Eye axis multi-parameter measurement method and system

Abstract

The invention discloses an eye axis multi-parameter measurement method and system, and belongs to the technical field of ophthalmology medical treatment, and the method comprises the steps: carrying out the band-pass filtering of a collected interference signal of an eye axis, and obtaining a filtered interference signal; extracting the envelope of the filtered interference signal by using stochastic resonance; performing peak extraction on the envelope; retrieving the extracted wave crest to obtain an original peak position corresponding to multiple parameters of the eye axis; and according to the original peak position and the refractive index corresponding to the multiple parameters of the eye axis, obtaining the multiple parameters of the original eye axis. According to the eye axis multi-parameter measurement method and system, the envelope of the interference signal can be effectively extracted based on stochastic resonance, the peak-to-peak position of the interference signal of the human eye tissue is fully utilized, and the measurement accuracy of the interference signal under the unstable condition is greatly improved.

Description

Technical Field

[0001] The present application relates to the field of ophthalmic medicine, and particularly to a method and system for measuring multiple eye axis parameters. Background Art

[0002] A biometer is a high-precision eye axis parameter measurement device based on optical coherence reflection technology, which can obtain key data such as corneal thickness, anterior chamber depth, lens thickness, vitreous thickness, and eye axis length. As an important organ for perceiving the world, the health of the human eye is crucial for individuals, especially teenagers. Many ophthalmic diseases (such as congenital glaucoma, myopia, hyperopia, etc.) are closely related to eye axis parameters. For example, for every 1 mm increase in eye axis length, the myopia degree increases by about 300 degrees. The normal eye axis length is about 24 mm, while the eye axis of myopic patients can reach 25 mm - 26 mm. Therefore, accurately and stably measuring multiple eye axis parameters is of great significance for disease diagnosis and treatment. For example, the anterior chamber depth directly affects the calculation accuracy of intraocular lens power.

[0003] The processing quality of optical coherence reflection signals directly determines the accuracy of eye axis parameter calculation. However, in the actual measurement process, due to the influence of the system working environment and electronic component noise, the interference signals returned by the human eye are often weak and accompanied by strong noise interference. How to effectively extract useful signals from the noise has become the key to improving measurement accuracy. In addition, the accurate retrieval of the peak value of the interference signal is crucial. For example, the calculation formula of the eye axis length depends on the peak position and refractive index. If the peak recognition error is large, it will directly affect the final measurement result. In the prior art, wavelet thresholding is mostly used for denoising. However, the selection of wavelet transform basis functions has a great influence on the analysis results, and the computational complexity is high, making it difficult to meet the requirements of large-scale data processing. Summary of the Invention

[0004] The present application aims to provide a method and system for measuring multiple eye axis parameters to solve the problem that the difficulty in extracting interference signals affects the measurement results.

[0005] To achieve the above object, the technical solution of the present application is as follows:

[0006] In a first aspect, the present application provides a method for measuring multiple eye axis parameters, including:

[0007] Performing band-pass filtering on the collected interference signals of the eye axis to obtain filtered interference signals;

[0008] Using stochastic resonance to extract the envelope of the filtered interference signals;

[0009] Performing peak extraction on the envelope;

[0010] Retrieving the extracted peaks to obtain the original peak positions corresponding to multiple eye axis parameters;

[0011] The original eye axis multi-parameters are obtained based on the refractive indices corresponding to the original peak positions and the eye axis multi-parameters.

[0012] Optionally, narrowband band-pass filtering is used to perform band-pass filtering on the interference signal.

[0013] Optionally, the use of stochastic resonance to extract the envelope of the filtered interference signal includes:

[0014] Defining a stochastic resonance system based on a non-linear multi-stable system described by a non-linear Langevin equation; inputting the filtered interference signal into the stochastic resonance system to solve the non-linear Langevin equation, so as to obtain the interference signal after stochastic resonance;

[0015] Performing a Hilbert transform on the interference signal after stochastic resonance to obtain an analytic signal;

[0016] Using a sliding window and an interpolation method to extract the analytic signal, so as to obtain the envelope of the interference signal.

[0017] Optionally, the formula of the non-linear Langevin equation is as follows:

[0018] dx(t) / dt = -dU(x) / dx + D*sin(2πft) + ε(t)

[0019] Wherein, D represents the amplitude of the input filtered interference signal, t represents time, f represents the frequency of the input filtered interference signal, x(t) represents the interference signal after stochastic resonance, ε(t) represents Gaussian white noise with a mean of 0, and U(x) represents the potential function of the non-linear Langevin equation.

[0020] Optionally, the window length of the sliding window ≥ the step length, the window length ≤ 2048 and is 2^n.

[0021] Optionally, before performing peak extraction on the envelope, it further includes:

[0022] Calculating the mean value of the analytic signal;

[0023] Analyzing the mode of the lowest amplitude of the human eye tissue surface wave peak in the analytic signal;

[0024] Obtaining a threshold coefficient based on the proportional relationship between the mode of the amplitude and the mean value of the analytic signal;

[0025] Obtaining a dynamic threshold based on the product of the threshold coefficient and the mean value of the analytic signal;

[0026] Setting the mode of the lowest amplitude of the front and back surfaces of the human eye lens in the analytic signal as a fixed threshold.

[0027] Optionally, the second-order difference discrimination method is used to extract the wave peaks of the envelope.

[0028] Optionally, retrieving the original peak positions corresponding to the multi-parameters of the eye axis from the extracted wave peaks includes: screening the wave peaks that meet the dynamic threshold or the fixed threshold from the extracted wave peaks, and using the peak values of the screened wave peaks as the original peak positions corresponding to the multi-parameters of the eye axis.

[0029] Optionally, the formula for obtaining the original multi-parameters of the eye axis according to the original peak positions and the refractive indices corresponding to the multi-parameters of the eye axis is as follows:

[0030] L = f(P, R, coe) = P * R / (4 * coe) / 1000 / 1000

[0031] Wherein, L is the original eye axis parameter, P is the original index difference of the original eye axis parameter, R is the wavelength of the narrowband light source, coe is the refractive index of the original multi-parameters of the eye axis; * represents multiplication; / represents division.

[0032] An eye axis multi-parameter measurement system that executes an eye axis multi-parameter measurement method as described in any one of the above, includes:

[0033] A band-pass filtering processing module for performing band-pass filtering processing on the interference signal of the eye axis collected;

[0034] An envelope extraction module for extracting the envelope of the filtered interference signal by using stochastic resonance;

[0035] A wave peak extraction module for extracting wave peaks of the envelope;

[0036] An original peak position retrieval module for retrieving the original peak positions corresponding to the multi-parameters of the eye axis from the extracted wave peaks;

[0037] An original multi-parameter calculation module for obtaining the original multi-parameters of the eye axis according to the original peak positions and the refractive indices corresponding to the multi-parameters of the eye axis;

[0038] The band-pass filtering processing module, the envelope extraction module, the wave peak extraction module, the original peak position retrieval module, and the original multi-parameter calculation module are connected in sequence.

[0039] The axial length multi-parameter measurement method proposed in this application utilizes the characteristics of the peak-to-peak position of the interference signal of the human eye tissue, significantly improving the measurement accuracy of the system under unstable signal conditions. It can efficiently and accurately extract the envelope of the interference signal, solve the technical problem of inaccurate envelope extraction in complex biological signal processing, and make the measurement results of key parameters such as axial length have higher repeatability and reliability. Even in the face of large individual differences and uneven signal quality, stable and accurate axial length multi-parameter measurement data can still be obtained.

[0040] To make the above features and advantages of the application more obvious and understandable, specific embodiments are hereinafter given and described in detail in conjunction with the accompanying drawings as follows. Brief Description of the Drawings

[0041] Figure 1 It is a flowchart of the axial length multi-parameter measurement method provided by this application.

[0042] Figure 2 It is a structural block diagram of the axial length multi-parameter measurement system provided by this application. Detailed Embodiments

[0043] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below in conjunction with the drawings of the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0044] In an embodiment of this application, please refer to Figure 1 , Figure 1 It is a flowchart of the axial length multi-parameter measurement method provided by this application. This application provides an axial length multi-parameter measurement method, including: Step S1 to Step S5.

[0045] Step S1: Perform band-pass filtering on the collected interference signal of the axial length to obtain the filtered interference signal. [[ID=3}}

[0046] Step S2: Use stochastic resonance to extract the envelope of the filtered interference signal.

[0047] Step S3: Extract the wave peaks of the envelope.

[0048] Step S4: Retrieve the extracted wave peaks to obtain the original peak positions corresponding to the axial length multi-parameters.

[0049] Step S5: Obtain the original axial length multi-parameters according to the original peak positions and the refractive indices corresponding to the axial length multi-parameters.

[0050] The axial length multi-parameter measurement method of the present application can effectively extract the envelope of the interference signal, make full use of the peak positions of the interference signal of the human eye tissue, and greatly improve the measurement accuracy of the interference signal under unstable conditions.

[0051] In step S1, refer to Figure 1 the S1 step in it, and perform band-pass filtering on the collected interference signal of the axial length to obtain the filtered interference signal.

[0052] As an example, first measure the axial length of the human eye to obtain N groups of interference signals. The multi-parameters of the measured axial length can include: corneal thickness, anterior chamber depth, lens thickness, vitreous thickness, and axial length.

[0053] As an example, use narrow-band band-pass filtering to perform band-pass filtering on the interference signal. The center frequency F can be expressed by the following formula:

[0054] F = f(v,λ)

[0055] where v represents the tangential velocity and λ represents the central wavelength of the light source.

[0056] Furthermore, set the bandwidth to ΔMHz, then the cut-off frequencies of the narrow-band band-pass filtering are: (F - Δ / 2)ΔMHz and (F + Δ / 2)ΔMHz. Use the cut-off frequencies (F - Δ / 2)ΔMHz and (F + Δ / 2)ΔMHz to perform band-pass filtering on the collected interference signal to obtain the filtered interference signal.

[0057] In step S2, refer to Figure 1 the S2 step in it, and use stochastic resonance to extract the envelope of the filtered interference signal.

[0058] As an example, when a weak signal submerged in a strong noise background passes through a non-linear system, by adjusting the system parameters, the output signal-to-noise ratio reaches the maximum, and part of the noise energy is converted into interference signal energy, so as to achieve the best matching relationship among the non-linear system, the interference signal, and the noise. Therefore, the present application uses stochastic resonance to extract the envelope of the filtered interference signal. The strong noise in the filtered interference signal will not weaken the role of the interference signal, but will transfer energy to the weak interference signal, greatly improving the signal-to-noise ratio of the output interference signal, that is, using stochastic resonance to weaken the energy of the noise signal and enhance the energy of the weak interference signal to obtain the interference signal after stochastic resonance.

[0059] As an example, step S2 can include the following steps: S21~S23.

[0060] S21: Define a stochastic resonance system based on a non - linear multi - stable system described by a non - linear Langevin equation; input the filtered interference signal into the stochastic resonance system to solve the non - linear Langevin equation, so as to obtain the interference signal after stochastic resonance.

[0061] S22: Perform a Hilbert transform on the interference signal after stochastic resonance to obtain an analytic signal.

[0062] S23: Extract the analytic signal by using a sliding window and interpolation method to obtain the envelope of the interference signal.

[0063] The formula of the non - linear Langevin equation can be expressed as:

[0064] dx(t) / dt = - dU(x) / dx + D*sin(2πft)+ε(t)

[0065] Where, D represents the amplitude of the input filtered interference signal, t represents time, f represents the frequency of the input filtered interference signal, x(t) represents the interference signal after stochastic resonance, ε(t) represents Gaussian white noise with a mean of 0, and U(x) represents the potential function of the non - linear Langevin equation, which is expressed by the formula:

[0066] U(x)=A*x 6 +B*x 4 +C*x 2

[0067] Where, x represents the interference signal after stochastic resonance, and A, B, and C all represent the parameters of the structure of the stochastic resonance system.

[0068] As an example, the window length of the sliding window can be set to H, and the step - shift length can be set to W, where the window length H≥step - shift length W, the window length H≤2048 and is 2^n; calculate the maximum peak value within the real - time sliding window of the analytic signal, and obtain the envelope of the entire interference signal through linear interpolation.

[0069] In step S3, refer to Figure 1 step S3 in, and extract the wave peaks of the envelope.

[0070] As an example, before extracting the wave peaks of the envelope, it may also include: steps S31 to S35.

[0071] Step S31: Calculate the mean value M of the analytic signal.

[0072] Step S32: Analyze the mode of the amplitude with the lowest wave peak on the human eye tissue surface in the analytic signal.

[0073] Step S33: Obtain a threshold coefficient based on the proportional relationship between the mode of the amplitude and the mean of the analytical signal.

[0074] Specifically, there is a proportional relationship between the amplitude of each surface wave peak of different human eye tissue surfaces and the mean value M. Determine the coefficient of the proportional relationship according to the mean value M and the mode of the amplitude, and obtain a threshold coefficient k suitable for most human eye analytical signals.

[0075] Step S34: Obtain a dynamic threshold k*M based on the product of the threshold coefficient k and the mean value M of the analytical signal; wherein, the value range of the threshold coefficient k is 1.65 - 2.2.

[0076] Step S35: Set the mode T of the lowest amplitude of the front and back surface wave peaks of the human eye lens in the analytical signal as a fixed threshold.

[0077] As an example, adopt the second-order difference discrimination method to extract the wave peaks of the envelope, that is, find the points where the first-order difference has different signs and the second-order difference is less than zero. Further, screen the wave peaks that meet the dynamic threshold k*M or the fixed threshold T, and retrieve the peak positions of each screened wave peak.

[0078] In step S4, please refer to Figure 1 the S4 step in

[0079] to retrieve the original peak positions corresponding to multiple eye axis parameters from the extracted wave peaks.

[0080] In step S5, please refer to Figure 1 the S5 step in

[0081] As an example, according to the original peak positions corresponding to multiple eye axis parameters, combined with the refractive indices corresponding to multiple eye axis parameters, obtain the original multiple eye axis parameters, which is expressed by the formula:

[0082] L = f(P, R, coe) = P*R / (4*coe) / 1000 / 1000

[0083] Among them, L represents the original axial length parameter of the eye, P represents the original index difference of the original axial length parameter of the eye, R represents the wavelength of the narrowband light source, and coe represents the refractive index of the original multi-parameter of the axial length of the eye; * represents multiplication; / represents division, corresponding to unit conversion. The unit of the narrowband light source wavelength R is nm, and the unit of the original axial length parameter L is mm.

[0084] In a specific embodiment, N groups of interference signals are measured once, and finally N groups of original multi-parameters of the axial length of the eye are obtained, including: the original axial length parameter, the original corneal thickness, the original anterior chamber depth, the original lens thickness, and the original vitreous thickness, which are respectively represented by the following formulas:

[0085] Axl = ((P5 - P1) / n) * r1310 / (4 * axl_coe) / 1000 / 1000

[0086] Cornea = ((P2 - P1) / n) * r1310 / (4 * cornea_coe) / 1000 / 1000

[0087] Atria = ((P3 - P2) / n) * r1310 / (4 * atria_coe) / 1000 / 1000

[0088] Lens = ((P4 - P3) / n) * r1310 / (4 * lens_coe) / 1000 / 1000

[0089] Vitreous = ((P5 - P4) / n) * r1310 / (4 * vitreous_coe) / 1000 / 1000

[0090] Among them, Axl represents the original axial length parameter, Cornea represents the original corneal thickness, Atria represents the original anterior chamber depth, Lens represents the original lens thickness, and Vitreous represents the original vitreous thickness.

[0091] P1 represents the original peak position on the anterior surface of the cornea; P2 represents the original peak position on the posterior surface of the cornea, P3 represents the original peak position on the anterior surface of the lens, P4 represents the original peak position on the posterior surface of the lens, and P5 represents the original peak position on the posterior surface of the retina.

[0092] n represents the number of interpolations between zero crossings.

[0093] ((P5 - P1) / n) represents the original index difference between the posterior surface of the retina and the anterior surface of the cornea, ((P2 - P1) / n) represents the original index difference between the posterior surface of the cornea and the anterior surface of the cornea, ((P3 - P2) / n) represents the original index difference between the anterior surface of the lens and the posterior surface of the cornea, ((P4 - P3) / n) represents the original index difference between the posterior surface of the lens and the anterior surface of the lens, and ((P5 - P4) / n) represents the original index difference between the posterior surface of the retina and the posterior surface of the lens.

[0094] r1310 indicates that the wavelength R of the narrowband light source is 1310 nm.

[0095] axl_coe represents the refractive index of the eye axis, cornea_coe represents the refractive index of the cornea, atria represents the refractive index of the anterior chamber, lens represents the refractive index of the lens, and vitreous_coe represents the refractive index of the vitreous body.

[0096] Specifically, the accuracy of peak retrieval directly affects the calculation results of multi-parameters of the eye axis.

[0097] As an example, after step S5, it further includes: step S6, processing the original multi-parameters of the eye axis and outputting the measured results of the filtered original multi-parameters of the eye axis.

[0098] As an example, filter out the original multi-parameters of the eye axis with obvious abnormalities, sort each of the N groups of original multi-parameters of the eye axis in ascending order for subsequent selection of the original eye axis parameters. Calculate the statistics of the N groups of original eye axis parameters, including: mean, standard deviation, and coefficient of variation. The average level of the original multi-parameters of the eye axis is reflected by the mean of the original multi-parameters of the eye axis, and the degree of dispersion of the original multi-parameters of the eye axis is measured by the standard deviation and the coefficient of variation. Further, compare the statistics of the original multi-parameters of the eye axis obtained from the N groups of interference signals, and select the original eye axis parameters with the smallest fluctuation of the statistics as the measurement results, and output the measured results of the filtered original multi-parameters of the eye axis.

[0099] In another embodiment of the present application, please refer to Figure 2 , Figure 2 is the module diagram of the eye axis multi-parameter measurement system provided by the present application. The present application provides an eye axis multi-parameter measurement system that executes an eye axis multi-parameter measurement method as described above, including:

[0100] Band-pass filtering processing module 1, used for performing band-pass filtering processing on the interference signal of the eye axis collected;

[0101] Envelope extraction module 2, used for extracting the envelope of the filtered interference signal by using stochastic resonance;

[0102] Peak extraction module 3, used for performing peak extraction on the envelope;

[0103] The original peak position retrieval module 4 is used to retrieve the extracted wave peaks to obtain the original peak positions corresponding to the multi-parameters of the eye axis;

[0104] The original multi-parameter calculation module 5 is used to obtain the original multi-parameters of the eye axis according to the original peak positions and the refractive indices corresponding to the multi-parameters of the eye axis.

[0105] As an example, the band-pass filtering processing module 1, the envelope extraction module 2, the wave peak extraction module 3, the original peak position retrieval module 4, and the original multi-parameter calculation module 5 are connected in sequence.

[0106] As an example, the eye axis multi-parameter measurement system further includes: an original multi-parameter processing module 6, which is connected to the original multi-parameter calculation module 5 and is used to process the original multi-parameters of the eye axis and output the filtered original eye axis multi-parameter measurement results.

[0107] The eye axis multi-parameter measurement method and system proposed in this application utilize the peak position characteristics of the interference signal wave peaks of the human eye tissue, significantly improving the measurement accuracy of the system under unstable signal conditions. It can efficiently and accurately extract the envelope of the interference signal, solve the technical problem of inaccurate envelope extraction in complex biological signal processing, and make the measurement results of key parameters such as the eye axis length have higher repeatability and reliability. Even in the face of large individual differences and uneven signal quality, stable and accurate eye axis multi-parameter measurement data can still be obtained.

[0108] Although this application has been disclosed as above by way of examples, it is not intended to limit this application. Any person with ordinary knowledge in the technical field to which this application pertains may make some modifications and refinements without departing from the spirit and scope of this application. Therefore, the protection scope of this application shall be subject to that defined by the appended patent application scope.

Claims

1. A method for measuring multiple eye axis parameters, characterized in that, including Performing band-pass filtering on the collected interference signal of the eye axis to obtain a filtered interference signal; Using stochastic resonance to extract the envelope of the filtered interference signal; Performing peak extraction on the envelope; Searching the extracted peaks to obtain the original peak positions corresponding to multiple eye axis parameters; Obtaining the original multiple eye axis parameters according to the original peak positions and the refractive indices corresponding to the multiple eye axis parameters; 2. The eye axis multi-parameter measurement method according to claim 1, wherein, Performing band-pass filtering on the interference signal by using narrow-band band-pass filtering; 3. The axial length multi-parameter measurement method according to claim 1, characterized in that The using of stochastic resonance to extract the envelope of the filtered interference signal includes: Defining a stochastic resonance system based on a non-linear multi-stable system described by a non-linear Langevin equation; inputting the filtered interference signal into the stochastic resonance system to solve the non-linear Langevin equation, so as to obtain the interference signal after stochastic resonance; Performing Hilbert transform on the interference signal after stochastic resonance to obtain an analytic signal; Using a sliding window and an interpolation method to extract the analytic signal to obtain the envelope of the interference signal; 4. The axial length multi-parameter measurement method according to claim 3, wherein The formula of the non-linear Langevin equation is as follows: dx(t) / dt=-dU(x) / dx + D*sin(2πft)+ε(t) Wherein, D represents the amplitude of the input filtered interference signal, t represents time, f represents the frequency of the input filtered interference signal, x(t) represents the interference signal after stochastic resonance, ε(t) represents Gaussian white noise with a mean of 0, and U(x) represents the potential function of the non-linear Langevin equation; 5. The axial length multi-parameter measurement method according to claim 3, wherein The window length of the sliding window ≥ the step length, the window length ≤ 2048 and is 2^n; 6. The axial length multi-parameter measurement method according to claim 3, wherein, Before performing peak extraction on the envelope, it further includes: Calculating the mean value of the analytic signal; Analyzing the mode of the lowest amplitude of the surface wave of the human eye tissue in the analytic signal; Obtaining a threshold coefficient based on the proportional relationship between the mode of the amplitude and the mean value of the analytic signal; Obtaining a dynamic threshold based on the product of the threshold coefficient and the mean value of the analytic signal; Setting the mode of the lowest amplitude of the front and back surfaces of the human eye lens in the analytic signal as a fixed threshold; 7. The axial length multi-parameter measurement method according to claim 6, characterized in that, Performing peak extraction on the envelope by using a second-order difference discrimination method; 8. The axial length multi-parameter measurement method according to claim 7, characterized in that, The searching of the extracted peaks to obtain the original peak positions corresponding to multiple eye axis parameters includes: screening the peaks that meet the dynamic threshold or the fixed threshold from the extracted peaks, and using the peak value of the screened peaks as the original peak positions corresponding to the multiple eye axis parameters; 9. The axial length multi-parameter measurement method according to any one of claims 1 to 8, characterized in that, The formula for obtaining the original multiple eye axis parameters according to the original peak positions and the refractive indices corresponding to the multiple eye axis parameters is as follows: L = f(P, R, coe)=P*R / (4*coe) / 1000 / 1000 Wherein, L is the original eye axis parameter, P is the original index difference of the original eye axis parameter, R is the wavelength of the narrow-band light source, coe is the refractive index of the original multiple eye axis parameters; * represents multiplication; / represents division; 10. An eye axis multi-parameter measurement system, which executes an eye axis multi-parameter measurement method according to any one of claims 1-9, characterized in that, including: A band-pass filtering processing module, configured to perform band-pass filtering on the collected interference signal of the eye axis; An envelope extraction module, configured to use stochastic resonance to extract the envelope of the filtered interference signal; A peak extraction module, configured to perform peak extraction on the envelope; An original peak position retrieval module, configured to retrieve the extracted wave peaks to obtain the original peak positions corresponding to the multi-parameters of the eye axis; An original multi-parameter calculation module, configured to obtain the original multi-parameters of the eye axis according to the original peak positions and the refractive indices corresponding to the multi-parameters of the eye axis; The band-pass filtering processing module, the envelope extraction module, the wave peak extraction module, the original peak position retrieval module, and the original multi-parameter calculation module are connected in sequence.

Images