery and determine the vision correction surgery based on the corneal shape factor prediction output from the recommended method for vision correction surgery in Figure 35 and the vision correction surgery, such as standard LASIK, Standard LASEK, standard SMILE, customized LASIK, customized LASEK, customized SMILE, etc.

36 is a diagram illustrating a second example of a recommended method of vision correction surgery according to the embodiment.

Referring to FIG. 36 , the method for recommending vision correction surgery according to the embodiment may include the operation of obtaining the examination data of the subject S2100, the operation of predicting whether the vision correction surgery is suitable for the subject S2200, and the operation of predicting whether the subject can use laser. The operation S2300 of the vision correction surgery, the operation S2400 of calculating the predicted corneal shape factor value of the subject, and the operation S2500 of recommending the vision correction surgery corresponding to the subject.

Since the recommended method of vision correction surgery of Fig. 36 is similar to Fig. 35, the differences from Fig. 35 will be mainly explained.

Referring to FIG. 36 , in the operation S2500 of recommending a vision correction surgery corresponding to the subject, the cornea shape factor prediction value after the standard vision correction surgery calculated in the operation of calculating the cornea shape factor prediction value of the subject and Customized prediction of corneal shape factor after vision correction surgery to recommend vision correction surgery. Vision correction surgery may be a surgery determined by considering the need for customized vision correction surgery. For example, vision correction surgery may include standard LASIK, standard LASEK, standard SMILE, custom LASIK, custom LASEK, and custom SMILE. In addition, vision correction surgery may include lens implantation.

37 is a diagram illustrating a third example of a recommended method of vision correction surgery according to the embodiment.

Referring to FIG. 37 , the method for recommending vision correction surgery according to the embodiment may include the operation of obtaining the examination data of the subject S3100, the operation of predicting whether the vision correction surgery is suitable for the subject S3200, and the operation of predicting whether the subject can use laser. The operation S3300 of vision correction surgery, the operation S3400 of predicting whether the subject needs customized vision correction surgery, and the operation S3500 of recommending vision correction surgery corresponding to the subject.

Since the recommended method of vision correction surgery of Fig. 37 is similar to Fig. 35, the differences from Fig. 35 will be mainly explained.

37 , the operation S3400 of predicting whether the subject needs customized vision correction surgery may include predicting whether the subject needs custom vision correction surgery by inputting a third set of data obtained from the subject's examination data into a third prediction model. vision correction surgery. Whether to perform operation S3400 may depend on whether the examinee can undergo vision correction surgery using laser. For example, when the subject is capable of vision correction surgery using laser, operation S3400 may be performed. The third prediction model may be a customized prediction model for the necessity of vision correction surgery. The third prediction model can predict the necessity of customized vision correction surgery based on the third set of data. In operation, the necessity of customized vision correction surgery can be predicted based on the predicted corneal shape factor value of the subject after standard vision correction surgery and the predicted corneal shape factor value of the subject after customized vision correction surgery.

The operation S3400 of predicting whether the subject needs customized vision correction surgery and the operation S3500 of recommending vision correction surgery corresponding to the subject may be connected. For example, the output of operation S3500 of recommending vision correction surgery for the subject may be calculated based on the output of operation S3400 of predicting whether the subject requires customized vision correction surgery. Alternatively, the output of operation S3400 of predicting whether the subject requires customized vision correction surgery may be calculated based on the output of operation S3500 of recommending vision correction surgery corresponding to the subject.

For example, in the operation S3500 of recommending a vision correction surgery corresponding to the subject, the vision correction surgery may be recommended based on the necessity of the customized vision correction surgery determined in the operation S3400 of predicting whether the subject needs the customized vision correction surgery. . For example, vision correction surgery may include standard LASIK, standard LASEK, standard SMILE, custom LASIK, custom LASEK, and custom SMILE. In addition, vision correction surgery may include lens implantation.

As another example, in the operation S3400 of predicting whether the subject needs customized vision correction surgery, a second vision correction surgery may be output based on the first vision correction surgery determined in the operation S3500 of recommending a vision correction surgery corresponding to the subject. Vision correction surgery. Here, the first vision correction surgery may be a surgery determined without considering the necessity of customized vision correction surgery, and the second vision correction surgery may be a surgery determined considering the necessity of customized vision correction surgery.

As yet another example, it may be determined whether to perform the operation S3400 of predicting whether the subject needs a customized vision correction surgery based on the type of the first vision correction surgery determined in the operation S3500 of recommending a vision correction surgery corresponding to the subject. For example, when the first vision correction surgery is a first type of vision correction surgery, a customized vision correction surgery necessity prediction model may not be performed, and when the first vision correction surgery is a second type of vision correction surgery, a custom vision correction surgery necessity prediction model may be performed. Developing a predictive model for the necessity of vision correction surgery.

The first type can be distinguished from the second type based on whether laser ablation of the cornea is used. For example, the first type may be non-laser vision correcting surgery (eg, lens implantation), and the second type may be laser vision correcting surgery (eg, LASIK, LASEK, SMILE).

Type 1 can be distinguished from type 2 based on whether customized surgery is possible. For example, a first type may be a vision correction surgery in which custom surgery is not possible, and a second type may be a vision correction surgery in which custom surgery is possible. Here, there may be predetermined criteria for whether vision correction surgery is a customizable surgery. However, the criteria may vary based on technological advances, hospital, surgical device, physician's circumstances and decisions, etc. For example, when custom SMILE surgery is not possible, the first type may include SMILE and lens implantation, and the second type may include LASIK and LASEK. On the other hand, when custom SMILE procedures are available, the first type may include lens implants, and the second type may include LASIK, LASEK, and SMILE.

38 is a diagram illustrating a fourth example of a recommended method of vision correction surgery according to the embodiment.

Referring to Figure 38, the method for recommending vision correction surgery according to the embodiment may include the operation of obtaining the examination data of the subject S4100, the operation of predicting whether the vision correction surgery is suitable for the subject S4200, and the operation of predicting whether the subject can use laser. The operation S4300 of vision correction surgery, and the operation S4400 of recommending vision correction surgery corresponding to the subject.

Since the operation S4100 of obtaining the examination data of the subject, the operation S4200 of predicting whether the subject is suitable for vision correction surgery, and the operation S4300 of predicting whether the subject can undergo vision correction surgery using laser in FIG. 38 are the same as in FIG. 35 , so they will not be described again.

The operation S4400 of recommending a vision correction surgery corresponding to the subject may include recommending a vision correction surgery corresponding to the subject by inputting a third set of data obtained from the examination data of the subject into a third prediction model.

Whether to perform operation S4400 may depend on whether the examinee can undergo vision correction surgery using laser. For example, when the subject is capable of vision correction surgery using laser, operation S4400 may be performed. In operation S4400, a vision correction surgery may be recommended based on the predicted corneal shape factor value of the subject after the standard vision correction surgery and the predicted corneal shape factor value of the subject after the customized vision correction surgery.

The third prediction model may be a model in which a customized vision correction surgery necessity prediction model and a vision correction surgery recommendation model are merged with each other. The third prediction model may be trained based on at least one selected from the group consisting of: a plurality of examination data of a plurality of subjects who have undergone vision correction surgery, and vision correction corresponding to the plurality of subjects. Surgery and the vision of the plurality of subjects following vision correction surgery. The third predictive model may output information regarding at least one of the need for customized vision correction surgery or vision correction surgery based on the third set of data. For example, the output of the third predictive model may include at least one of customizing the need for vision correction surgery or vision correction surgery.

39 is a diagram illustrating a fifth example of a recommended method of vision correction surgery according to the embodiment.

Referring to FIG. 39 , the recommendation method for vision correction surgery according to the embodiment may include an operation of obtaining examination data of the subject S5100, an operation of predicting whether the vision correction surgery is suitable for the subject S5200, and recommending a vision correction corresponding to the subject. Surgical Operation S5300.

Since the operation S5100 of obtaining the examination data of the subject and the operation S5200 of predicting whether the vision correction surgery is suitable for the subject shown in Fig. 39 are the same as those shown in Fig. 35, they will not be described again.

The operation S5300 of recommending a vision correction surgery corresponding to the subject may include recommending a vision correction surgery corresponding to the subject by inputting a second set of data obtained from the examination data of the subject into a second prediction model.

Whether to perform operation S5300 may depend on whether vision correction surgery is suitable for the subject. For example, when vision correction surgery is suitable for the subject, operation S5300 may be performed. In operation S5300, a vision correction surgery may be recommended based on the predicted corneal shape factor value of the subject after the standard vision correction surgery and the predicted corneal shape factor value of the subject after the customized vision correction surgery.

The second prediction model may be a model in which a laser surgery availability prediction model, a customized vision correction surgery necessity prediction model, and a vision correction surgery recommendation model are combined. The second prediction model may be trained based on at least one selected from the group consisting of: a plurality of examination data of a plurality of subjects who have undergone vision correction surgery, and vision correction corresponding to the plurality of subjects. Surgery and the vision of the plurality of subjects following vision correction surgery. The second predictive model may output information related to at least one selected from the group consisting of laser surgery availability, custom vision correction surgery necessity, and vision correction surgery based on the second set of data. For example, the output of the second predictive model may include at least one selected from the group consisting of laser surgery availability, custom vision correction surgery necessity, and vision correction surgery.

40 is a diagram illustrating a sixth example of a recommended method of vision correction surgery according to the embodiment.

Referring to FIG. 40 , the recommending method for vision correction surgery according to the embodiment may include an operation S6100 of obtaining examination data of the subject, and an operation S6200 of recommending a vision correction surgery corresponding to the subject.

Since the operation S6100 of obtaining the examination data of the subject in Fig. 40 is the same as that in Fig. 35, it will not be described again.

The operation S6200 of recommending a vision correction surgery corresponding to the subject may include recommending a vision correction surgery corresponding to the subject by inputting a set of data obtained from the examination data of the subject into a prediction model. In operation S6200, a vision correction surgery may be recommended based on the predicted corneal shape factor value of the subject after the standard vision correction surgery and the predicted corneal shape factor value of the subject after the customized vision correction surgery.

The prediction model may be a model that combines a surgical suitability prediction model, a laser surgery availability prediction model, a customized vision correction surgery necessity prediction model, and a vision correction surgery recommendation model. The prediction model may be trained based on at least one selected from the group consisting of: a plurality of examination data of a plurality of subjects who have undergone vision correction surgery, and vision correction corresponding to the plurality of subjects. Surgery and the vision of the plurality of subjects following vision correction surgery. The predictive model may output information related to at least one selected from the group consisting of suitability for surgery, availability of laser surgery, necessity for custom vision correction surgery, and vision correction surgery based on the input data. For example, the output of the predictive model may include at least one selected from the group consisting of suitability for surgery, availability of laser surgery, necessity for custom vision correction surgery, and vision correction surgery.

The combination and/or merging of the recommended methods for vision correction surgery and the models related to vision correction surgery are exemplary only, and in addition, the recommended methods for vision correction surgery may be performed, or the models related to vision correction surgery may be combined and/or combined in various ways .

In the following, examples of methods of providing visual information for vision correction surgery will be described.

Methods of providing vision correction surgery visualization information may be implemented using one or more vision correction surgery related models. When the method is implemented using multiple vision correction surgery-related models, whether to implement at least one vision correction surgery-related model may depend on the prediction results of at least one other vision correction surgery-related model. For example, whether to implement the second vision correction surgery-related model may depend on the prediction results of the first vision correction surgery-related model.

Methods of providing visualization information for vision correction surgery may include methods of providing expected visual acuity images, methods of providing corneal topography images, and methods of providing reasons for calculating predicted results.

The method of providing an expected visual acuity image may be implemented using an expected visual acuity image generation model. Methods of providing corneal topography images may be implemented using corneal topography image prediction models. The method of providing the cause of the prediction result calculation may be implemented using a prediction result calculation cause analysis model.

The operations of the method of providing visual information for vision correction surgery described below may be performed by a prediction device.

Figure 41 is a diagram showing a first example of a method for providing visualization information for vision correction surgery according to an embodiment.

Referring to FIG. 41 , a method for providing visualization information for vision correction surgery according to an embodiment may include an operation S7100 of obtaining inspection data of the subject, an operation S7200 of calculating a predicted value of the eye characteristic data of the subject after the vision correction surgery, and generating Operation S7300 of expected visual acuity image. In addition, although not shown, the method of providing vision correction surgery visualization information according to embodiments may further include outputting an expected vision image.

The operation S7100 of obtaining the examination data of the subject may include obtaining, from the computing device, examination data including measurement values of the examination data and the eye characteristic data.

The operation S7200 of calculating the predicted value of the eye characteristic data of the subject after the vision correction surgery may include calculating the predicted value of the eye characteristic data by inputting a first set of data obtained from the examination data of the subject into a first prediction model. , the eye characteristic data includes at least one of the predicted value of the subject's vision or the predicted value of the corneal shape factor after vision correction surgery.

The first prediction model may be trained based on at least one selected from the group consisting of: measured values of a plurality of eye characteristic data of a plurality of subjects who have undergone vision correction surgery before the surgery, a measurement of the plurality of eye characteristic data of the plurality of subjects who have undergone vision correction surgery. The surgical parameters of the vision correction surgery performed by the subject, and the measured values of a plurality of pieces of eye characteristic data after the surgery of the plurality of subjects. The first prediction model may include at least one of a vision prediction model or a corneal shape factor prediction model. As another option, the first prediction model may be a model in which a vision prediction model and a corneal shape factor prediction model are merged. The first prediction model can calculate the prediction value of the subject's eye characteristic data after the vision correction surgery based on the first set of data.

The operation S7300 of generating the expected vision image may include generating the expected vision image based on the predicted value of the eye characteristic data.

42 is a diagram illustrating a second example of a method of providing visual information for vision correction surgery according to an embodiment.

Referring to FIG. 42 , the method of providing vision correction surgery visualization information according to the embodiment may further include the operation of calculating and/or selecting a filter based on predicted value of the eye characteristic data S7600, and the operation of applying the filter to the original image S7700.

The operation S7600 of calculating and/or selecting a filter based on the predicted value of the eye characteristic data may be performed by the first sub-model M171 of FIG. 11 .

The operation of applying the filter to the original image S7700 may include applying the filter to the original image to generate an expected vision image. Operation S7700 may be performed by the second sub-model M172 of FIG. 11 .

43 is a diagram illustrating a third example of a method of providing visual information for vision correction surgery according to an embodiment.

Referring to FIG. 43 , the method for providing vision correction surgery visualization information according to the embodiment may further include an operation S7400 of predicting the corneal topography image of the subject after the vision correction surgery.

The operation S7400 of predicting the subject's corneal topography image after the vision correction surgery may include predicting the subject's cornea after the vision correction surgery by inputting a second set of data obtained from the examination data of the subject into a second prediction model. Terrain images.

The second prediction model may be trained based on at least one selected from the group consisting of: pre-surgery corneal topography images of a plurality of subjects who have undergone vision correction surgery, images performed on the plurality of subjects The surgical parameters of the vision correction surgery, and the corneal topography images of the plurality of treated subjects after the surgery. The second prediction model may be a corneal topography image prediction model. The second prediction model can predict the corneal topography image of the subject after vision correction surgery based on the second set of data.

44 is a diagram illustrating a fourth example of a method of providing visual information for vision correction surgery according to an embodiment.

Referring to FIG. 44 , the method for providing vision correction surgery visualization information according to the embodiment may further include an operation S7500 of calculating the dependence of the predicted value of the eye characteristic data on the first set of data. In addition, although not shown, the method of providing vision correction surgery visualization information according to the embodiment may further include outputting a dependence coefficient.

The output of the dependence coefficients may include outputting a dependence coefficient greater than a predetermined value among the dependence coefficients, or outputting a predetermined number of dependence coefficients.

Methods according to embodiments may be implemented in the form of program instructions executable by various computer units and recorded in computer-readable media. Computer-readable media may include program instructions, data files, data structures, or combinations thereof. The program instructions recorded in the computer-readable medium may be specially designed and prepared for the embodiments, or may be instructions that may be familiar to those skilled in the computer software field. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tapes, such as compact disc read-only memory (CD-ROM) and digital video disc (DVD), etc. Optical media, such as magneto-optical media such as soft magnetic optical disks, and hardware devices specially designed to store and execute program instructions, such as ROM, random access memory (random access memory, RAM) or flash memory. Examples of program instructions may include machine code generated by a compiler and high-level language code that may be executed on a computer using an interpreter. Such hardware devices may be configured as at least one software module to perform the operations of the embodiments, and vice versa.

The configuration and features of the present invention have been described above based on the embodiments, but the present invention is not limited thereto, and it is obvious to those skilled in the art that various changes or modifications can be made within the spirit and scope of the present invention. Accordingly, it is disclosed that such changes or modifications fall within the scope of the appended claims. [Embodiments of the invention]

As mentioned above, in the technical solution of the present invention, relevant matters have been explained.

10: Vision correction surgery assistance system 100:Training device 300: Prediction device 500:Server device 700, 700a, 700b: client device 1000, 1000a, 1000b: control unit 5000, 5000a, 5000b: memory unit 9000, 9000a, 9000b: Communication unit CI1, CI2: Corneal topography image I1: First expected visual acuity image I2: second expected visual acuity image I3: Third expected visual acuity image I4: fourth expected vision image I5: fifth expected visual acuity image I6: sixth expected visual acuity image M: Vision correction surgery related models M1: sub-model/first sub-model M2: Sub-model/Second sub-model M3: Output sub-model M10: Surgical suitability prediction model M11: Laser surgery availability prediction model M12: Corneal shape factor prediction model M13: Customized vision correction surgery necessity prediction model M14: Vision Correction Surgery Recommendation Model M15: Surgical parameter recommendation model M16: Vision prediction model M17: Expected visual acuity image generation model M18: Corneal topography image prediction model M19: Reason analysis model for prediction result calculation M25, M27, M38: merged models M171: First sub-model M172: Second sub-model M191: Input data perturbation model M192: Vision correction surgery related models M193: Prediction result analysis model Ma: First vision correction surgery related model Mb: Second vision correction surgery related model Mab: model Mc: Third vision correction surgery related models OV: predicted visual acuity value S10: Training operations S30: Prediction operation S110: Training data acquisition operation S150: Model training operation S310: Enter data to obtain operation S350: Model execution operation S500: Preprocessing operation S700: Operation S1100, S1200, S1300, S1400, S1500: Operation S2100, S2200, S2300, S2400, S2500: Operation S3100, S3200, S3300, S3400, S3500: Operation S4100, S4200, S4300, S4400: Operation S5100, S5200, S5300: Operation S6100, S6200: Operation S7100, S7200, S7300, S7400, S7500, S7600, S7700: Operation V1:Cornea_Back_Rmin V2: astigmatism V3: Single eye V4: Myopia V5:Op_flag V6:Pupil_Dia

FIG. 1 is a diagram showing a vision correction surgery assisting system according to an embodiment. FIG. 2 is a diagram for explaining a training device/prediction device according to an embodiment. FIG. 3 is a diagram illustrating a server device and a client device according to an embodiment. 4 is a diagram showing the configuration of a server device and a client device according to an embodiment. FIG. 5 is a diagram showing a vision correction surgery related model according to the embodiment. 6 is a diagram for explaining the training operation and the prediction operation of the vision correction surgery related model according to the embodiment. FIG. 7 is a diagram illustrating preprocessing of input/output data according to an embodiment. 8 is a diagram illustrating a vision correction surgery related model including sub-models connected in series, according to an embodiment. 9 is a diagram illustrating a vision correction surgery related model including sub-models connected in parallel, according to an embodiment. Figure 10 shows a view of an expected visual acuity image according to an embodiment. 11 is a diagram illustrating an expected vision image generation model using a filter according to an embodiment. 12 is a diagram illustrating a corneal topography image according to an embodiment. 13 is a diagram illustrating a prediction result calculation cause analysis model including a vision correction surgery related model according to an embodiment. FIG. 14 is a diagram showing calculation reasons for prediction results calculated according to the embodiment. 15 is a diagram illustrating vision correction surgery-related models connected in series according to an embodiment. 16 to 19 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of the corneal shape factor prediction model, according to an embodiment. 20 is a diagram illustrating the calculation of the output of a vision correction surgery-related model based on the output of a customized vision correction surgery necessity prediction model, according to one embodiment. 21 to 22 are diagrams illustrating calculation of an output of a vision correction surgery-related model based on an output of a vision correction surgery recommendation model, according to an embodiment. 23 is a diagram illustrating calculation of output of a vision correction surgery-related model based on output of a surgical parameter recommendation model, according to an embodiment. 24 to 25 are diagrams illustrating calculation of an output of a vision correction surgery-related model based on an output of a vision prediction model according to an embodiment. 26 to 27 are diagrams illustrating calculation of an output of a vision correction surgery-related model based on an output of a corneal topography image prediction model according to an embodiment. 28 is a diagram illustrating a combination of three or more vision correction surgery related models according to an embodiment. 29 to 30 are diagrams illustrating calculation of the output of a vision correction surgery-related model based on the output of the corneal shape factor prediction model and the vision prediction model, according to an embodiment. 31 is a diagram for explaining the incorporation of vision correction surgery related models according to the embodiment. 32 to 34 are diagrams illustrating an implementation example of the merger of vision correction surgery-related models according to the embodiment. 35 is a diagram illustrating a first example of a vision correction surgery recommendation method according to an embodiment. 36 is a diagram illustrating a second example of the vision correction surgery recommendation method according to the embodiment. 37 is a diagram illustrating a third example of the vision correction surgery recommendation method according to the embodiment. 38 is a diagram illustrating a fourth example of the vision correction surgery recommendation method according to the embodiment. 39 is a diagram illustrating a fifth example of the vision correction surgery recommendation method according to the embodiment. 40 is a diagram illustrating a sixth example of the vision correction surgery recommendation method according to the embodiment. 41 is a diagram illustrating a first example of a method of providing visual information for vision correction surgery according to an embodiment. 42 is a diagram illustrating a second example of a method of providing visual information for vision correction surgery according to an embodiment. 43 is a diagram illustrating a third example of a method of providing visual information for vision correction surgery according to an embodiment. 44 is a diagram illustrating a fourth example of a method of providing visual information for vision correction surgery according to an embodiment.

S1100, S1200, S1300, S1400, S1500: Operation

Claims (1)

A method performed by a computing device, the method comprising: Obtain examination data of the subject, wherein the examination data includes consultation data and measured values of eye characteristic data before vision correction surgery; Calculate one or more first prediction values and one or more second prediction values based on the first set of data and the first prediction model, wherein said first set of data is obtained from said examination data of said subject, Wherein the first predictive value represents the predicted corneal shape of the subject after undergoing laser vision correction surgery using predetermined surgical parameters, and the second predictive value represents the predicted corneal shape after the subject undergoes laser vision correction surgery using customized surgical parameters. The subject's predicted corneal shape after laser vision correction surgery; Output the first predicted value and the second predicted value; and recommending a type of vision correction surgery corresponding to the subject based on the second set of data and the second prediction model, wherein said second set of information is obtained from said examination information of said subject, and The recommended vision correction surgery for the subject is determined by combining the results obtained by the first prediction model and the second prediction model.

Images