Optical imaging analysis method for non-contact tonometer

Abstract

The invention provides an optical imaging analysis method for a non-contact tonometer, an optical imaging analysis system for the non-contact tonometer, the non-contact tonometer, a self-adaptive blowing control method of the non-contact tonometer and a computer readable storage medium. The optical imaging analysis method for the non-contact tonometer comprises the following steps: acquiring blowing air pressure data including time information and optical imaging data; aligning and splicing the blowing pressure data and the optical imaging data in a time dimension to determine spliced data; the splicing data are input into the constructed neural network model, a matrix result including the predicted intraocular pressure value and the deformation size confidence coefficient is output, the neural network model comprises a convolution kernel, a splicing kernel and a pooling kernel, and the deformation size confidence coefficient comprises an invalid value, a blowing large value, a blowing small value and a qualified value.

Description

Technical Field

[0001] The present invention relates to the field of intraocular pressure detection devices, and particularly to an optical imaging analysis method for a non-contact tonometer, an optical imaging analysis system for a non-contact tonometer, a non-contact tonometer, an adaptive air blowing control method for a non-contact tonometer, and a computer-readable storage medium. Background Art

[0002] A non-contact tonometer is an intraocular pressure detection device based on an optical sensor, which can accurately measure the intraocular pressure without contact measurement. Since the intraocular pressure is closely related to the occurrence of eye diseases, the non-contact tonometer plays an important role in ophthalmic diagnosis.

[0003] Existing non-contact tonometers usually have problems of poor accuracy and precision, which limit their popularity in clinical applications. When detecting patients, the existing non-contact tonometer has few gear positions for the nozzle air blowing pressure and low control precision, which not only affects the patient's detection experience, but also results in low measurement precision and inaccurate measurement results of the non-contact tonometer.

[0004] In order to overcome the above-mentioned defects existing in the prior art, there is an urgent need in the art for an optical imaging analysis technology for a non-contact tonometer and a non-contact tonometer and its adaptive air blowing control method, which can quickly and accurately analyze the optical imaging in combination with the air blowing pressure data, and control the air blowing system of the non-contact tonometer based on the analysis and prediction results of the optical imaging, so as to improve the comfort of the patient during measurement, and at the same time, improve the measurement precision of the non-contact tonometer and the accuracy of the measurement results. Summary of the Invention

[0005] The following gives a brief overview of one or more aspects to provide a basic understanding of these aspects. This overview is not an exhaustive survey of all contemplated aspects, and is neither intended to identify key or decisive elements of all aspects nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description to follow.

[0006] In order to overcome the above-mentioned defects existing in the prior art, the present invention provides an optical imaging analysis method and system for a non-contact tonometer, a non-contact tonometer and its adaptive air blowing control method, which can quickly and accurately analyze the optical imaging in combination with the air blowing pressure data, and control the air blowing system of the non-contact tonometer based on the analysis and prediction results of the optical imaging, so as to improve the comfort of the patient during measurement, and at the same time, improve the measurement precision of the non-contact tonometer and the accuracy of the measurement results.

[0007] Specifically, the optical imaging analysis method for a non-contact tonometer provided according to the first aspect of the present invention comprises the steps of: acquiring air puff pressure data and optical imaging data including time information; aligning and splicing the air puff pressure data and the optical imaging data in the time dimension to determine the spliced ​​data; and inputting the spliced ​​data into a constructed neural network model, and outputting a matrix result including a predicted intraocular pressure value and a deformation size confidence, wherein the neural network model comprises a convolution kernel, a splicing kernel and a pooling kernel, and the deformation size confidence comprises an invalid value, an air puffing value that is too large, an air puffing value that is too small and a qualified value.

[0008] Preferably, in one embodiment of the present invention, the step of aligning and splicing the air blowing pressure data and the optical imaging data in the time dimension to determine the spliced ​​data includes: adding an optical imaging image to the optical imaging data to form the optical imaging data aligned with the air blowing pressure data set in the time dimension, the optical imaging image being fitted through a neural network operation; and splicing the air blowing pressure data and the optical imaging data aligned in the time dimension to determine the spliced ​​data.

[0009] Preferably, in one embodiment of the present invention, a training step of the neural network model is also included, and the training step includes: acquiring a plurality of the air blowing pressure data and the optical imaging data including time information; aligning and splicing the plurality of the air blowing pressure data and the optical imaging data in the time dimension to determine a data set including a plurality of spliced ​​data, the data set including an intraocular pressure value label and a confidence label of each spliced ​​data; inputting the spliced ​​data into the constructed neural network model for training, and outputting a matrix result including a predicted intraocular pressure value and a deformation size confidence; and based on the predicted intraocular pressure value and the deformation size confidence of the matrix result and the intraocular pressure value label and the confidence label of the spliced ​​data, calculating the loss function to reversely optimize the parameters of the neural network model.

[0010] Preferably, in one embodiment of the present invention, the neural network model includes multiple convolutional layers, and an attention mechanism is added to at least one of the multiple convolutional layers.

[0011] In addition, the optical imaging analysis system for non-contact tonometer provided according to the second aspect of the present invention includes a memory and a processor. The memory stores computer instructions. The processor is connected to the memory and is configured to execute the computer instructions stored in the memory to implement the optical imaging analysis method for non-contact tonometer provided in any one of the above embodiments.

[0012] In addition, the above non-contact tonometer provided according to the third aspect of the present invention includes: a blowing system for blowing air at a target and detecting the air pressure value at the nozzle to form blowing air pressure data; a camera sensor for recording the optical imaging of the target to form optical imaging data; and an optical imaging analysis system for a non-contact tonometer provided according to the second aspect of the present invention.

[0013] Preferably, in an embodiment of the present invention, the above non-contact tonometer includes a slit lamp for illuminating the target.

[0014] Preferably, in an embodiment of the present invention, the camera sensor is a area array camera.

[0015] In addition, the adaptive blowing control method of the above non-contact tonometer provided according to the fourth aspect of the present invention includes the steps of: S1: The blowing system of the non-contact tonometer provided according to the third aspect of the present invention blows air at a target according to a set pressure and detects the air pressure value at the nozzle to form blowing air pressure data; S2: The camera sensor records the optical imaging of the target and forms optical imaging data; S3: Input the blowing air pressure data and the optical imaging data into the optical imaging analysis system for a non-contact tonometer, and output a matrix result including a predicted intraocular pressure value and a confidence level of the deformation size; S4: Determine the analysis result according to the maximum value among the invalid value, the overblown value, the underblown value, and the qualified value included in the confidence level of the deformation size; S5: In response to the analysis result being invalid, repeat steps S1 to S4; S6: In response to the analysis result being overblown or underblown, change the set pressure to the predicted intraocular pressure value, and repeat steps S1 to S4; and S7: In response to the analysis result being qualified, output the predicted intraocular pressure value as the measurement result.

[0016] In addition, computer instructions are stored on the computer-readable storage medium provided according to the fifth aspect of the present invention. When the computer instructions are executed by a processor, the method provided in any one of the above embodiments is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] After reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings, the above features and advantages of the present invention can be better understood. In the drawings, the components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0018] Figure 1 Shows a schematic diagram of an optical imaging analysis system for a non-contact tonometer provided according to some embodiments of the present invention; Figure 2 Shows a schematic diagram of a non-contact tonometer provided according to some embodiments of the present invention; Figure 3The flowchart of an optical imaging analysis method for a non-contact tonometer provided according to some embodiments of the present invention is shown; Figure 4 The schematic diagram of an optical imaging analysis method for a non-contact tonometer provided according to some embodiments of the present invention is shown; Figure 5 The predicted effect diagram of an optical imaging analysis system for a non-contact tonometer provided according to some embodiments of the present invention is shown; and Figure 6 The flowchart of an adaptive blowing control method for a non-contact tonometer provided according to some embodiments of the present invention is shown.

[0019] Reference numerals: 100: Optical imaging analysis system for non-contact tonometer; 110: Memory; 111: Computer-readable storage medium; 120: Processor; 200: Non-contact tonometer; 210: Camera sensor; 220: Lens group; 230: Nozzle; 240: Slit lamp emitter; 300: Optical imaging analysis method; S310~S330: Steps; 401: Blowing air pressure data; 402: Optical imaging data; 410: Neural network operation; 420: Stitching data; 430: Convolutional layer; 431: Convolutional neural network; 432: Activation function; 433: Dropout; 434: Pooling layer; 440: Fully connected layer; 450: Matrix result; 600: Adaptive blowing control method for non-contact tonometer; and S1~S6: Steps. Detailed implementation manners

[0020] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the aspects described below in conjunction with the accompanying drawings and specific embodiments are merely exemplary and should not be construed as imposing any limitation on the protection scope of the present invention.

[0021] In the description of the present invention, it should be noted that, unless otherwise clearly defined and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0022] In addition, the "upper", "lower", "left", "right", "top", "bottom", "horizontal", and "vertical" used in the following description should be understood as the orientations shown in this section and the relevant drawings. Such relative terms are only for convenience of description and do not represent that the devices described need to be manufactured or operated in a specific orientation, and thus should not be construed as a limitation to the present invention.

[0023] It can be understood that although terms such as "first", "second", and "third" can be used herein to describe various components, regions, layers, and / or parts, these components, regions, layers, and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers, and / or parts. Therefore, the first component, region, layer, and / or part discussed below can be referred to as the second component, region, layer, and / or part without departing from some embodiments of the present invention.

[0024] As described above, existing non-contact tonometers generally have problems of poor accuracy and precision, which limits their popularity in clinical applications. When detecting patients, existing non-contact tonometers have few gears for nozzle blowing pressure and low control precision, resulting in low measurement precision and inaccurate measurement results.

[0025] To overcome the above-mentioned defects existing in the prior art, the present invention provides an optical imaging analysis method and system for a non-contact tonometer, as well as a non-contact tonometer and its adaptive blowing control method, which can quickly and accurately analyze optical imaging in combination with blowing pressure data, and control the blowing system of the non-contact tonometer based on the analysis and prediction results of the optical imaging, thereby improving the comfort of patients during measurement, and at the same time improving the measurement precision of the non-contact tonometer and the accuracy of measurement results.

[0026] In some non-limiting embodiments, the above-mentioned optical imaging analysis method for a non-contact tonometer provided by the first aspect of the present invention can be implemented via the above-mentioned optical imaging analysis system for a non-contact tonometer provided by the second aspect of the present invention.

[0027] Please refer to Figure 1 , Figure 1The figure shows a schematic diagram of an optical imaging analysis system for a non-contact tonometer according to some embodiments of the present invention.

[0028] As Figure 1 shown, the optical imaging analysis system 100 for a non-contact tonometer may be configured with a memory 110 and a processor 120. The memory 110 includes but is not limited to the above-mentioned computer-readable storage medium 111 provided in the fifth aspect of the present invention, on which computer instructions are stored. The processor 120 is connected to the memory 110 and is configured to execute the computer instructions stored on the memory 110 to implement the optical imaging analysis method for a non-contact tonometer provided in the first aspect of the present invention.

[0029] The above-mentioned optical imaging analysis system for a non-contact tonometer may be configured in the non-contact tonometer provided in the third aspect of the present invention.

[0030] The non-contact tonometer may include a blowing system for blowing air at a target and detecting the air pressure value at the nozzle to form blowing air pressure data, a camera sensor for recording the optical imaging of the target to form optical imaging data, and an optical imaging analysis system for the non-contact tonometer configured in the non-contact tonometer.

[0031] Please refer to Figure 2 , Figure 2 The figure shows a schematic diagram of a non-contact tonometer according to some embodiments of the present invention.

[0032] In Figure 2 the shown embodiment, the non-contact tonometer 200 may include a support base, on which there are provided air path hardware for forming a blowing system, optical hardware including a slit lamp and a camera sensor 210, and a lens group 220. The air path hardware and the optical hardware may be detachably connected to the support base. The air path hardware may include a nozzle 230, and air flow is ejected through the nozzle of the nozzle 230 towards the patient's eyeball. The air path hardware may further include an air pump, a cylinder, a solenoid valve, a tee, and / or a pressure sensor. One end of the cylinder is communicated with the air pump, and the other end of the cylinder is successively communicated with the pressure sensor and the nozzle through the solenoid valve and the tee.

[0033] The optical hardware of the non-contact tonometer 200 includes a slit lamp for illuminating the target and a camera sensor 210. Here, the camera sensor 210 may be a area array camera. The slit lamp emitter 240 of the slit lamp may be used to generate and adjust the light beam. The slit lamp emitter 240 and the camera sensor 210 may correspond to the nozzle of the nozzle 230 to facilitate optical imaging, and the slit lamp emitter 240 may be located on the side of the camera sensor 210.

[0034] The slit lamp can clearly illuminate the corneal structure of the target, and the camera sensor 210 can record the whole process of the deformation of the eyeball with the airflow. In this way, the optical imaging data obtained by the camera sensor 210 can clearly show the shape and thickness of the cornea and the whole process of corneal deformation, thereby making the obtained optical imaging data have more obvious features and improving the accuracy of the analysis results and subsequent measurement results.

[0035] The working principles of the above-mentioned optical imaging analysis system for non-contact tonometers and non-contact tonometers will be described below in conjunction with some embodiments of optical imaging analysis methods for non-contact tonometers. Those skilled in the art can understand that these embodiments of optical imaging analysis methods for non-contact tonometers are only some non-restrictive implementation manners provided by the present invention, aiming to clearly show the main concept of the present invention and provide some specific solutions convenient for the public to implement, rather than restricting all functions or all working modes of the optical imaging analysis system for non-contact tonometers. Similarly, the optical imaging analysis system for non-contact tonometers is also a non-restrictive implementation manner provided by the present invention, and does not limit the execution subject and execution order of each step in these optical imaging analysis methods for non-contact tonometers.

[0036] Please refer to Figure 3 and Figure 4 , Figure 3 which shows a flowchart of an optical imaging analysis method for a non-contact tonometer provided according to some embodiments of the present invention, Figure 4 and which shows a schematic diagram of an optical imaging analysis method for a non-contact tonometer provided according to some embodiments of the present invention.

[0037] As Figure 3 shown, the optical imaging analysis method 300 for a non-contact tonometer may include step S310: obtaining blowing air pressure data and optical imaging data including time information.

[0038] As Figure 4 shown, the optical imaging analysis system for a non-contact tonometer can obtain blowing air pressure data 401 and optical imaging data 402 including time information respectively through the blowing system and the camera sensor of the non-contact tonometer.

[0039] When the non-contact tonometer detects a target, it will operate the blowing system to blow air at the target. In some embodiments, the blowing system blows air at a relatively comfortable blowing range pressure of 10 - 60 mmHg. After that, the blowing system can detect the air pressure value at the nozzle of the nozzle and collect the blowing air pressure data during the blowing process at a certain frequency.

[0040] In some embodiments, the blowing system of the non-contact tonometer can collect blowing air pressure data at a frequency of 20 KHz, and the collected blowing air pressure data can include 600 data points. Drawing an image of the collected blowing air pressure data based on time points, it can be observed that the air pressure data collected at most times is 0, and the really useful data is the data during the blowing process in the middle, and the amount of data is approximately 360. Therefore, the optical imaging analysis system can intercept 360 air pressure data during the effective time period in the middle to form blowing air pressure data 401, and arrange them to form a 1×360 matrix data.

[0041] While the non-contact tonometer operates the blowing system to blow air at the target, the camera sensor can take pictures of the target to record the optical imaging of the target and form optical imaging data.

[0042] The camera sensor can take pictures to form video data, and this video data can be input into the optical imaging analysis system as optical imaging data 402. In some embodiments, the camera sensor is an area array camera, and each unit time, the area array camera can take pictures of 4-byte data with a size of 768×200 including RGBA information (color information, RGBA represents the color space of Red / red, Green / green, Blue / blue, and Alpha). The number of frames of the video data contains time information. For example, each video data can contain 60 unit times. Thus, the size of each video data is 768×200×4×60. Since the obtained optical imaging data does not require complete RGBA information, the video data can be processed into a 768×200×1×60 matrix.

[0043] After that, as Figure 3 shown, the optical imaging analysis method 300 can include step S320: align and splice the blowing air pressure data and the optical imaging data in the time dimension to determine the spliced data.

[0044] Specifically, the optical imaging analysis system can first determine the unit time of the blowing air pressure data 401 and the optical imaging data 402 to scale the sizes of the blowing air pressure data 401 and / or the optical imaging data 402. For example, when the number of the blowing air pressure data 401 is greater than the number of the optical imaging data 402 during the effective time period of the blowing process, it can be considered that the length of the unit time of the blowing air pressure data 401 is less than the length of the unit time of the optical imaging data 402. In this way, the optical imaging analysis system can scale the optical imaging data 402 according to the data volume and unit time of the blowing air pressure data 401, so that the optical imaging data 402 can be aligned with the blowing air pressure data 401 in the time dimension.

[0045] In Figure 4In the illustrated embodiment, the optical imaging analysis system can fit the optical imaging image through neural network operation 410, and then add the fitted optical imaging image to the optical imaging data 402 to form optical imaging data aligned with the blowing air pressure data 401 in the time dimension. The neural network operation 410 fits the optical imaging image based on the original video data, which can improve the magnification efficiency of the optical imaging data.

[0046] After that, splice the blowing air pressure data and the optical imaging data aligned in the time dimension to determine the spliced data 420.

[0047] For example, in some embodiments, according to the effective time period of the blowing process, the blowing air pressure data within this time period can be determined. The blowing air pressure data can be represented as a 1×360 matrix data. Correspondingly, the optical imaging data within this time period can be represented as a 768×200×1×60 matrix data. During the time period of the blowing process, there are 360 unit-time data for the blowing air pressure data, while there are 60 unit-time data for the optical imaging data. Therefore, the optical imaging analysis system can slice the optical imaging images of 60 unit-time through neural network operation, and form another 5 optical imaging images within each unit-time. Thus, the optical imaging data within this time period can also include 360 unit-time data, forming a 360×1 matrix. Splice the optical imaging data represented as a 360×1 matrix with the blowing air pressure data to synthesize a 360×2 matrix data as the spliced data. By splicing the blowing air pressure data and the optical imaging data aligned in the time dimension, it can make the influence ratio of the blowing air pressure data and the optical imaging data on the final result 1:1 in the subsequent operation process.

[0048] Please continue to refer to Figure 3 that the optical imaging analysis method 300 can include step S330: input the spliced data into the constructed neural network model, and output a matrix result including the predicted intraocular pressure value and the confidence level of the deformation size.

[0049] In recent years, with the wide application of deep learning in the medical field, neural network technology has also attracted more and more attention in the research of non-contact tonometer. The neural network model of the optical imaging analysis method 300 provided by the present invention is constructed based on deep learning. The neural network model can include convolutional kernels, splicing kernels, and max pooling kernels.

[0050] The constructed neural network model can also include multiple convolutional layers. In some embodiments, the multiple convolutional layers can include 26 layers. In other embodiments, through preprocessing of the deployment language and input data, the convolutional layer can be 17 layers.

[0051] In Figure 4In the illustrated embodiment, the neural network model may include a plurality of convolutional layers 430 and fully connected layers 440. Each convolutional layer 430 may include a convolutional neural network 431, an activation function 432, Dropout 433, and a pooling layer 434. Here, the activation function 432 may be a Relu function (Rectified Linear Unit). The neural network model extracts features through a plurality of convolutional layers 430 and then converts them into an output format through the fully connected layer 440, and outputs a matrix result 450 including the predicted intraocular pressure value and the confidence in the deformation size.

[0052] The neural network model may add an attention mechanism to at least one of the plurality of convolutional layers. The attention mechanism may be a mechanism that focuses on local information. For example, a certain image region in an image, so as to locate the information of interest and suppress useless information. The results of the attention mechanism are usually presented in the form of a probability map or a probability feature vector. By adding an attention mechanism to the key convolutional layer, the constructed neural network model can be more adaptable to the field of image and video.

[0053] The optical imaging analysis system inputs the processed stitched data into the constructed neural network model and outputs a 1×5 matrix result including the predicted intraocular pressure value and the confidence in the deformation size.

[0054] The predicted intraocular pressure value is a continuous non-linear regression model. For example, it can vary within the range of 0 to 70 mmHg. The confidence in the deformation size can be a 4-bit discrete model, including invalid value, over-inflation value, under-inflation value, and qualified value, corresponding to the four cases of invalid measurement result, over-inflation, under-inflation, and qualified inflation.

[0055] In some embodiments, the optical imaging analysis system may determine the confidence in the deformation size according to the index number of the maximum value. For example, the output confidence in the deformation size may be a 1×4 matrix, and the 4-column data respectively correspond to the index numbers of invalid measurement result, over-inflation, under-inflation, and qualified inflation. When the output confidence in the deformation size is [0.1, 0.95, 0.3, 0.22], the maximum value is the second 0.95, corresponding to the index number of over-inflation. Thus, the optical imaging analysis system may consider that the confidence in the deformation size at this measurement is the case of over-inflation.

[0056] Those skilled in the art can understand that the output confidence in the deformation size can also be adjusted according to the actual required accuracy requirements. For example, a threshold can be set to divide the optical imaging analysis results of intraocular pressure deformation into six cases: invalid, over-inflated, moderately inflated, under-inflated, severely under-inflated, and qualified, and output as a 6-bit discrete model, so as to further improve the measurement accuracy.

[0057] In some embodiments, the confidence level of the output deformation size may be a 1×4 matrix, and the four columns of data respectively correspond to the index numbers of invalid, over-inflation, under-inflation, and qualified. The size of the fourth column corresponding to the qualified condition can be used as a score to facilitate the optical imaging analysis system to evaluate the qualified condition. In some preferred embodiments, the optical imaging analysis system can also set a threshold for the size of the fourth column corresponding to the qualified condition. For example, when the size of the fourth column corresponding to the qualified condition is above 0.9, it can be considered a reliable over-inflation qualified condition.

[0058] In addition, the neural network model also includes an offline training process. The offline training process can first obtain a plurality of blowing air pressure data and optical imaging data including time information. Then, align and splice the plurality of blowing air pressure data and optical imaging data in the time dimension to determine a data set including a plurality of spliced data. The data set includes the intraocular pressure value label and the confidence level label of each spliced data.

[0059] The optical imaging analysis system can label the spliced data formed by the blowing air pressure data and the optical imaging data within the time period of the same blowing process with an intraocular pressure value label and a confidence level label. In some embodiments, the confidence level label can be a 1×4 matrix data. For example, [0, 1, 0, 0] is used to represent over-inflation.

[0060] After that, the spliced data is input into the constructed neural network model for training, and a matrix result including the predicted intraocular pressure value and the confidence level of the deformation size is output.

[0061] Then, based on the predicted intraocular pressure value and the confidence level of the deformation size in the matrix result and the intraocular pressure value label and the confidence level label of the spliced data, a loss function is calculated to reversely optimize the parameters of the neural network model.

[0062] During the training process, the optimizer of the optical imaging analysis system can use the Adam optimization algorithm, and the loss function can use MSE (Mean-Square Error). In some embodiments, the overall parameters of the trained neural network model are 603,233, which are very suitable for edge deployment.

[0063] Please refer to Figure 5 , Figure 5 which shows the prediction effect diagram of the optical imaging analysis system for the non-contact tonometer according to some embodiments of the present invention.

[0064] As Figure 5As shown, the horizontal axis is the number of samples, and the vertical axis on the left is intraocular pressure (Pressure), in mmhg. The true value of the target's intraocular pressure is the blue true value data point, and the predicted intraocular pressure value obtained by the optical imaging analysis system is the orange predicted value data point. The vertical axis on the right is the relative error (Relative Error), in mmhg. Based on the error values ​​of each sample represented by the green line segments on the horizontal axis, the error between the predicted intraocular pressure value of each sample and the true value of the intraocular pressure can be determined. Figure 5 It can be seen that the error between the predicted intraocular pressure value and the actual value of the optical imaging analysis system used for non-contact tonometer is mostly within 3 mmHg.

[0065] Based on the optical imaging analysis method for a non-contact tonometer provided in the first aspect of the present invention, the non-contact tonometer can implement the adaptive air blowing control method for a non-contact tonometer provided in the fourth aspect of the present invention.

[0066] Please refer to Figure 6 , Figure 6 A flow chart of an adaptive air puff control method for a non-contact tonometer provided according to some embodiments of the present invention is shown.

[0067] like Figure 6 As shown, the adaptive air blowing control method 600 of the non-contact tonometer may include step S1: the air blowing system of the non-contact tonometer blows air toward the target according to the set pressure and detects the air pressure value at the nozzle to form air blowing pressure data.

[0068] Afterwards, the non-contact tonometer may perform step S2: the camera sensor records the optical imaging of the target and forms optical imaging data.

[0069] Then, the non-contact tonometer can execute step S3: input the air puff pressure data and the optical imaging data into the optical imaging analysis system for the non-contact tonometer, and output a matrix result including the predicted intraocular pressure value and the confidence level of the deformation size.

[0070] The blowing pressure data and optical imaging data within the time period of the blowing process are input into the optical imaging analysis system for the non-contact tonometer configured in the non-contact tonometer, and through the optical imaging analysis method of the non-contact tonometer provided by the first aspect of the present invention, a matrix result including the predicted intraocular pressure value and the confidence level of the deformation size is output.

[0071] Afterwards, the non-contact tonometer may execute step S4: determining the analysis result according to the maximum value among the invalid value, the excessive air-blowing value, the insufficient air-blowing value and the qualified value included in the confidence level of the deformation magnitude.

[0072] like Figure 6As shown in steps S5, S6 and S7, when the analysis result is invalid, steps S1 - S4 are repeated to re - measure the target. When the analysis result shows that the blowing air pressure is too large or too small, the set pressure is changed to the predicted intraocular pressure value, and steps S1 - S4 are repeated until the analysis result is qualified, and the predicted intraocular pressure value is output as the measurement result.

[0073] Specifically, in some embodiments, during measurement, there may be a situation of occlusion. When the optical imaging analysis system faces such optical imaging data, it can output the deformation size confidence level representing the invalidity of the measurement result and re - measure the target.

[0074] In some embodiments, if the intraocular pressure of the target is 30 mmHg, then it is possible that when using a 50 mmHg blowing air bag to blow the cornea, the blowing air pressure is too large; when using a 40 mmHg blowing air bag to blow the cornea, the blowing air pressure is relatively large; when using a 30 mmHg blowing air bag to blow the cornea, the blowing air pressure is appropriate; and when using a 20 mmHg blowing air bag to blow the cornea, the blowing air pressure is too small.

[0075] In the above - mentioned embodiments, since most people have an intraocular pressure of 10 - 20 mmHg, the non - contact tonometer can set the initial air pressure to 20 mmHg. The non - contact tonometer uses a 20 mmHg blowing air bag to blow the target cornea. The optical imaging analysis system can output a predicted intraocular pressure value of 27 mmHg (with a relatively large deviation from the target intraocular pressure of 30 mmHg) and output the deformation size confidence level. The deformation size confidence level can be a 1×4 matrix result, for example, [0.1, 0.12, 0.94, 0.1]. According to the index number of the maximum value, the analysis result can be determined, that is, in this case of blowing, the situation is that the blowing is too small, and the qualified score is 0.1.

[0076] In this way, the non - contact tonometer can decide to use the air bag with a predicted intraocular pressure value of 27 mmHg to blow the target again. The optical imaging analysis system can analyze based on the blowing air pressure data and optical imaging data input in this blowing, output a predicted intraocular pressure value of 31 mmHg, and output the deformation size confidence level representing that the blowing is too small.

[0077] The optical imaging analysis system can also set a threshold for the size of the fourth column corresponding to the qualified situation in terms of the index number. In this embodiment, when the size of the fourth column corresponding to the qualified situation is above 0.9, it can be considered as a reliable situation of qualified blowing.

[0078] In the deformation size confidence level representing that the blowing is too small output for the second time, the value of the index number representing qualified can be 0.8, which indicates that the blowing of the optical imaging analysis system is already close to being appropriate, but still does not exceed the set threshold of 0.9. Therefore, the analysis result cannot be determined as qualified.

[0079] Then, in response to the analysis result indicating that the blowing is too weak, the optical imaging analysis system can blow air at the target again using an air bag with a predicted intraocular pressure value of 31 mmHg. Based on this blowing, the optical imaging analysis system can output the predicted intraocular pressure value of 30.4 mmHg for the third time and output the confidence level of the deformation size representing compliance. For example, the value of the index number representing compliance in the confidence level of the deformation size can be 0.98, and the score is higher than the set threshold of 0.9. In this way, it can be considered that this blowing is appropriate and the confidence level of the obtained optical imaging image is high, and the predicted intraocular pressure value of 30.4 mmHg is output as the measurement result of the non-contact tonometer.

[0080] The above embodiments are only some optional embodiments provided by the present invention. In actual situations, the non-contact tonometer usually meets the compliance standard when outputting for the second time, and the error between the measured intraocular pressure value result and the actual intraocular pressure value is very small.

[0081] In summary, an optical imaging analysis method, system, and computer-readable storage medium for a non-contact tonometer provided by the present invention can perform optical imaging analysis and intraocular pressure value prediction by combining a neural network model through real-time monitoring of the air blowing pressure of the tonometer nozzle and recording of optical imaging data of the deformation process of the target. Through the optical imaging analysis method for a non-contact tonometer provided by the present invention, the non-contact tonometer can dynamically adjust the precise control of the nozzle air blowing pressure according to the real-time predicted intraocular pressure value, improve the comfort of patient measurement, and improve the accuracy of intraocular pressure value measurement.

[0082] Although the above methods are illustrated and described as a series of actions for simplicity of explanation, it should be understood and appreciated that these methods are not limited by the order of the actions, because according to one or more embodiments, some actions may occur in a different order and / or concurrently with other actions not illustrated and described herein but understood by those skilled in the art.

[0083] Those skilled in the art will understand that information, signals, and data can be represented using any of a variety of different technologies and techniques. For example, the data, instructions, commands, information, signals, bits, symbols, and chips described throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or optical particles, or any combination thereof.

[0084] Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability of hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

[0085] The various illustrative logical modules and circuits described in connection with the embodiments disclosed herein can be implemented or performed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0086] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read from, and write to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In the alternative, the processor and the storage medium may reside as discrete components in a user terminal.

[0087] In one or more exemplary embodiments, the described functionality may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions may be stored on or transmitted via a computer-readable medium as one or more instructions or code. The computer-readable medium includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. The storage media may be any available media that can be accessed by a computer. By way of example and not limitation, such computer-readable media can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a web site, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. As used herein, the disk and disc include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray disc, where the disk typically reproduces data magnetically and the disc optically with a laser. Combinations of the above should also be included within the scope of computer-readable media.

[0088] The foregoing description of the disclosure has been provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An optical imaging analysis method for non-contact tonometer, characterized in that: Includes steps: Acquiring air blowing pressure data and optical imaging data including time information; Aligning and splicing the blowing air pressure data and the optical imaging data in a time dimension to determine spliced ​​data; as well as The spliced ​​data is input into the constructed neural network model, and a matrix result including the predicted intraocular pressure value and the deformation size confidence is output, wherein the neural network model includes a convolution kernel, a splicing kernel and a pooling kernel, and the deformation size confidence includes an invalid value, an over-inflated value, a under-inflated value and a qualified value.

2. The optical imaging analysis method according to claim 1, wherein: The step of aligning and splicing the blowing air pressure data and the optical imaging data in the time dimension to determine the spliced ​​data comprises: adding an optical imaging image to the optical imaging data to form the optical imaging data aligned with the insufflation air pressure data set in a time dimension, the optical imaging image being fitted by a neural network operation; and The insufflation air pressure data and the optical imaging data aligned in the time dimension are stitched to determine stitched data.

3. The optical imaging analysis method according to claim 1, characterized in that: The method also includes a training step of the neural network model, wherein the training step includes: Acquire a plurality of the blowing air pressure data and the optical imaging data including time information; Aligning and splicing the plurality of the puff pressure data and the optical imaging data in a time dimension to determine a data set including a plurality of spliced ​​data, wherein the data set includes an intraocular pressure value label and a confidence label of each spliced ​​data; Inputting the spliced ​​data into the constructed neural network model for training, and outputting a matrix result including the predicted intraocular pressure value and the confidence of the deformation size; and Based on the predicted intraocular pressure value and the deformation size confidence of the matrix result and the intraocular pressure value label and the confidence label of the spliced ​​data, a loss function is calculated to reversely optimize the parameters of the neural network model.

4. The optical imaging analysis method according to claim 1, characterized in that: The neural network model includes multiple convolutional layers, and an attention mechanism is added to at least one of the multiple convolutional layers.

5. An optical imaging analysis system for non-contact tonometer, characterized in that: include: a memory having computer instructions stored thereon; as well as A processor is connected to the memory and is configured to execute computer instructions stored in the memory to implement the optical imaging analysis method for non-contact tonometer according to any one of claims 1 to 4.

6. A non-contact tonometer, characterized in that: include: The blowing system is used to blow air to the target and detect the air pressure value at the nozzle to form the blowing air pressure data; A camera sensor records the optical imaging of the target to form optical imaging data; as well as An optical imaging analysis system for non-contact tonometer as claimed in claim 5.

7. The non-contact tonometer according to claim 6, characterized in that: A slit lamp is included for illuminating the target.

8. The non-contact tonometer according to claim 6, wherein: The camera sensor is an area array camera.

9. An adaptive air blowing control method for a non-contact tonometer, characterized in that: Includes steps: S1: The air blowing system of the non-contact tonometer according to any one of claims 6 to 8 blows air toward the target according to the set pressure and detects the air pressure value at the nozzle to form air blowing pressure data; S2: The camera sensor records the optical imaging of the target and forms optical imaging data; S3: inputting the air puff pressure data and the optical imaging data into an optical imaging analysis system for a non-contact tonometer, and outputting a matrix result including a predicted intraocular pressure value and a confidence level of the deformation size; S4: determining the analysis result according to the maximum value among the invalid value, the excessive blowing value, the insufficient blowing value and the qualified value included in the deformation confidence level; S5: In response to the analysis result being invalid, repeating steps S1 to S4; S6: in response to the analysis result that the air blowing is too large or too small, changing the set pressure to the predicted intraocular pressure value, and repeating steps S1 to S4; and S7: In response to the analysis result being qualified, outputting the predicted intraocular pressure value as a measurement result.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed by the processor, the optical imaging analysis method for non-contact tonometer according to any one of claims 1 to 4 is implemented.

Images