Intraocular pressure measurement method and system based on ultrasonic waves and deep learning

Abstract

The invention provides an intraocular pressure measurement method and system based on ultrasonic waves and deep learning. The method comprises the steps that eye ultrasonic data and intraocular pressure data are collected; performing four-dimensional ultrasonic reconstruction on the continuously acquired ultrasonic intraocular pressure data sequence to obtain four-dimensional ultrasonic data reflecting the three-dimensional dynamic change of the internal structure of the eye; establishing a relation model of the four-dimensional eye ultrasonic data and the corresponding intraocular pressure data; constructing and training an LSTM neural network intraocular pressure recognition model through the reconstructed four-dimensional ocular internal structure ultrasonic data, the intraocular pressure data and the relation model; and identifying and displaying the intraocular pressure of the patient based on the trained LSTM neural network intraocular pressure identification model. According to the method, the relationship model of the four-dimensional eye ultrasonic data and the corresponding intraocular pressure data is established, the ultrasonic data features closely related to the intraocular pressure are mined, the LSTM neural network intraocular pressure recognition model is constructed and trained for intraocular pressure recognition, dynamic intraocular pressure measurement is achieved, and the measurement precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intraocular pressure measurement, and specifically to an intraocular pressure measurement method and system based on ultrasound and deep learning. Background Art

[0002] Intraocular pressure is a key indicator in ophthalmic diagnosis and treatment. It is of vital importance for the early detection, disease monitoring and treatment effect evaluation of eye diseases such as glaucoma. Glaucoma is one of the main causes of irreversible blindness worldwide. Its main feature is pathological increase in intraocular pressure, which leads to damage to the optic nerve and gradual loss of visual field. Therefore, accurate and reliable intraocular pressure measurement is crucial for the prevention and treatment of glaucoma.

[0003] Traditional intraocular pressure measurement methods are mainly divided into two categories: contact and non-contact. Among the contact measurement methods, the Goldmann applanation tonometer is regarded as the "gold standard" for intraocular pressure measurement. Its principle is to calculate the intraocular pressure by flattening the cornea to a certain area based on the amount of external force required for flattening. However, this method requires direct contact with the eyeball, which may cause discomfort to the patient and even pose a certain risk of infection, such as corneal abrasions that can easily cause infection. In addition, the measurement process has high technical requirements for the operator, and the measurement results between different operators may differ. The indentation tonometer is also a contact measurement that estimates intraocular pressure by measuring the depth of the indented eyeball, but this method also has risks caused by contact, and its accuracy is easily affected by factors such as the hardness of the eyeball.

[0004] Non-contact measurement methods such as non-contact tonometer (NCT) use airflow pulses to flatten the cornea to a certain extent and calculate intraocular pressure by measuring the changes in reflected light. Although it avoids direct contact with the eyeball, it is easily affected by multiple factors such as the surface condition of the eye (such as tear film stability, corneal surface irregularities, etc.), eye movement, and patient cooperation, and the measurement accuracy is relatively low. Although the dynamic contour tonometer has improved the measurement accuracy to a certain extent, the equipment is expensive and the operation is relatively complicated, making it difficult to be widely popularized.

[0005] In recent years, with the rapid development of ultrasound technology and deep learning algorithms, new opportunities have been brought to intraocular pressure measurement technology. Ultrasound technology has the advantages of being non-invasive and highly penetrable, and can obtain detailed information about the internal structure of the eye. Deep learning algorithms have shown excellent capabilities in processing complex data and pattern recognition. By learning from large amounts of data, they can automatically extract features and establish accurate prediction models. The organic combination of the two is expected to overcome the limitations of traditional intraocular pressure measurement methods, achieve continuous, dynamic, and accurate measurement of intraocular pressure, and provide stronger support for clinical diagnosis and research in ophthalmology. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide an intraocular pressure measurement method and system based on ultrasound and deep learning, so as to solve the technical problems existing in the existing intraocular pressure measurement as described in the background technology.

[0007] To solve the above technical problems, the embodiments of the present invention provide the following technical solutions:

[0008] A method for measuring intraocular pressure based on ultrasound and deep learning, comprising the following steps:

[0009] Collect ocular ultrasound data and intraocular pressure data through head-mounted ultrasound equipment and intraocular pressure measurement equipment;

[0010] Performing four-dimensional ultrasound reconstruction on the continuously collected ultrasound intraocular pressure data sequence to obtain four-dimensional ultrasound data reflecting the three-dimensional dynamic changes of the internal structure of the eye;

[0011] Establish a relationship model between four-dimensional ocular ultrasound data and corresponding intraocular pressure data;

[0012] By reconstructing the four-dimensional ultrasound data of the internal structure of the eye, the intraocular pressure data and the relationship model, the LSTM neural network intraocular pressure recognition model is constructed and trained;

[0013] Based on the trained LSTM neural network intraocular pressure recognition model, the patient's intraocular pressure is classified, identified and displayed.

[0014] Optionally, the continuously collected ultrasonic intraocular pressure data sequence is subjected to four-dimensional ultrasonic reconstruction, specifically: the data sequence is subjected to four-dimensional ultrasonic reconstruction using an image processing algorithm based on spatiotemporal features, assuming that the two-dimensional ultrasonic image sequence has T frames, spatial coordinates (x, y), and the ultrasonic signal intensity of the t-th frame image at the point (x, y) is I t (x, y), and the eye tissue displacement vector field (Δx t (x,y),Δy t (x,y)), the reconstruction formula is:

[0015]

[0016] Where ω(tt′) is the time interpolation weight function.

[0017] Optionally, the establishing of the relationship model between the four-dimensional ultrasound data and the corresponding intraocular pressure data is specifically as follows:

[0018] Data annotation: annotate the four-dimensional ultrasound data and the corresponding intraocular pressure data, and divide the samples into a normal intraocular pressure group and a patient intraocular pressure group according to the intraocular pressure value range;

[0019] Feature extraction and analysis: extract features from four-dimensional ultrasound data, analyze the feature differences of ultrasound data of internal eye structures under different intraocular pressure states, and perform correlation analysis with intraocular pressure values ​​to determine the feature vector related to intraocular pressure;

[0020] A relationship model was constructed, and the mathematical model between ultrasound data features and intraocular pressure values ​​was constructed using the multivariate linear regression method. The characteristic vector was used as the independent variable and the intraocular pressure value was used as the dependent variable. The model parameters were determined by training the model, and the quantitative correspondence between ultrasound data and intraocular pressure was established.

[0021] Optionally, the reconstructed four-dimensional intraocular structure ultrasound data, intraocular pressure data and relationship model are used to construct and train an LSTM neural network intraocular pressure recognition model, specifically:

[0022] The model was trained using ultrasound data and intraocular pressure data samples with established correspondence, and the sample data was divided into training set, validation set and test set in a ratio of 7:2:1 or 8:1:1;

[0023] Construct an LSTM neural network model with 1-3 LSTM layers, 30-100 hidden units in each layer, followed by a fully connected layer for classification output, using the four-dimensional ultrasound data sequence of the internal structure of the eye as input and the corresponding intraocular pressure data as output to train the model;

[0024] The stochastic gradient descent optimization algorithm was used to adjust the model weights and bias parameters through the back propagation algorithm according to the difference between the model prediction results and the actual intraocular pressure data to minimize the loss function.

[0025] Optionally, the intraocular pressure of the patient is identified based on the trained LSTM neural network intraocular pressure recognition model, specifically: the ultrasonic intraocular pressure data is collected and reconstructed in real time, and input into the trained LSTM neural network intraocular pressure recognition model to determine whether the intraocular pressure is normal and whether there is any abnormal change trend, and at the same time calculate the dynamic change indicators of intraocular pressure, including the intraocular pressure fluctuation amplitude and the intraocular pressure change frequency.

[0026] Optionally, the feature extraction of the four-dimensional ultrasound data and the analysis of the feature differences of the ultrasound data of the internal structure of the eye under different intraocular pressure states are specifically as follows:

[0027] The division and location of eye tissue regions uses an image segmentation algorithm to analyze the features of ultrasound images, not limited to grayscale and texture, to identify the boundaries and ranges of eye structures, not limited to the cornea, anterior chamber, lens, vitreous body, and retina. The cornea region is located based on its arc structure and echo intensity; the anterior chamber region is determined based on echo characteristics and the positional relationship with the cornea and lens; the lens is identified by its shape and echo performance; the vitreous body region is divided by its echo characteristics; and the retina is located based on the interface reflection characteristics between it and the choroid.

[0028] Establish an independent coordinate system for each divided eye tissue area, and then accurately calculate and analyze the features in different areas;

[0029] Thickness change rate calculation: for each ocular tissue area, the thickness change rate over time is calculated in the time dimension of the four-dimensional ultrasound data;

[0030] Echo intensity change analysis: statistics on the echo intensity distribution of each eye tissue area in the four-dimensional ultrasound data, dividing the echo intensity into several intervals, and calculating the number or proportion of pixels in each interval at different times;

[0031] Analyze the changing trend of the echo intensity distribution of eye tissues during the change of intraocular pressure, and calculate the average echo intensity change and echo intensity change rate;

[0032] Texture feature extraction and analysis: The texture features of the ultrasound images of eye tissues are extracted using the gray-level co-occurrence matrix analysis method. Multiple characteristic parameters of the gray-level co-occurrence matrix, including but not limited to contrast, correlation, energy, and entropy, are calculated to compare the differences in texture features of eye tissues under different intraocular pressure states.

[0033] Feature correlation analysis and feature vector construction, calculate the correlation coefficient between the extracted thickness change rate, echo intensity change related parameters, texture feature parameters and intraocular pressure values, and based on the correlation analysis results, screen out features that are significantly correlated with intraocular pressure and construct feature vectors.

[0034] Optionally, the correlation coefficient between the calculated extracted thickness change rate, echo intensity change related parameters, texture feature parameters and intraocular pressure value is specifically calculated using the Pearson correlation coefficient method, and the formula is:

[0035]

[0036] Where n is the number of samples, a i and b i are the observed values ​​of variables A and B respectively, and are the means of variables A and B, respectively, where a i represents the thickness change rate, echo intensity change related parameters, texture feature parameters, b i Indicates that i The corresponding intraocular pressure value.

[0037] The present invention proposes an intraocular pressure measurement system based on ultrasound and deep learning, comprising:

[0038] A data acquisition module, used to collect eye ultrasound data and intraocular pressure data;

[0039] A four-dimensional reconstruction module, used to perform four-dimensional ultrasonic reconstruction on the continuously collected ultrasonic intraocular pressure data sequence to obtain four-dimensional ultrasonic data reflecting the three-dimensional dynamic changes of the internal structure of the eye;

[0040] Ultrasound and intraocular pressure relationship building module, used to establish a relationship model between four-dimensional eye ultrasound data and corresponding intraocular pressure data;

[0041] A neural network model training module is used to construct and train an LSTM neural network intraocular pressure recognition model through reconstructed four-dimensional ultrasound data of the internal structure of the eye, intraocular pressure data, and a relationship model;

[0042] The intraocular pressure analysis and recognition module is used to identify the patient's intraocular pressure based on the trained LSTM neural network intraocular pressure recognition model.

[0043] Optionally, the data acquisition module includes an eye ultrasound acquisition submodule, an intraocular pressure data acquisition submodule, and a data transmission module, and the eye ultrasound acquisition submodule and the intraocular pressure data acquisition submodule are both communicatively connected to the data transmission module, wherein:

[0044] (1) The ocular ultrasound acquisition submodule includes:

[0045] Ultrasonic transducer, array surrounds the eye, operating frequency 2-10MHz, multi-angle transmission and reception of ultrasonic signals;

[0046] The signal transmitting and receiving module controls the ultrasonic transducer to transmit pulses and receive reflected echo signals;

[0047] Echo acquisition module, converts analog echo signals into digital signals, with a sampling frequency of 15-45MHz;

[0048] A microprocessor for preprocessing the collected data;

[0049] A communication interface module, used to send the collected ultrasonic signal to an external device, and in the present invention, the ultrasonic signal is transmitted to the data transmission module;

[0050] Power module, using rechargeable battery;

[0051] (4) an intraocular pressure data acquisition submodule, using a tonometer, including but not limited to an applanation tonometer, an indentation tonometer, a non-contact tonometer, and a dynamic contour tonometer;

[0052] (5) A data transmission module, used to send the eye data collected by the eye ultrasound acquisition submodule and the intraocular pressure data acquisition submodule to a server for data analysis and modeling.

[0053] Optionally, it also includes an intraocular pressure display module, which displays the intraocular pressure measurement results and related information in the form of charts and data lists, including but not limited to three-dimensional dynamic charts, line charts, and bar charts.

[0054] The beneficial effects of the above technical solution of the present invention are as follows:

[0055] 1. The present invention uses a head-mounted ultrasound device to continuously collect ocular ultrasound data, and combined with four-dimensional ultrasound reconstruction technology, can obtain three-dimensional dynamic change information of the internal structure of the eye over time, thereby realizing continuous dynamic monitoring of intraocular pressure. This continuous dynamic measurement method can provide more comprehensive and accurate information on intraocular pressure changes, which helps to gain a deeper understanding of the changing laws and trends of intraocular pressure. For patients with large intraocular pressure fluctuations or potential risks of eye diseases, abnormal conditions can be discovered in a timely manner, providing an important basis for early diagnosis and treatment.

[0056] 2. The present invention establishes a relationship model between four-dimensional eye ultrasound data and corresponding intraocular pressure data, and performs correlation analysis on the four-dimensional eye ultrasound data and the corresponding intraocular pressure data. It can deeply explore the impact mechanism of intraocular pressure changes on the internal structure of the eye in spatial and temporal dimensions, and can mine ultrasound data features that are closely related to intraocular pressure. These features can then be applied to the subsequent construction of an intraocular pressure recognition model, which can more accurately reflect the true state of intraocular pressure.

[0057] 3. The present invention establishes a relationship model between four-dimensional ocular ultrasound data and intraocular pressure data, and uses these data to build and train an LSTM neural network intraocular pressure recognition model, giving full play to the deep learning algorithm's ability to process complex data and feature learning capabilities. With the continuous accumulation of measurement data, the system can use new data to continuously update and optimize the LSTM neural network model, and can continuously adapt to changes in different patient groups and various eye conditions, thereby improving the generalization ability and predictive performance of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a flow chart of the intraocular pressure measurement method based on ultrasound and deep learning of the present invention;

[0059] Figure 2 This is a principle block diagram of the intraocular pressure measurement system based on ultrasound and deep learning of the present invention;

[0060] Figure 3 This is a principle block diagram of a data acquisition module of an intraocular pressure measurement system based on ultrasound and deep learning of the present invention;

[0061] Figure 4 This is a principle block diagram of the eye ultrasound acquisition submodule of the intraocular pressure measurement system based on ultrasound and deep learning of the present invention. DETAILED DESCRIPTION

[0062] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0063] like Figure 1 As shown, a method for measuring intraocular pressure based on ultrasound and deep learning comprises the following steps:

[0064] S1. Collect ocular ultrasound data and intraocular pressure data through a head-mounted ultrasound device and an intraocular pressure measurement device;

[0065] S2, performing four-dimensional ultrasound reconstruction on the continuously collected ultrasound intraocular pressure data sequence to obtain four-dimensional ultrasound data reflecting the three-dimensional dynamic changes of the internal structure of the eye;

[0066] S3, establishing a relationship model between four-dimensional eye ultrasound data and corresponding intraocular pressure data;

[0067] S4. Construct and train an LSTM neural network intraocular pressure recognition model through the reconstructed four-dimensional ultrasound data of the internal structure of the eye, intraocular pressure data and relationship model;

[0068] S5. Based on the trained LSTM neural network intraocular pressure recognition model, the patient's intraocular pressure is classified, identified and displayed.

[0069] In step S1, the patient wears the head-mounted ultrasound device and adjusts the position to ensure that the ultrasound transducer fits the eye well without any sense of blockage or pressure. The medical staff enters the patient's basic information (name, age, gender, medical history, etc.) in the supporting software and sets the measurement parameters according to the patient's condition, such as the measurement time interval (5-15 minutes) and the measurement duration (1-24 hours).

[0070] Then, the eye ultrasound data and intraocular pressure data are collected, and the head-mounted ultrasound device automatically starts the ultrasonic transducer to transmit ultrasonic signals at preset time intervals. The transmission frequency is initially 5MHz, and is automatically adjusted between 2-10MHz according to the strength of the reflected signal from the eye tissue. The pulse width is 2-8 microseconds, and the transmission interval is 60-180 microseconds. The ultrasonic transducer array receives the echo signal reflected from the eye tissue, and the data acquisition module converts it into a digital signal with a sampling frequency of 15-45MHz to obtain the eye ultrasound data sequence.

[0071] While collecting ocular ultrasound data, use a standard intraocular pressure measurement device (such as applanation tonometer, indentation tonometer, non-contact tonometer, or dynamic contour tonometer) to measure intraocular pressure. The probe of the tonometer is gently touched to the surface of the patient's eyeball to measure the intraocular pressure value.

[0072] In this embodiment, step S2 performs four-dimensional ultrasound reconstruction on the continuously collected ultrasound intraocular pressure data sequence, specifically including:

[0073] (1) Time dimension integration:

[0074] Assume that the two-dimensional ultrasound image sequence has T frames in the time dimension. For each frame of the image, mark its corresponding time point t (t = 0, 1, ..., T-1). By analyzing the time interval Δt between adjacent frames, the uniformity and accuracy of the time series are ensured. If the time interval is found to be uneven, it may be caused by equipment failure or interference during the acquisition process. Corresponding time correction processing is required, such as adjusting the time series by linear interpolation or resampling, so that the image sequence can accurately reflect the physiological changes of the eye in the time dimension.

[0075] The optical flow method is used to track the motion information of eye tissue between different time frames and calculate the displacement vector field (Δx t (x,y),Δy t (x,y)). The optical flow method is based on the continuity assumption of pixel intensity in the image. By calculating the movement speed of pixels between different frames, the movement direction and displacement of eye tissue are obtained. For key structures of the eye, such as the cornea and lens, their position changes in the time dimension are accurately tracked to provide dynamic information for subsequent spatial integration.

[0076] (2) Spatial Dimension Integration:

[0077] Determine the spatial coordinates (x, y) of the two-dimensional ultrasound image, and clarify the position of each pixel in each frame of the image. According to the characteristics of the anatomical structure of the eye, classify and mark the pixels of different areas (such as the cornea area, anterior chamber area, lens area, etc.) so that the spatial relationship of different tissues can be accurately processed during the integration process.

[0078] Combined with the knowledge of eye anatomy, the two-dimensional ultrasound images of different time frames are spatially aligned and spliced. Considering the possible small rotation and translation movements of the eye during physiological activities, the image registration technology based on feature point matching is adopted. By identifying the key feature points in the image (such as the corneal vertex, lens edge, etc.), the correspondence between different frames is established, and then the image is geometrically transformed (such as rotation, translation, scaling, etc.) according to these correspondences, so that the images of different time frames are accurately aligned in space, and a three-dimensional eye structure image sequence is constructed.

[0079] (3) Four-dimensional ultrasound data generation:

[0080] Using the integrated information of the time dimension and space dimension obtained above, the four-dimensional ultrasound data is calculated according to a specific reconstruction formula. Assume that the ultrasound signal intensity at the point (x, y) of the frame image is I t (x, y), then the reconstruction formula of four-dimensional ultrasound data P(x, y, t) is:

[0081]

[0082] Where ω(tt′) is the time interpolation weight function, according to the time interval Δ t The specific form is determined by the characteristics of the data. Through this formula, the ultrasound signal intensity information of different time frames and different spatial positions is fused to generate four-dimensional ultrasound data that can reflect the dynamic changes of the internal structure of the eye over time. Its spatial resolution can reach 0.1-0.5mm and its temporal resolution can reach 0.1-1 second, which can clearly show the fine dynamic changes of eye tissue and intraocular pressure over time.

[0083] In this embodiment, step S3 establishes a relationship model between four-dimensional ultrasound data and corresponding intraocular pressure data, specifically:

[0084] Data annotation: annotate the four-dimensional ultrasound data and the corresponding intraocular pressure data, and divide the samples into a normal intraocular pressure group and a patient intraocular pressure group according to the intraocular pressure value range (for example, the normal intraocular pressure range is 10-21 mmHg, which can be adjusted according to actual clinical standards).

[0085] Feature extraction and analysis: extract features from four-dimensional ultrasound data and analyze the feature differences of ultrasound data of internal eye structures under different intraocular pressure conditions. For example, calculate characteristic parameters such as the thickness change rate and echo intensity change of eye tissue, and perform correlation analysis with intraocular pressure values ​​to determine the feature vectors that are closely related to intraocular pressure. Methods such as principal component analysis (PCA) can be used to reduce feature dimensions and improve computational efficiency.

[0086] Construct a relationship model, use the multivariate linear regression method to construct a mathematical model between ultrasound data features and intraocular pressure values, take the feature vector as the independent variable, and the intraocular pressure value as the dependent variable. Determine the model parameters by training the model, and establish a quantitative correspondence between ultrasound data and intraocular pressure. Suppose the intraocular pressure value is y, and the ultrasound data feature vector is X = (x1, x2, ..., x n )(where n is the dimension of the feature vector), then the multivariate linear regression model can be expressed as y=β0+β1x1+β2x2+......+β n x n +ε(where β0, β1, β2,…β n is the regression coefficient to be determined, and ε is the error term). i ,y i )(i is the number of samples), and the regression coefficient is estimated by the least squares method. After the model is built, the validation set data is used to evaluate the model and calculate indicators such as the mean square error (MSE). Based on the evaluation results, if the model performance is poor, the model parameters can be adjusted or other improvement measures can be adopted.

[0087] Among them, feature extraction is performed on the four-dimensional ultrasound data to analyze the feature differences of ultrasound data of the internal structure of the eye under different intraocular pressure states, specifically:

[0088] (1) Division and location of ocular tissue regions

[0089] Image segmentation algorithms are used to accurately divide and locate different tissue areas in eye ultrasound images. For example, by analyzing the grayscale, texture and other features of ultrasound images, the boundaries and ranges of major eye structures such as the cornea, anterior chamber, lens, vitreous body, and retina can be accurately identified. For the corneal area, it is located based on its unique arc structure and relatively high echo intensity; the anterior chamber area is determined based on its relatively low echo characteristics and its positional relationship with the cornea and lens; the lens is identified by its special shape and high echo performance; the vitreous area is divided by its relatively uniform low echo characteristics; and the retina is located based on the interface reflection characteristics between it and the choroid.

[0090] An independent coordinate system is established for each divided eye tissue area to facilitate the subsequent accurate calculation and analysis of the features in different areas. For example, a two-dimensional plane coordinate system is established with the corneal vertex as the origin, with the horizontal direction as the x-axis and the vertical direction as the y-axis, to describe the changes in features within the corneal area; for the lens, a three-dimensional coordinate system is established with its center as the origin. In addition to the x and y axes, an additional axis is added to represent the thickness direction of the lens, so that changes in the lens in different dimensions can be more comprehensively analyzed.

[0091] (2) Calculation of thickness change rate

[0092] For each eye tissue area, the rate of change of its thickness over time is accurately calculated in the time dimension of the four-dimensional ultrasound data. Taking the cornea as an example, in each frame of ultrasound image, by measuring the distance between the front and back surfaces of the cornea (the front and back surface boundaries of the cornea tissue are determined based on the grayscale changes in the corneal tissue in the ultrasound image), the corneal thickness value T(t) at different times is obtained (where t represents time).

[0093] Calculate thickness change rate The difference method is used for approximate calculation, that is, (Δt is the time interval between two adjacent frames of images.) By calculating the difference in corneal thickness change rate under different intraocular pressure states, it can be found that when intraocular pressure increases, the cornea may show a trend of increasing thickness or thinning due to pressure, and the rate of change may be related to the magnitude of intraocular pressure change.

[0094] Similarly, the thickness change rate of other eye tissues such as the lens and retina is calculated. Changes in lens thickness may affect its refractive power, while changes in retinal thickness may be related to the effect of intraocular pressure on the retinal nerve fiber layer. By comparing the differences in the thickness change rates of these tissues under normal and high intraocular pressure conditions, important characteristic information related to intraocular pressure can be obtained.

[0095] (3) Analysis of echo intensity changes

[0096] The echo intensity distribution of each eye tissue area in the four-dimensional ultrasound data is counted. The echo intensity is divided into several intervals (for example, divided into three intervals of low, medium, and high according to experience and data characteristics), and the number or proportion of pixels in each interval at different times is calculated.

[0097] Analyze the changing trend of the echo intensity distribution of ocular tissues during changes in intraocular pressure. For example, when intraocular pressure rises, some tissues may change their internal acoustic properties due to structural deformation or changes in blood circulation, which in turn causes changes in echo intensity. This may manifest as an increase in echo intensity in some areas (such as increased tissue density) or a decrease (such as echo attenuation caused by tissue edema).

[0098] Calculation of relevant parameters of echo intensity changes, such as the average echo intensity change and the echo intensity change rate, can serve as important features reflecting the impact of intraocular pressure changes on ocular tissues.

[0099] (4) Texture feature extraction and analysis

[0100] Texture analysis methods such as gray-level co-occurrence matrix (GLCM) are used to extract the texture features of ultrasound images of ocular tissues. Multiple characteristic parameters of GLCM, such as contrast, correlation, energy, entropy, etc., are calculated to describe the uniformity, directionality, and complexity of the internal structure of ocular tissues.

[0101] Compare the differences in the texture features of eye tissues under different intraocular pressure states. For example, under normal intraocular pressure, the texture of eye tissues may be relatively uniform and regular, but when the intraocular pressure increases, the tissue may deform, displace, or undergo microstructural changes, resulting in changes in texture features, such as roughening, enhanced or disordered directionality, etc. By analyzing the changes in these texture features, we can further explore information related to intraocular pressure and provide more basis for intraocular pressure measurement and eye disease diagnosis.

[0102] (5) Feature correlation analysis and feature vector construction

[0103] Calculate the correlation coefficient between the various features extracted above (thickness change rate, echo intensity change related parameters, texture feature parameters, etc.) and the intraocular pressure value. Use the Pearson correlation coefficient method to calculate and determine the strength of the correlation between each feature and intraocular pressure. Among them, the Pearson correlation coefficient method, the formula is:

[0104]

[0105] Where n is the number of samples, a i and b i are the observed values ​​of variables A and B respectively, and are the means of variables A and B, respectively, where a i represents the thickness change rate, echo intensity change related parameters, texture feature parameters, b i Indicates that i The corresponding intraocular pressure value.

[0106] According to the results of correlation analysis, the features with significant correlation with intraocular pressure are screened out and feature vectors are constructed. These features are combined into vector form in a certain order and used as input variables for the subsequent construction of the ultrasound data and intraocular pressure relationship model to accurately reflect the intrinsic relationship between the ocular ultrasound data features and the intraocular pressure value.

[0107] In this embodiment, the step S4 constructs and trains an LSTM neural network intraocular pressure recognition model by reconstructing four-dimensional intraocular structure ultrasound data, intraocular pressure data and a relationship model, specifically:

[0108] The model was trained using ultrasound data and intraocular pressure data samples with established correspondence, and the sample data was divided into training set, validation set and test set in a ratio of 7:2:1 or 8:1:1;

[0109] Construct an LSTM neural network model with 1-3 LSTM layers, 30-100 hidden units in each layer, followed by a fully connected layer for classification output, using the four-dimensional ultrasound data sequence of the internal structure of the eye as input and the corresponding intraocular pressure data as output to train the model;

[0110] The stochastic gradient descent optimization algorithm is used. According to the difference between the model prediction results and the actual intraocular pressure data, the model weights and bias parameters are adjusted through the back propagation algorithm to minimize the loss function (such as mean square error). The model performance is regularly evaluated with the validation set. When the model loss no longer decreases or the accuracy no longer improves on the validation set, the training is stopped to prevent overfitting. Finally, the model performance is evaluated with the test set to ensure that the model has high accuracy and reliability.

[0111] In this embodiment, step S5 classifies, identifies and displays the patient's intraocular pressure based on the trained LSTM neural network intraocular pressure recognition model, specifically: real-time acquisition and reconstructed ultrasonic intraocular pressure data are input into the trained LSTM neural network intraocular pressure recognition model to determine whether the intraocular pressure is normal and whether there is any abnormal change trend, and at the same time calculate the intraocular pressure dynamic change indicators, including the intraocular pressure fluctuation amplitude and the intraocular pressure change frequency.

[0112] like Figure 2 As shown, the present invention proposes an intraocular pressure measurement system based on ultrasound and deep learning, comprising:

[0113] The data acquisition module 101 is used to collect eye ultrasound data and intraocular pressure data;

[0114] A four-dimensional reconstruction module 102 is used to perform four-dimensional ultrasound reconstruction on the continuously collected ultrasound intraocular pressure data sequence to obtain four-dimensional ultrasound data reflecting the three-dimensional dynamic changes of the internal structure of the eye;

[0115] Ultrasound and intraocular pressure relationship building module 103, used to establish a relationship model between four-dimensional eye ultrasound data and corresponding intraocular pressure data;

[0116] A neural network model training module 104 is used to construct and train an LSTM neural network intraocular pressure recognition model through the reconstructed four-dimensional intraocular structure ultrasound data, intraocular pressure data and relationship model;

[0117] The intraocular pressure analysis and recognition module 105 is used to recognize the intraocular pressure of the patient based on the trained LSTM neural network intraocular pressure recognition model.

[0118] In this embodiment, the data acquisition module 101 includes an eye ultrasound acquisition submodule 1011, an intraocular pressure data acquisition submodule 1012, and a data transmission module 1013. The eye ultrasound acquisition submodule 1011 and the intraocular pressure data acquisition submodule 1012 are both communicatively connected to the data transmission module 1013. The data transmission module 1013 can be an RS485 module, an Ethernet module, such as Figure 3 As shown. Among them:

[0119] (1) Figure 4 As shown, the eye ultrasound acquisition submodule 1011 includes:

[0120] Ultrasonic transducer, array surrounds the eye, operating frequency 2-10MHz, multi-angle transmission and reception of ultrasonic signals;

[0121] The signal transmitting and receiving module controls the ultrasonic transducer to transmit pulses and receive reflected echo signals;

[0122] Echo acquisition module, converts analog echo signals into digital signals, with a sampling frequency of 15-45MHz;

[0123] A microprocessor for preprocessing the collected data;

[0124] A communication interface module, used to send the collected ultrasonic signal to an external device, and in the present invention, the ultrasonic signal is transmitted to the data transmission module;

[0125] The power module adopts a rechargeable battery.

[0126] (6) an intraocular pressure data acquisition submodule, using a tonometer, including but not limited to an applanation tonometer, an indentation tonometer, a non-contact tonometer, and a dynamic contour tonometer;

[0127] (7) A data transmission module, used to send the eye data collected by the eye ultrasound acquisition submodule and the intraocular pressure data acquisition submodule to a server for data analysis and modeling.

[0128] In this embodiment, an intraocular pressure display module 106 is also included to display the intraocular pressure measurement results and related information in the form of charts and data lists, including but not limited to three-dimensional dynamic charts, line charts, and bar charts.

[0129] In summary, the intraocular pressure measurement method and system based on ultrasound and deep learning of the present invention uses a variety of advanced technical means to achieve accurate measurement and analysis of intraocular pressure. Its core principle is to collect eye ultrasound data and intraocular pressure data through a head-mounted ultrasound device, use the non-invasiveness and penetration of ultrasound technology to obtain information on the internal structure of the eye, and use a tonometer to obtain accurate intraocular pressure values.

[0130] First, in terms of data processing, an image processing algorithm based on spatiotemporal features is used to reconstruct the ultrasound data sequence in four dimensions. Through the integration of the time dimension, the optical flow method is used to track the movement of eye tissues to ensure that the time series is uniform and accurate, and to obtain the tissue displacement vector field; in the spatial dimension, image registration is performed based on the anatomical structure of the eye to achieve accurate alignment of different frame images, thereby generating four-dimensional ultrasound data that can clearly present the fine dynamic changes of eye tissues and intraocular pressure over time.

[0131] Secondly, a relationship model between four-dimensional ocular ultrasound data and intraocular pressure data was established. The samples were grouped through data annotation, and the features of the four-dimensional ultrasound data were extracted and analyzed, including the calculation of the thickness change rate of each tissue area of ​​the eye, echo intensity change-related parameters, and texture feature parameters, etc., and correlation analysis was performed with the intraocular pressure value, and significant related features were screened to construct feature vectors. Then, the mathematical model was constructed using the multivariate linear regression method to determine the quantitative correspondence between the ultrasound data features and the intraocular pressure value.

[0132] Then, based on the reconstructed four-dimensional ultrasound data, intraocular pressure data and relationship model, the LSTM neural network intraocular pressure recognition model was constructed and trained. Using the data samples with established corresponding relationships, the training set, validation set and test set were divided to construct an LSTM neural network with a specific structure. The four-dimensional ultrasound data sequence was used as input and the intraocular pressure data was used as output. The stochastic gradient descent optimization algorithm was used. According to the difference between the model prediction and the actual data, the model parameters were adjusted through back propagation to minimize the loss function and train a high-precision intraocular pressure recognition model.

[0133] Finally, the real-time collected and reconstructed ultrasound intraocular pressure data is input into the trained LSTM neural network model to realize the identification of the patient's intraocular pressure and the calculation of dynamic change indicators. The intraocular pressure display module displays the results in the form of intuitive charts and data lists, providing comprehensive, accurate and dynamic intraocular pressure information for ophthalmic clinical diagnosis and research.

[0134] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for measuring intraocular pressure based on ultrasound and deep learning, characterized in that: The following steps are involved: Collect ocular ultrasound data and intraocular pressure data through head-mounted ultrasound equipment and intraocular pressure measurement equipment; Performing four-dimensional ultrasound reconstruction on the continuously collected ultrasound intraocular pressure data sequence to obtain four-dimensional ultrasound data reflecting the three-dimensional dynamic changes of the internal structure of the eye; Establish a relationship model between four-dimensional ocular ultrasound data and corresponding intraocular pressure data; By reconstructing the four-dimensional ultrasound data of the internal structure of the eye, the intraocular pressure data and the relationship model, the LSTM neural network intraocular pressure recognition model is constructed and trained; Based on the trained LSTM neural network intraocular pressure recognition model, the patient's intraocular pressure is identified and displayed.

2. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 1, characterized in that: The method of reconstructing the ultrasonic intraocular pressure data sequence collected continuously by four-dimensional ultrasound is specifically as follows: using an image processing algorithm based on spatiotemporal features to reconstruct the data sequence by four-dimensional ultrasound, assuming that the two-dimensional ultrasound image sequence has T frames, spatial coordinates (x, y), and the ultrasound signal intensity of the t-th frame image at the point (x, y) is I t (x, y), and the eye tissue displacement vector field (Δx t (x,y),Δy t (x,y)), the reconstruction formula is: Where ω(tt′) is the time interpolation weight function.

3. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 1, characterized in that: The relationship model between the four-dimensional ultrasound data and the corresponding intraocular pressure data is established as follows: Data annotation: annotate the four-dimensional ultrasound data and the corresponding intraocular pressure data, and divide the samples into a normal intraocular pressure group and a patient intraocular pressure group according to the intraocular pressure value range; Feature extraction and analysis: extract features from four-dimensional ultrasound data, analyze the feature differences of ultrasound data of internal eye structures under different intraocular pressure states, and perform correlation analysis with intraocular pressure values ​​to determine the feature vector related to intraocular pressure; A relationship model was constructed, and the mathematical model between ultrasound data features and intraocular pressure values ​​was constructed using the multivariate linear regression method. The characteristic vector was used as the independent variable and the intraocular pressure value was used as the dependent variable. The model parameters were determined by training the model, and the quantitative correspondence between ultrasound data and intraocular pressure was established.

4. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 1, characterized in that: The LSTM neural network intraocular pressure recognition model is constructed and trained by reconstructing the four-dimensional intraocular structure ultrasound data, intraocular pressure data and relationship model, specifically: The model was trained using ultrasound data and intraocular pressure data samples with established correspondence, and the sample data was divided into training set, validation set and test set in a ratio of 7:2:1 or 8:1:1; Construct an LSTM neural network model with 1-3 LSTM layers, 30-100 hidden units in each layer, followed by a fully connected layer for classification output, using the four-dimensional ultrasound data sequence of the internal structure of the eye as input and the corresponding intraocular pressure data as output to train the model; The stochastic gradient descent optimization algorithm was used to adjust the model weights and bias parameters through the back propagation algorithm according to the difference between the model prediction results and the actual intraocular pressure data to minimize the loss function.

5. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 1, characterized in that: The intraocular pressure of the patient is identified based on the trained LSTM neural network intraocular pressure recognition model, specifically: the ultrasonic intraocular pressure data is collected and reconstructed in real time, input into the trained LSTM neural network intraocular pressure recognition model, and judge whether the intraocular pressure is normal and whether there is any abnormal change trend, and at the same time calculate the dynamic change index of intraocular pressure, including the intraocular pressure fluctuation amplitude and the intraocular pressure change frequency.

6. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 3, characterized in that: The feature extraction of the four-dimensional ultrasound data and the analysis of the feature differences of the ultrasound data of the internal structure of the eye under different intraocular pressure states are specifically as follows: The division and location of eye tissue regions uses an image segmentation algorithm to analyze the features of ultrasound images, not limited to grayscale and texture, to identify the boundaries and ranges of eye structures, not limited to the cornea, anterior chamber, lens, vitreous body, and retina. The cornea region is located based on its arc structure and echo intensity; the anterior chamber region is determined based on echo characteristics and the positional relationship with the cornea and lens; the lens is identified by its shape and echo performance; the vitreous body region is divided by its echo characteristics; and the retina is located based on the interface reflection characteristics between it and the choroid. Establish an independent coordinate system for each divided eye tissue area, and then accurately calculate and analyze the features in different areas; Thickness change rate calculation: for each ocular tissue area, the thickness change rate over time is calculated in the time dimension of the four-dimensional ultrasound data; Echo intensity change analysis: statistics on the echo intensity distribution of each eye tissue area in the four-dimensional ultrasound data, dividing the echo intensity into several intervals, and calculating the number or proportion of pixels in each interval at different times; Analyze the changing trend of the echo intensity distribution of eye tissues during the change of intraocular pressure, and calculate the average echo intensity change and echo intensity change rate; Texture feature extraction and analysis: The texture features of the ultrasound images of eye tissues are extracted using the gray-level co-occurrence matrix analysis method. Multiple characteristic parameters of the gray-level co-occurrence matrix, including but not limited to contrast, correlation, energy, and entropy, are calculated to compare the differences in texture features of eye tissues under different intraocular pressure states. Feature correlation analysis and feature vector construction, calculate the correlation coefficient between the extracted thickness change rate, echo intensity change related parameters, texture feature parameters and intraocular pressure values, and based on the correlation analysis results, screen out features that are significantly correlated with intraocular pressure and construct feature vectors.

7. The intraocular pressure measurement method based on ultrasound and deep learning according to claim 6, characterized in that: The correlation coefficient between the calculated and extracted thickness change rate, echo intensity change related parameters, texture feature parameters and intraocular pressure value is specifically calculated using the Pearson correlation coefficient method, and the formula is: Where n is the number of samples, a i and b i are the observed values ​​of variables A and B respectively, and are the means of variables A and B, respectively, where a i represents the thickness change rate, echo intensity change related parameters, texture feature parameters, b i Indicates that i The corresponding intraocular pressure value.

8. An intraocular pressure measurement system based on ultrasound and deep learning, characterized in that: include: A data acquisition module, used to collect eye ultrasound data and intraocular pressure data; A four-dimensional reconstruction module, used to perform four-dimensional ultrasonic reconstruction on the continuously collected ultrasonic intraocular pressure data sequence to obtain four-dimensional ultrasonic data reflecting the three-dimensional dynamic changes of the internal structure of the eye; Ultrasound and intraocular pressure relationship building module, used to establish a relationship model between four-dimensional eye ultrasound data and corresponding intraocular pressure data; A neural network model training module is used to construct and train an LSTM neural network intraocular pressure recognition model through reconstructed four-dimensional ultrasound data of the internal structure of the eye, intraocular pressure data, and a relationship model; The intraocular pressure analysis and recognition module is used to identify the patient's intraocular pressure based on the trained LSTM neural network intraocular pressure recognition model.

9. The intraocular pressure measurement system based on ultrasound and deep learning according to claim 8, characterized in that: The data acquisition module includes an eye ultrasound acquisition submodule, an intraocular pressure data acquisition submodule, and a data transmission module. The eye ultrasound acquisition submodule and the intraocular pressure data acquisition submodule are both communicatively connected to the data transmission module, wherein: (1) The ocular ultrasound acquisition submodule includes: Ultrasonic transducer, array surrounds the eye, operating frequency 2-10MHz, multi-angle transmission and reception of ultrasonic signals; The signal transmitting and receiving module controls the ultrasonic transducer to transmit pulses and receive reflected echo signals; Echo acquisition module, converts analog echo signals into digital signals, with a sampling frequency of 15-45MHz; A microprocessor for preprocessing the collected data; A communication interface module, used to send the collected ultrasonic signal to an external device; Power module, using rechargeable battery; (2) an intraocular pressure data acquisition submodule, using a tonometer, including but not limited to an applanation tonometer, an indentation tonometer, a non-contact tonometer, and a dynamic contour tonometer; (3) A data transmission module, used to send the eye data collected by the eye ultrasound acquisition submodule and the intraocular pressure data acquisition submodule to a server for data analysis and modeling.

10. The intraocular pressure measurement system based on ultrasound and deep learning according to claim 8, characterized in that: It also includes an intraocular pressure display module, which displays intraocular pressure measurement results and related information in the form of charts and data lists, including but not limited to three-dimensional dynamic charts, line charts, and bar charts.

Images