OCT-based eye axis multi-point measurement method
Through the OCT-based multi-point measurement method, the optical fiber coupler and multimodal light source are used, combined with detailed geometric and optical models, the problems of traditional eye axial measurement are solved, and high precision and High reliability measurement of eye length.
Patent Information
- Application Number
- CN202411317915.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-20
- Publication Date
- 2024-11-22
AI Technical Summary
Traditional methods of measuring eye axial lengths such as ultrasonic measurement and ordinary OCT systems have problems such as low accuracy, poor reliability and insufficient adaptability, especially when dealing with complex eye structures, it is difficult to obtain accurate measurement results.
Using OCT-based multi-point measurement method, high-resolution generation of optical fiber coupler spectroscopy, interference signal processing and multi-dimensional signal optimization, combined with detailed geometric and optical models, multi-modal light source and wavelength modulation technology are used to generate high-resolution Grayscale images are processed through complex filters and spatial kernel functions to calculate the eye axis length.
It significantly improves the accuracy and reliability of the measurement of eye axis length, can maintain high accuracy under different physiological conditions, adapt to physiological differences in different individuals, and provides more reliable data support in the measurement of lesional eyeballs or postoperative eye structure.
Smart Images

Figure CN118986264A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to but is not limited to the field of axial length measurement technology, and in particular to an axial length multi-point measurement method based on OCT. Background Art
[0002] Commonly used axial length measurements are divided into two categories: ultrasonic measurement and optical measurement. The principle of ultrasonic measurement is that the eye A-ultrasound transmits ultrasonic waves through the probe. The ultrasonic waves pass through the cornea, retina and other eye tissues, and the interface between the ultrasonic waves and the eye tissues is reflected to form echo signals. These echo signals are captured by the receiver. Based on the time difference and sound speed between the transmitted and received ultrasonic signals, the equipment can calculate the axial length of the eye. However, since ultrasonic measurement requires contact with the ocular surface, the measurement accuracy is low and there is no visual eye image, it is gradually being replaced by optical measurement methods. The optical measurement method mainly uses the principle of light interference to measure the axial length of the eye.
[0003] Optical coherence tomography technology is divided into time-domain OCT and Fourier domain OCT technology. Time-domain OCT technology limits the imaging speed because the reference mirror needs to be moved mechanically during scanning. In addition, the interference of time-domain OCT is limited by the coherence length, and the imaging signal-to-noise ratio performance is low. Therefore, the current mainstream biological measurement OCT technology uses Fourier domain OCT technology. There are two ways to measure the spectrum of Fourier domain OCT: one is to use a grating to separate the interference signal by wavelength, and use a linear array charge-coupled device to simultaneously measure the interference signals of different wavelengths, and divide the spectrum in the spatial domain, which is called spectral domain OCT; the other is to use a swept frequency light source with a scannable wavelength output, output a single wavelength light at each moment, and use a point detection CCD to detect the interference signal of a single wavelength, and divide the spectrum in the time domain. When collecting images, the anterior and posterior segments of the eye are imaged separately by moving the reference mirror, and the whole eye tomographic image is obtained after splicing, and then the axial length of the eye is obtained based on the pixel points and resolution. In terms of optical path design, since traditional axial length measurement only focuses on the distance from the anterior surface of the cornea to the fovea, the imaging of the posterior segment of the eye is often focused on the fovea, and a flat fundus image is obtained. When measuring the axial length, only the position information of the fovea and the anterior surface of the cornea is extracted for calculation.
[0004] Traditional methods for measuring axial length mainly rely on ultrasonic measurement technology, especially A-ultrasound technology. A-ultrasound technology measures axial length by propagating high-frequency sound waves in ocular tissues. Its principle is to use the reflection and refraction of sound waves at different tissue interfaces to calculate the optical path difference and finally obtain the axial length. Although this technology has been widely used in clinical practice and has a certain measurement accuracy, it still has some defects that cannot be ignored. First, ultrasonic measurement requires contact with the surface of the eyeball, which not only causes discomfort to the patient, but also may increase the risk of infection. Secondly, ultrasonic measurement is highly dependent on the operator, and the measurement results are easily affected by human factors, resulting in poor repeatability and accuracy. In addition, ultrasonic technology is easily interfered by the cornea and lens during the measurement process, especially when dealing with complex eye structures such as lesions or post-operative eyeballs, its measurement accuracy is greatly limited. Summary of the invention
[0005] The present disclosure provides a multi-point axial length measurement method based on OCT, which significantly improves the accuracy, reliability and adaptability of axial length measurement.
[0006] In order to solve the above problems, the technical solution of the present invention is achieved as follows:
[0007] A multi-point measurement method of eye axis based on OCT, the method comprising:
[0008] Step 1: The light source emits laser, and the fiber coupler receives and splits the laser to obtain a first split light and a second split light; the first split light enters the reference light path, and the second split light enters the sample light path;
[0009] Step 2: The light in the reference light path is reflected by the reference mirror, and the light in the sample light path is reflected in the human eye tissue; both reflected lights return to the fiber coupler, where they interfere with each other and generate interference signals;
[0010] Step 3: The detector receives the interference signal and transmits the interference signal to the computer. At the same time, a whole-eye cross-sectional image of the human eye tissue is obtained and sent to the computer. After the computer receives the interference signal, it uses the interference signal to generate a grayscale image based on the whole-eye cross-sectional image, analyzes the generated grayscale image, and calculates the axial length of the eye.
[0011] Furthermore, when acquiring the whole-eye tomographic image, an SLD light source or a long coherence swept-frequency light source is used; when using the SLD light source, the reference light path reflector is moved to image the anterior segment of the eye and the posterior segment of the eye respectively, and then they are spliced to obtain the whole-eye tomographic image; when using the long coherence swept-frequency light source, the whole-eye tomographic image is directly acquired; after obtaining the whole-eye tomographic image, the position information of the center of the anterior surface of the cornea and the 5mm range of the macular area is extracted, and after data processing, the length information of the position of each point in the range of 5mm from the center of the anterior surface of the cornea to the macular area is obtained, and the maximum value is selected as the best reference value of the eye axis.
[0012] Furthermore, a measuring probe is arranged in the sample optical path; the measuring probe is used to control and guide the second split light to enter the measured human eye tissue; the measuring probe can be moved in the X-axis direction and the Y-axis direction; by moving the measuring probe, the center of the second split light is aligned with the pupil center in the human eye tissue, ensuring that the second split light passes through the corneal reflection point and makes the corneal reflection point appear in the center of the window.
[0013] Furthermore, the light source in step 1 is a multimodal light source; when emitting laser light, the light source increases the depth resolution by wavelength modulation, and the formula is as follows:
[0014]
[0015] Wherein, λ(t) is the time-varying wavelength of the laser obtained by wavelength modulation; λ0 = 850nm, which is the central wavelength; A n =10 -n nm is the amplitude of each mode; n is the number of modes; ω n =2πn·10 6 rad / s, is the angular frequency of each mode; is the initial phase; σ n =10 -3-n s is the standard deviation of the Gaussian envelope corresponding to each mode; t is time in seconds; s represents seconds.
[0016] Furthermore, in step 2, the intensity of the interference signal generated in the optical fiber coupler is expressed using the following formula:
[0017]
[0018] Where, I(t) is the intensity of the interference signal; I0 = 1mW, which is the reference intensity; OPD(t) is the time-varying optical path difference; is the phase modulation term; is the coherence length; Δλ is the light source bandwidth.
[0019] Furthermore, the time-varying optical path difference OPD(t) is expressed using the following formula:
[0020]
[0021]
[0022] Among them, L0 = 23mm, which is the reference axial length; Δn is the amplitude of the refractive index change; n0 = 1.336, which is the average refractive index in the eye; ω T is the temperature change frequency; φ T is the initial phase of temperature change; τ T is the temperature change time constant; θ x (t) = 0.02 sin(2π 0.1t) is the eye rotation angle function in the X-axis direction; θ y (t) = 0.015·sin(2π·0.15t+π / 4) is the eyeball rotation angle function in the Y-axis direction; C IOP is the intraocular pressure influence coefficient; T IOP is the time constant of intraocular pressure change; ω) IOP is the frequency of intraocular pressure changes;
[0023] Furthermore, in step 3, the grayscale image I is generated by the following formula: gray (x, y):
[0024]
[0025] Among them, x is the X-axis coordinate; y is the Y-axis coordinate; and denote Fourier transform and inverse Fourier transform respectively; H(ω) is a filter; W(ω) is a weighted window function; K(x, y) is a spatial kernel function; V(x, y) is a tissue weight function based on the whole-eye tomographic image; is the convolution operation.
[0026] Furthermore, the filter H(ω) is expressed using the following formula:
[0027]
[0028] Where p is the filter order; α is the smoothing parameter; ω0 is the center frequency; σ ω is the frequency bandwidth; β is the edge suppression coefficient; ωω max is the maximum frequency; the weighted window function W(ω) is expressed using the following formula:
[0029]
[0030] Where q is the window function index; γ is the window function modulation coefficient; the spatial kernel function K(x, y) is expressed using the following formula:
[0031]
[0032] Among them, σ x and σ y are the spatial resolution parameters in the X-axis direction and the Y-axis direction respectively; J0 is the zero-order Bessel function; R is the characteristic radius; the tissue weight function V(x, y) based on the whole eye tomographic image is expressed using the following formula:
[0033]
[0034] Among them, I OCT (x, y, z) is the whole eye tomographic image; z is the image depth; θ is the preset threshold parameter.
[0035] Furthermore, the axial length L is calculated by the following formula: axis :
[0036]
[0037] Among them, κ avg is the central radius of curvature of the cornea; R eye is the average radius of the eyeball; c is the speed of light; n avg is the average reflectivity of the eyeball; τ is the time delay; σ c It is the mean value of the length from the center of the anterior surface of the cornea to each point within 5mm of the macular area; H is the anterior chamber depth; D is the vitreous cavity depth; Δx is the X-axis offset; Δy is the Y-axis offset.
[0038] The present invention discloses an axial length multi-point measurement method based on OCT, which has the following beneficial effects: the present invention significantly improves the accuracy of axial length measurement through complex mathematical modeling and optimization algorithms. Existing axial length measurement technologies, such as traditional ultrasonic measurement and ordinary OCT systems, are often limited by various factors, such as corneal curvature, geometric shape of the eyeball, changes in the refractive index in the eye, etc., resulting in insufficient accuracy of the measurement results. The present invention introduces detailed geometric and optical models, especially considering the central radius of curvature of the cornea, the average radius of the eyeball, and the propagation speed of light in different media in the eye, which can more accurately describe the propagation path of light in the eye, so that high-precision measurement results can still be maintained under different physiological conditions. This precise optical path difference calculation method effectively overcomes the limitations of traditional technologies in processing complex eye structures and provides more reliable data support for clinical diagnosis. The present invention greatly improves the image quality and the reliability of the measurement results through multi-dimensional optimization processing of OCT signals. In the OCT system, the generation and analysis of grayscale images play a key role in the final axial length measurement. Traditional OCT systems often rely on simple image processing methods and are easily affected by noise interference or image blur, resulting in unstable measurement results. The present invention effectively filters and optimizes the noise and irregular components in the OCT signal by introducing complex filters, weighted window functions and spatial kernel functions. The design of the filter not only takes into account the signal characteristics in the frequency domain, but also combines the image characteristics in the spatial domain. By weighted processing of signals of different frequencies and spatial positions, the signal components valuable for measurement are retained to the maximum extent. At the same time, the introduction of the spatial kernel function allows the edges and details in the image to be better preserved and enhanced, thereby generating a high-resolution grayscale image. This multi-dimensional signal optimization processing not only improves the contrast and clarity of the image, but also significantly reduces the impact of noise on the measurement results, ensuring the reliability and consistency of the axial length measurement. The present invention also improves the personalization and adaptability of the measurement results through in-depth analysis of the characteristics of ocular tissues. In practical applications, the structure and physiological state of the eyeball vary from person to person, and traditional fixed models often cannot fully reflect these differences, resulting in significant deviations in the measurement results between different individuals. The present invention can assign different weights to different tissue regions by introducing a tissue weight function based on the whole-eye tomographic image, thereby highlighting the characteristics of these regions in the generated grayscale image. The addition of this tissue characteristic analysis enables the present invention to more flexibly adapt to the physiological differences of different individuals, especially when dealing with diseased eyeballs or postoperative eye structures, and still maintain high-precision measurement results. This not only improves the accuracy of the measurement, but also provides stronger support for the diagnosis and treatment of complex eye cases. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1A schematic diagram of the system structure of a vacuum packaging sealing detection system based on a convolutional neural network provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0040] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present disclosure clearer and more understandable, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.
[0041] Example 1: Reference Figure 1 , a multi-point measurement method of eye axis based on OCT, the method comprising:
[0042] Step 1: The light source emits laser, and the fiber coupler receives and splits the laser to obtain a first split light and a second split light; the first split light enters the reference light path, and the second split light enters the sample light path;
[0043] The choice of light source is crucial to the performance of the OCT system. In this method, the light source used is usually a low-coherence broadband light source, such as a superluminescent diode (SLD) or a broadband light source, in order to provide a wide spectral range, so as to achieve a high depth resolution in the measurement of the eye axis. This low-coherence light source can emit light with a certain bandwidth, rather than a single wavelength laser. The characteristics of the broadband light source enable it to produce high-contrast depth information through interference signals when penetrating the ocular tissue. The laser emitted by the light source enters the entire optical system through a fiber coupler. The fiber coupler plays a key role here. It divides the light beam from the light source into two parts: one part of the light is called the reference light, and the other part of the light is called the sample light. The separation of the reference light and the sample light is achieved by a semi-transparent and semi-reflective mirror in the fiber coupler. The reference light and the sample light enter different optical paths, the reference light path and the sample light path, respectively, and finally interfere with each other in the OCT system to generate an interference signal containing rich depth information. The design of the fiber coupler makes the separation of the light beam very precise, ensuring that the light entering the reference light path and the sample light path has the same initial phase and intensity, thereby ensuring that the interference conditions are met. In the reference optical path, the separated reference light is transmitted in a specific length of optical fiber or free space, and finally reflected by a highly reflective reference mirror. The reflection optical path length of the reference mirror is an adjustable parameter, which together with the optical path difference of the reflected light in the sample optical path determines the phase and intensity of the final interference signal. On the other hand, the light in the sample optical path enters the human eye tissue to be measured. Different eye structures (such as cornea, lens and retina, etc.) have different reflection characteristics. Therefore, at the interface of these structures, the sample light will produce reflected light of different intensities and phases. These reflected lights carry the depth information of the measured eye axis length, and when returning to the fiber coupler, they interfere with the reference light. It is worth noting here that since OCT technology uses a low-coherence light source, a significant interference signal will only be generated when the optical path difference between the reference light and the sample light is within the coherence length range. The coherence length is determined by the bandwidth of the light source. The wider the bandwidth, the shorter the coherence length, thus achieving higher depth resolution. This feature enables OCT technology to very accurately distinguish the different depths of ocular tissue, thereby providing high-precision multi-point measurement results in the axial length measurement of the present invention. In the OCT system, through the spectroscopic operation of the fiber optic coupler, factors such as the loss of light in the optical path and the coupling efficiency must also be considered to ensure that the light beam that ultimately enters the reference optical path and the sample optical path can generate a sufficiently strong interference signal. Any loss or imbalance in coupling efficiency in the fiber optic coupler will cause the interference signal to be attenuated or distorted, thereby affecting the final axial length measurement accuracy. Therefore, when designing a fiber optic coupler, its high precision and low loss characteristics must be ensured to improve the reliability and accuracy of the measurement.
[0044] Step 2: The light in the reference light path is reflected by the reference mirror, and the light in the sample light path is reflected in the human eye tissue; both reflected lights return to the fiber coupler, where they interfere with each other and generate interference signals;
[0045] First, the reference optical path is designed to ensure that the reference light can be reflected back to the fiber coupler with a specific optical path. The position and reflection characteristics of the reference mirror are adjustable, which allows the optical path of the reference light to be precisely controlled to match the optical path experienced by the sample light when it is reflected in different hierarchical structures of the eye. By adjusting the position of the reference mirror, the optical path difference between the reference light and the sample light can be made within the coherence length range, thereby generating a significant interference signal. The coherence length is determined by the bandwidth of the light source. The wider the bandwidth, the shorter the coherence length, thus achieving higher depth resolution. In the reference optical path, the reflection of light is ideal, which means that the reference light can be reflected back to the fiber coupler with almost 100% efficiency, which is crucial for generating a strong interference signal. In the sample optical path, the light enters the eye tissue being measured and is reflected at different eye structures. The structure of the human eye is very complex, including multiple optical medium layers such as the cornea, anterior chamber, lens, vitreous body, and retina, each with different refractive index and reflection characteristics. When the sample light propagates in these layer structures, due to the mismatch of refractive index, the light wave will be partially reflected and partially transmitted at the interface of each layer. The advantage of OCT technology is that it can detect these weak reflection signals, the intensity of which is closely related to the optical properties of the sample medium. When the sample light is reflected back to the fiber coupler in these different levels of ocular structures, it carries the depth information of each layer of structure, which will interfere with the reflected signal of the reference light to form an interference signal. The generation of the interference signal depends on whether the optical path difference between the reference light and the sample light is within the coherence length range. When the optical path difference between the reference light and the sample light is within the coherence length range, the phase difference between the two light waves is small, and a clear interference pattern can be formed. The intensity of the interference signal depends not only on the phase difference of the light waves, but also on the intensity of the reflected light. Since the intensity of the reflected signal of the sample light in different ocular structures is different, the intensity of the interference signal also varies, providing information about different depths of the ocular tissue. At this time, the fiber coupler plays an important role again, which recombines the reference light and the sample light and outputs the interference signal. The accuracy of the fiber coupler directly affects the quality of the interference signal, because even a slight difference in the optical path or a change in the coupling efficiency can cause the interference signal to be distorted or weakened. Therefore, the design of the fiber coupler must ensure efficient light coupling and minimal light loss to maintain the intensity and accuracy of the interference signal. The generated interference signal is a direct reflection of the optical path length of the eye structure, which contains all the depth information from the cornea to the retina. By further processing these interference signals, the specific position and thickness information of different eye structures can be extracted, which is particularly important for the accurate measurement of the axial length of the eye. In traditional axial length measurement methods, the accuracy of the measurement results is limited due to the inability to obtain such high-resolution depth information.The present invention utilizes the interference signal generation and processing mechanism of OCT technology to achieve high-precision measurement of the axial length of the eye. In particular, in the scenario of multi-point measurement, by analyzing the interference signals at multiple depth positions, the entire length of the axial length can be accurately depicted, significantly improving the reliability and accuracy of the measurement.
[0046] Step 3: The detector receives the interference signal and transmits the interference signal to the computer. At the same time, a whole-eye cross-sectional image of the human eye tissue is obtained and sent to the computer. After the computer receives the interference signal, it uses the interference signal to generate a grayscale image based on the whole-eye cross-sectional image, analyzes the generated grayscale image, and calculates the axial length of the eye.
[0047] First, the reception of the interference signal is completed by a highly sensitive photodetector. The photodetector can convert the optical interference signal into an electrical signal. This conversion process requires extremely high sensitivity and low noise performance to ensure that the weak interference signal will not be masked by the detector noise. The performance of the detector directly affects the quality of subsequent signal processing. Therefore, it is crucial to select a suitable detector technology. In the present invention, the detector used can not only capture the interference signal from the reference optical path and the sample optical path, but also distinguish signals of different intensities, thereby providing accurate raw data for subsequent signal processing. Once the interference signal is received by the detector and converted into an electrical signal, these signals will be transmitted to the computer for further processing. In the computer, the received electrical signal is first digitized, including steps such as signal sampling, quantization and filtering. The purpose of digital processing is to convert continuous analog signals into discrete data that can be processed by the computer, while removing noise and other interference components in the signal by filtering, thereby retaining useful information representing the structure of the eye. Since OCT signals often contain high-frequency components, the selection of the sampling frequency needs to satisfy the Nyquist sampling theorem to ensure the integrity of the signal. Next, the computer generates a grayscale image based on the collected signal. Grayscale images are intuitive representations of the intensity of interference signals, reflecting the distribution of reflection intensity at different depths. In OCT technology, the generation process of grayscale images involves Fourier transforming the original signal or other suitable signal processing algorithms to convert the interference signal in the time domain into image information in the depth domain. In the generated grayscale image, different grayscale values correspond to the reflection intensity of different structures of the eye, where areas with higher brightness usually represent strong reflection structures, such as the cornea and retina. Areas with lower brightness may represent the propagation process of light in relatively transparent or less scattered tissues. After the grayscale image is generated, the computer will further analyze the image to extract the length information of the eye axis. The calculation of the eye axis length is based on the identification and positioning of different reflection levels in the grayscale image. In the OCT image, the reflection signals of major structures such as the cornea and retina will form obvious peaks in the image. By identifying the positions of these peaks and combining the optical parameters of the OCT system (such as the wavelength of the light source and the coherence length), the distance between each reflection layer can be accurately calculated. The accumulation of these distances is the length of the eye axis. In addition, the present invention particularly emphasizes the importance of multi-point measurement. Traditional axial length measurement methods usually rely on single-point measurement. Although this method can provide basic information about the length of the axial length, in complex eye structures, especially when there are lesions or morphological abnormalities, the results of single-point measurement may not be accurate enough. Multi-point measurement can obtain a more detailed distribution of axial length by analyzing the reflected signals at multiple different depths, thereby improving the comprehensiveness and accuracy of the measurement.In practical applications, multi-point measurement can not only provide information on the overall length of the eye axis, but also reveal the relative position relationship between different structures of the eye, providing richer information support for clinical diagnosis.
[0048] Embodiment 2: When acquiring a whole-eye tomographic image, an SLD light source or a long coherence swept-frequency light source is used; when using an SLD light source, the reference light path reflector is moved to image the anterior segment of the eye and the posterior segment of the eye respectively, and then they are spliced to obtain a whole-eye tomographic image; when using a long coherence swept-frequency light source, the whole-eye tomographic image is directly acquired; after obtaining the whole-eye tomographic image, the position information of the center of the anterior surface of the cornea and the 5mm range of the macular area is extracted, and after data processing, the length information of the position of each point in the range of 5mm from the center of the anterior surface of the cornea to the macular area is obtained, and the maximum value is selected as the best reference value of the eye axis.
[0049] Specifically, first of all, the SLD light source is a low-coherence light source, which is widely used in OCT systems. Its main advantage is that it can provide a wider spectral bandwidth, which makes the OCT system have a higher resolution in the depth direction. However, due to the short coherence length of the SLD light source, it is impossible to obtain a tomographic image of the entire eye at one time. Therefore, when using the SLD light source, the present invention realizes the separate imaging of the anterior segment and the posterior segment of the eye by moving the reflector in the reference light path. The anterior segment of the eye includes structures such as the cornea, the anterior chamber, and the lens, while the posterior segment of the eye mainly includes the vitreous body and the retina. By moving the reflector and adjusting the optical path length of the reference light, the OCT system can obtain the tomographic images of the anterior segment and the posterior segment of the eye respectively under different optical path conditions. At this time, the process of moving the reflector requires extremely high precision and stability to ensure that the imaging data at different positions can be seamlessly spliced. After the two parts of the image are acquired, the system will splice the tomographic images of the anterior segment and the posterior segment of the eye to obtain a full-eye tomographic image of the entire eyeball. This image stitching technology requires that the consistency of optical parameters and alignment accuracy must be strictly controlled during the imaging process at different positions to avoid measurement deviations caused by stitching errors. Compared with SLD light sources, long coherence swept light sources have longer coherence lengths, which enables them to directly obtain tomographic images of the entire eyeball in a single scan. This light source is becoming more and more common in modern OCT systems, especially in application scenarios that require high throughput and high measurement speed. Long coherence swept light sources can cover the entire length of the eye axis by adjusting the sweep range, so there is no need to perform multiple imaging and stitching operations like SLD light sources. This not only simplifies the measurement process, but also reduces the cumulative errors that may be caused by multiple imaging, further improving the accuracy of axial length measurement. After obtaining the full-eye tomographic image, the system will perform a detailed analysis of the center of the anterior corneal surface and the macular area in the image. The center of the anterior corneal surface is an important reference point for axial length measurement, and the macular area is a key area of the retina, and its precise positioning is crucial to the final measurement result of the axial length. Especially within the 5mm range of the macular area, by measuring the positions of multiple points, the length changes of the eye axis can be more comprehensively evaluated. In this process, image processing technology plays an important role. By extracting the position information of the center of the anterior surface of the cornea and the 5mm range of the macular area, the system can generate length information about the positions of these key points. These length information reflects the actual geometric dimensions of the eye axis. By comparing these length values, the maximum value is selected as the best reference value of the eye axis. This approach is based on the anatomical characteristics of the eye structure, because the maximum value of the eye axis is often directly related to the quality of vision and refractive status.
[0050] Embodiment 3: A measuring probe is arranged in the sample optical path; the measuring probe is used to control and guide the second split light to enter the human eye tissue to be measured; the measuring probe can be moved in the X-axis direction and the Y-axis direction; by moving the measuring probe, the center of the second split light is aligned with the pupil center in the human eye tissue, ensuring that the second split light passes through the corneal reflection point and makes the corneal reflection point appear in the center of the window.
[0051] Specifically, the main function of the measuring probe is to control and guide the second split light from the fiber coupler so that it can accurately enter the tissue of the human eye being measured. In the optical measurement system, the stability and accuracy of the optical path are the basis for obtaining high-quality measurement data. The measuring probe realizes fine control of the light beam through the combination of mechanical structure and optical system, so that it can move in the X-axis and Y-axis directions. This design allows the probe to adjust the incident angle and position of the light beam according to the actual measurement requirements to ensure that the light beam is accurately aligned with the center of the pupil. In the application of the present invention, the mobility of the measuring probe is a significant technical highlight. By moving the measuring probe in the X-axis and Y-axis directions, the system can dynamically adjust the incident path of the light beam to avoid measurement deviations caused by natural micro-movements of the human eye or initial alignment errors of the system. In traditional OCT systems, the alignment of the light beam usually depends on the initial settings and the experience of the operator. This method may cause inaccurate measurements due to improper operation or inherent errors of the equipment. The present invention, by introducing a movable measuring probe, can correct the position of the light beam in real time during the measurement process to ensure that the second split light is always accurately aligned with the center of the pupil, thereby significantly improving the stability and accuracy of the measurement. In addition, the alignment of the center of the pupil is crucial to the accuracy of OCT measurement. The center of the pupil is the main path for light to enter the eyeball, and its position directly affects the propagation path of light in different structures of the eye. If the light beam fails to accurately align with the center of the pupil, the reflection path of the light beam in the eye may deviate, thereby affecting the intensity and quality of the interference signal, and further affecting the measurement result of the axial length of the eye. The present invention aligns the center of the second split light perfectly with the center of the pupil through the precise movement of the measuring probe, ensuring that the light beam can pass through the corneal reflection point along the predetermined path, and making the corneal reflection point accurately appear in the center of the window of the OCT system. The corneal reflection point is the key point where the light beam meets the corneal surface, and its position is related to the stability of the entire optical path. In terms of the presentation of the corneal reflection point, the present invention adopts highly precise optical design and control algorithms, so that the measuring probe can not only ensure that the light beam is aligned with the center of the pupil, but also adjust the direction of the light beam so that the corneal reflection point appears stably in the center of the measurement window. This precise control not only improves the measurement accuracy, but also enhances the adaptability of the system. In particular, when facing the differences in eye structures of different individuals, the present invention can achieve a measurement effect with high consistency and small error through the adjustment of the measuring probe. In addition, the design of the measurement probe also takes into account the safety of the human eye and the comfort of the measurement. Since OCT measurement involves the light beam directly entering the eye tissue, how to reduce the stimulation and discomfort to the subject while ensuring the measurement accuracy is an important technical challenge. The present invention ensures that the optical stimulation during the measurement process is within a safe range and minimizes the eye burden on the subject by precisely controlling the beam power, spot size and incident angle.At the same time, the flexibility of the measurement probe also makes the measurement process more efficient, reducing the time for repeated debugging due to inaccurate beam alignment, thereby improving the overall measurement efficiency.
[0052] Embodiment 4: The light source in step 1 is a multimodal light source; when emitting laser light, the light source increases the depth resolution by wavelength modulation, and the formula is as follows:
[0053]
[0054] Wherein, λ(t) is the time-varying wavelength of the laser obtained by wavelength modulation; λ0 = 850nm, which is the central wavelength; A n =10 -n nm is the amplitude of each mode; n is the number of modes; ω n =2πn·10 6 rad / s, is the angular frequency of each mode; is the initial phase; σ n =10 -3-n s is the standard deviation of the Gaussian envelope corresponding to each mode; t is time in seconds; s represents seconds.
[0055] Specifically, the core of this formula is to use multi-modal superposition to make the wavelength of the light source no longer a fixed value, but present complex dynamic changes over time. The central wavelength λ0 is 850nm, which is located in the near-infrared region and has good tissue penetration ability, enabling the OCT system to effectively penetrate different tissue layers of the eye. However, a single fixed wavelength cannot take into account the measurement accuracy of the entire eye axis at the same time. Therefore, the present invention introduces multiple modes, each with a different amplitude A n , frequency ω n and phase φ n To modulate the wavelength. The amplitude of these modes gradually decreases with the increase of the mode number n, showing a step-by-step attenuation characteristic. This design makes the low mode mainly act on the rough global scan, while the high mode provides a fine resolution of the details. n t+φ n ) is the core of wavelength modulation, which introduces periodic fluctuations, making the wavelength change dynamically within a certain range. By adjusting the angular frequency ω of different modes n , the present invention can realize multi-scale depth scanning. Angular frequency ω n Increasing in multiples of n means that the high-mode fluctuation frequency is higher and can capture more subtle depth changes. This design cleverly balances the imaging speed and depth resolution of the OCT system, allowing the system to simultaneously obtain macroscopic structural information and microscopic details of eye tissue in a shorter measurement time. In addition, the Gaussian envelope in the formula It plays a role in adjusting the contribution of each mode wavelength. As time t changes, the Gaussian envelope will gradually shrink, making the wavelength modulation amplitude show a concentration effect on the time axis. The standard deviation of this concentration effect σ n As the number of modes n increases, it gradually decreases, so that the wavelength modulation of the high mode is mainly concentrated in a shorter time range. Through this design, the contribution of the high mode to the wavelength change is mainly concentrated in a certain range of depth, further enhancing the resolution of the system at this depth, while the low mode provides stable imaging of the overall structure in a wider time range. This wavelength modulation strategy not only improves the depth resolution of the OCT system, but also improves the coherence of the signal to a certain extent. The imaging quality of OCT depends on the coherent interference between light waves. The wavelength change after modulation will affect the phase difference of the light waves, thereby affecting the intensity and distribution of the interference signal. By accurately controlling the dynamic change of the modulated wavelength, the present invention can optimize the coherence conditions so that the system can obtain high-contrast interference signals within different depth ranges. This strategy for optimizing coherence is particularly suitable for the measurement of complex eye structures, such as in the multi-point measurement of the eye axis, the system needs to simultaneously and accurately image multi-layer structures such as the cornea, lens, vitreous body, and retina. Further, the application of wavelength modulation in the present invention not only improves the measurement accuracy of the system, but also expands the application scope of OCT technology. Traditional OCT systems, under fixed wavelength conditions, have difficulty in simultaneously taking into account the overall length of the eye axis and the resolution of each detail. By introducing wavelength modulation of a multimodal light source, the present invention can achieve a seamless transition from global scanning to local fine imaging. This macro-to-micro measurement capability greatly improves the system's adaptability to complex eye structures, especially in ophthalmic clinical diagnosis, and can provide more accurate and comprehensive eye axis measurement data.
[0056] Embodiment 5: In step 2, the intensity of the interference signal generated in the optical fiber coupler is expressed using the following formula:
[0057]
[0058] Where, I(t) is the intensity of the interference signal; I0 = 1mW, which is the reference intensity; OPD(t) is the time-varying optical path difference; is the phase modulation term; is the coherence length; Δλ is the light source bandwidth.
[0059] Specifically, the basic structure of the interference signal intensity formula can be divided into three parts: reference intensity, cosine term and Gaussian term. The reference intensity I0 represents the light intensity when it is not subjected to interference modulation. In the present invention, the reference intensity is set to 1mW, which is a standard light intensity value, which can ensure that the system has sufficient signal intensity during the measurement process to facilitate subsequent signal processing and analysis. The setting of the reference intensity provides a scale for the entire formula, so that the intensity of the interference signal can be directly compared with the standard light intensity. Cosine term It is the core part of the interference signal, which describes the modulation of the coherence of the light wave on the signal intensity. The wavelength λ(t) here is time-varying. This is because the aforementioned wavelength modulation makes the light source have different wavelengths at different time points. This wavelength change directly affects the periodicity of the interference fringes. In the OCT system, the optical path difference OPD(t) is the difference between the reference light path and the sample light path. It determines the phase difference when the two beams of light recombine in the fiber coupler. When the optical path difference meets the coherence condition, the interference signal intensity will have a maximum or minimum value. This phenomenon is reflected in the cosine term in the formula, which manifests as the oscillation of the interference signal. The phase modulation term ψ(t) is another important part of the cosine term. It is introduced by wavelength modulation and its specific form is This phase modulation reflects the effect of different modal wavelength modulation on the phase of the interference signal. Due to the dynamic change of wavelength, the phase will also change, thus affecting the period and amplitude of the interference signal. Through the precise control of phase modulation, the present invention can effectively optimize the coherence of the interference signal, so that the signal strength at different depths can be maintained at a high level, thereby improving the accuracy and resolution of multi-point measurement of the eye axis. The envelope intensity of the interference signal is further modulated. Here, L cis the coherence length, and its value is determined by the central wavelength λ0 of the light source and the bandwidth Δλ. The coherence length is the maximum optical path difference that describes the coherence of light in the system. For the OCT system, the longer the coherence length, the lower the depth resolution of the system, but the larger the depth range that can be detected. The introduction of the Gaussian term causes the interference signal intensity to gradually decay with the increase of the optical path difference. This attenuation characteristic limits the effective detection range of the interference signal, but also ensures the high-resolution measurement capability of the system. Especially in the multi-point measurement of the axial line, this natural attenuation of the interference signal intensity helps to distinguish signals at different depth levels, thereby avoiding signal aliasing. This interference signal intensity formula provides a strong theoretical support for the application of the OCT system in the multi-point measurement of the axial line. Through the precise control of wavelength modulation and optical path difference, the present invention can generate interference signals with high contrast and high resolution, thereby obtaining accurate axial length data at different depth levels. Compared with the traditional fixed-wavelength OCT system, the multimodal wavelength modulation of the present invention not only improves the depth resolution of the system, but also enhances the adaptability to complex eye structures. In practical applications, this technical advantage is manifested in the ability to accurately measure the entire axial length from the cornea to the retina, especially in complex or diseased eyes, the present invention can provide more reliable and precise measurement results. In addition, this formula also reflects the dynamic characteristics of the interference signal, that is, the signal intensity depends not only on the optical path difference, but also on multiple factors such as wavelength modulation and coherence length. This dynamic characteristic enables the OCT system to automatically adjust and optimize the signal quality during the measurement process, thereby maintaining a high level of measurement accuracy under different measurement conditions. For multi-point measurement of the axial length, especially in scenarios requiring high precision and high resolution, this technological innovation significantly improves the reliability and efficiency of the measurement, and reduces the measurement deviation caused by system errors or signal noise.
[0060] Embodiment 6: The time-varying optical path difference OPD(t) is expressed using the following formula:
[0061]
[0062] Among them, L0 = 23mm, which is the reference axial length; Δn is the amplitude of the refractive index change; n0 = 1.336, which is the average refractive index in the eye; ω T is the temperature change frequency; φ T is the initial phase of temperature change; τ T is the temperature change time constant; θ x (t) = 0.02 sin(2π 0.1t) is the eye rotation angle function in the X-axis direction; θ y (t) = 0.015·sin(2π·0.15t+π / 4) is the eyeball rotation angle function in the Y-axis direction; C IOP is the intraocular pressure influence coefficient; τ IOPis the time constant of intraocular pressure change; ω IOP is the frequency of intraocular pressure changes.
[0063] Specifically, the basic part of the formula is a fixed reference axial length L0, which is set to 23mm. This length is based on the average length of the normal axial length of an adult eye and serves as a reference value for the entire measurement system. Axial length is an important parameter of the eye structure, which directly affects the refractive state and visual quality of the eye. In the OCT system, the reference axial length L0 is the initial value of the optical path difference, and changes in all other factors are adjusted based on this basis. By setting this fixed reference length, the present invention can ensure that measurements under different conditions have a unified starting point, thereby reducing measurement errors caused by individual differences or changes in measurement conditions. The next part of the optical path difference is related to the change in the refractive index in the eye. The refractive index n0 is the ratio of the speed of light propagation in different media, which determines the speed at which light propagates in the medium. In the eye, there are slight differences in the refractive index of different tissues, but the overall average refractive index is generally 1.336. In the OCT system, the propagation path of light in the eye will be affected by the change in the refractive index, which is reflected in the change in the optical path difference. The Δn introduced in the formula represents the amplitude of the refractive index change, which is a quantity that changes over time. By using the sin(ω T t+φ T ) term, and this change in refractive index is associated with temperature changes. Temperature is one of the important factors affecting the refractive index. Temperature fluctuations can cause slight changes in tissue density, thereby affecting the refractive behavior of light. In order to more accurately describe this effect, the formula also introduces a time constant τ for temperature changes T , which describes the decay process of refractive index changes over time. In this way, the formula can capture the dynamic impact of refractive index fluctuations caused by temperature changes on the optical path difference, thereby improving the accuracy of the measurement. Next, the eye rotation part of the optical path difference formula is another key factor. The rotation of the eye in the X-axis and Y-axis directions will cause changes in the propagation path of light in these directions, resulting in fluctuations in the optical path difference. In actual operation of the OCT system, small movements of the eye must be taken into account, because even a small rotation can have a significant effect on the measurement results. In the formula, θ x (t) and θ y (t) represents the rotation angle function of the eyeball in the X-axis and Y-axis directions respectively. Through these functions, the formula can accurately describe the effect of the dynamic rotation of the eyeball in two directions on the optical path difference. Specifically, θ x (t) and θ y The sinusoidal form of (t) represents the periodic motion of the eyeball's rotation, which changes the incident angle of the light and thus affects the path length of the light in the different structures of the eyeball. The terms quantify this change, and they represent the increase or decrease in the optical path difference caused by rotation. Through the precise description of this part, the present invention can compensate for the change in optical path difference caused by eyeball rotation, thereby ensuring the stability and accuracy of the measurement. In addition, the formula also takes into account the effect of changes in intraocular pressure (IOP) on the optical path difference. Intraocular pressure is the pressure exerted by the fluid inside the eyeball on the wall of the eyeball. It not only affects the morphology of the eyeball, but also changes the optical properties inside the eye. Changes in intraocular pressure may cause slight changes in the shape of the eyeball, thereby affecting the propagation path of light in the eye. In order to accurately describe this effect, the intraocular pressure influence term C is introduced into the formula. IOP · Here C IOP is the intraocular pressure influence coefficient, which quantifies the specific effect of intraocular pressure changes on the optical path difference. Time constant τ IOP This describes the time course of intraocular pressure changes, indicating that the effect of intraocular pressure changes on optical path difference does not occur instantaneously, but gradually appears over time. IOP and phase φ IOP The periodic behavior of intraocular pressure changes is further defined, and through this term, the formula can capture the dynamic impact of intraocular pressure fluctuations on the optical path difference. Combined with the above factors, the time-varying optical path difference formula OPD(t) proposed in the present invention provides a highly accurate optical path difference model for the OCT system in multi-point measurement of the axial line. This formula not only takes into account the basic axial length, but also introduces multiple variables such as refractive index changes, eye rotation, temperature fluctuations and intraocular pressure changes, so that the system can dynamically adjust the optical path difference to adapt to different measurement conditions. This dynamic adjustment ensures that the OCT system can still maintain high-precision measurement results under various complex physiological and environmental conditions. Compared with the traditional fixed optical path difference model, the time-varying optical path difference model of the present invention has significant advantages. Traditional models often ignore the influence of multiple dynamic factors, resulting in a decrease in measurement accuracy under complex conditions. The present invention achieves real-time adjustment of the optical path difference by accurately modeling these dynamic factors, significantly improving the reliability and accuracy of the measurement. For example, in actual clinical applications, the natural movement of the eyeball and changes in intraocular pressure are inevitable, and if not compensated, it may lead to deviations in the measurement results. However, the model of the present invention calculates the impact of these factors on the optical path difference in real time and dynamically adjusts the phase and intensity of the interference signal, ensuring that the system can still output accurate measurement data even under such undesirable conditions. In addition, the time-varying optical path difference model of the present invention also has wide adaptability and scalability. By adjusting various parameters in the formula, such as the amplitude of refractive index change, temperature change frequency, eye rotation angle, intraocular pressure change frequency, etc., the system can adapt to the physiological characteristics of different individuals and different measurement environments. This flexibility allows the OCT system to be used not only for the measurement of normal eye structures, but also for the fine measurement of pathological or abnormal eye structures, providing strong technical support for the diagnosis and treatment of complex eye cases.
[0064] Example 7: In step 3, the grayscale image I is generated by the following formula gray (x, y):
[0065]
[0066] Among them, x is the X-axis coordinate; y is the Y-axis coordinate; and denote Fourier transform and inverse Fourier transform respectively; H(ω) is a filter; W(ω) is a weighted window function; K(x, y) is a spatial kernel function; V(x, y) is a tissue weight function based on the whole-eye tomographic image; is the convolution operation.
[0067] Specifically, the Fourier transform F and the inverse Fourier transform F -1 It is the core step of grayscale image generation. Fourier transform is a mathematical tool that converts time domain signals into frequency domain signals. In the present invention, Fourier transform is used to convert the interference signal I(t) into the frequency domain. In the frequency domain, the various frequency components of the signal are separated, which enables subsequent filtering and weighting processing to operate more accurately on specific frequency components. The inverse Fourier transform converts the processed frequency domain signal back to the time domain to generate the final grayscale image. In the OCT system, the advantage of frequency domain processing is that specific structural features can be enhanced or suppressed by adjusting different frequency components, thereby improving the quality and resolution of the image. In the frequency domain after Fourier transform, the filter H(ω) and the weighted window function W(ω) play a vital role. The filter H(ω) is mainly used to eliminate noise and other unnecessary frequency components in the interference signal. The presence of noise will interfere with the accuracy of the signal. Therefore, by designing a suitable filter, these interferences can be effectively filtered out, thereby retaining the signal part that is useful for image generation. The design of the filter usually needs to be customized according to the specific signal characteristics and application requirements. The present invention may use a bandpass filter or a low-pass filter to eliminate high-frequency noise or other unnecessary frequency components. The weighted window function W (ω) is used to perform weighted processing on the signal in the frequency domain. The introduction of the window function is to reduce the boundary effect and improve the resolution of the signal within a specific frequency range. The weighted window function usually takes the form of a Hanning window, a Hamming window or a Gaussian window. These window functions can smooth the spectrum, thereby reducing the spectrum leakage problem introduced during the Fourier transform process. In the present invention, the role of the weighted window function W (ω) may be to enhance the signal of certain specific frequency components, thereby improving the contrast and clarity of certain important structures in the image. After filtering and weighted processing in the frequency domain, the signal is inverse Fourier transformed F -1Back to the time domain, the signal has been optimized and processed, ready to generate the final grayscale image. However, in order to further enhance the spatial resolution and fineness of the image, the signal needs to be convolved in the spatial domain. It is a mathematical operation that superimposes one function on another, and is used to achieve multiple functions such as smoothing, sharpening, and edge detection in signal processing and image processing. In the present invention, the spatial kernel function K(x, y) is the core part of the convolution operation. The design of the spatial kernel function determines the specific effect of the convolution operation. For example, the Gaussian kernel function is often used for image smoothing, while the Laplace kernel function is used for edge detection. Through convolution with the spatial kernel function K(x, y), the present invention can optimize the grayscale image in the spatial domain so that the image achieves the best effect in terms of smoothness, sharpness, and edge retention. Spatial convolution can not only remove local noise in the image, but also enhance the local contrast of the image, so that the eye structure is presented in the grayscale image more clearly and delicately. Finally, the tissue weight function V(x, y) based on the whole eye tomographic image is another innovation of the present invention. The tissue weight function is used to weight different eye tissue areas, so as to highlight the characteristics of these areas in the generated grayscale image. Different eye tissues have different optical properties, for example, the cornea and retina have different reflectivities, so when generating a grayscale image, different areas need to be given different weights based on these properties.
[0068] Embodiment 8: The filter H(ω) is expressed using the following formula:
[0069]
[0070] Where p is the filter order; α is the smoothing parameter; ω0 is the center frequency; σ ω is the frequency bandwidth; β is the edge suppression coefficient; ω max is the maximum frequency; the weighted window function W(ω) is expressed using the following formula:
[0071]
[0072] Where q is the window function index; γ is the window function modulation coefficient; the spatial kernel function K(x,y) is expressed using the following formula:
[0073]
[0074] Among them, σ x and σ y are the spatial resolution parameters in the X-axis direction and the Y-axis direction respectively; J0 is the zero-order Bessel function; R is the characteristic radius; the tissue weight function V(x,y) based on the whole eye tomographic image is expressed using the following formula:
[0075]
[0076] Among them, I OCT (x, y, z) is the whole eye tomographic image; z is the image depth; θ is the preset threshold parameter.
[0077] Specifically, the filter H(ω) is designed to optimize the processing of interference signals in the frequency domain. Interference signals carry the depth information of the eye structure, but in actual operation, these signals are often masked by various noises, which affects the clarity of the image and the accuracy of the measurement. In order to effectively eliminate noise and highlight the useful signal components, the filter H(ω) is designed as a high-order filter, and its expression is The part is used to control the passband characteristics of the filter. Here, p represents the order of the filter, which determines the response curve of the filter to different frequency components. By selecting a suitable p value, the passing capacity of high-frequency signals can be enhanced while effectively suppressing low-frequency noise. At the same time, the filter also introduces a Gaussian distribution term It is used to focus on a specific frequency region in the frequency domain. This term is composed of the center frequency ω0 and the frequency bandwidth σ ω Determine and choose appropriate ω0 and σ ω The value can effectively isolate the signal frequency band of interest and reduce sideband interference. This selective processing can significantly improve the contrast of key structures in the image, making the structures of different levels of the eye clearer and easier to identify in the final grayscale image. In addition, the edge suppression coefficient β in the filter further enhances the fineness of image processing. This part can effectively suppress the noise at the edge of the spectrum, thereby reducing the impact of frequency domain edge effects on image quality. Through this precise frequency domain processing, the filter retains the useful signal components to the maximum extent, while suppressing noise and other unnecessary frequency components, ensuring that the generated grayscale image has high resolution and high contrast. Next, the weighted window function W(ω) further optimizes the frequency domain characteristics during signal processing. Window functions are often used in signal processing to reduce frequency domain boundary effects and spectrum leakage problems in Fourier transforms. The weighted window function W(ω) here is designed to be form, which can effectively smooth the signal spectrum and improve image quality. The window function index q determines the shape of the window function, thereby affecting the weighted effect of the signal in the frequency domain. By adjusting the q value, the system can control how each part of the spectrum is weighted, thereby optimizing the clarity of the image. The modulation coefficient γ in the window function further enhances the signal processing capability at a specific frequency. By adjusting the size of γ, the influence of non-essential frequencies can be reduced while retaining important signal components. This processing is particularly effective at the edges and details of the image, making the image not only smoother as a whole, but also clearer in detail, which is crucial for accurate measurement of the eye structure. In the spatial domain part of signal processing, the spatial kernel function K(x,y) plays a core role. The spatial kernel function is a convolution kernel used for image smoothing, edge detection or sharpening. By performing convolution operations with the signal, various spatial domain optimizations of the image can be achieved. In the present invention, the design of K(x,y) adopts a combination of Gaussian function and zero-order Bessel function. This design can highlight the edges and details in the image while maintaining the smoothness of the image. Gaussian part For smoothing images, σ x and σ y Represents the spatial resolution parameters in the X-axis and Y-axis directions respectively. By adjusting these parameters, the blur degree of the image can be controlled to achieve the best smoothing effect. Zero-order Bessel function It is used to enhance the edge features in the image. Bessel function is often used for edge detection and detail enhancement in image processing. Its characteristic is that it can maintain signal strength within a certain distance range, thereby enhancing local details in the image without destroying the overall smoothness of the image. This combined design enables the spatial kernel function to not only effectively remove noise and artifacts, but also maintain overall smoothness while enhancing image edge clarity, providing a high-quality image foundation for subsequent image analysis. Finally, the tissue weight function V(x, y) based on the whole-eye tomographic image is an important step in the present invention to weight the signal by tissue characteristics. The tissue weight function assigns different weight values to different tissue areas according to the reflection characteristics of different tissues in the OCT image. The expression of the weight function is It reflects the edge strength and contrast of the tissue by integrating the square of the image gradient. The gradient intensity reflects the degree of change between different areas in the image. A larger gradient intensity usually corresponds to a tissue boundary or an area with rich details, and these areas often have a higher diagnostic value in image analysis. The weight function uses the hyperbolic tangent function tanh to normalize the gradient intensity so that the final weight value varies between 0 and 1. By adjusting the threshold parameter θ, the system can control which areas are more prominently displayed in the generated grayscale image. This weighted processing based on tissue characteristics enables the OCT system to more flexibly adjust the focus of the image and ensure that the key tissue structure is clearly visible in the grayscale image, thereby improving the accuracy of the diagnosis. In general, these complex signal processing functions in Example 8 greatly improve the image generation quality of the OCT system in multi-point measurement of the eye axis through comprehensive optimization of the frequency domain, spatial domain and depth domain. The filter H(ω) eliminates noise and enhances the signal through precise frequency domain processing, and the weighted window function W(ω) further optimizes the smoothness of the spectrum and the detail performance of the image. The spatial kernel function K(x,y) preserves and enhances the edges and details of the image while smoothing the image, while the tissue weight function V(x,y) ensures that the key tissue structures are highlighted in the grayscale image.
[0078] Example 9: Calculate the axial length L using the following formula: axis :
[0079]
[0080] Among them, κ avg is the central radius of curvature of the cornea; R eye is the average radius of the eyeball; c is the speed of light; n avg is the average reflectivity of the eyeball; τ is the time delay; σ c It is the mean value of the length from the center of the anterior surface of the cornea to each point within 5mm of the macular area; H is the anterior chamber depth; D is the vitreous cavity depth; Δx is the X-axis offset; Δy is the Y-axis offset.
[0081] Specifically, after entering the eyeball, light will pass through multiple media such as the cornea, lens, and vitreous body. The refractive index of these media is different, which causes the propagation speed and path of light in each layer of the medium to change. In order to accurately calculate the optical path difference and ultimately obtain the axial length of the eye, the formula needs to take these changes into account. The first part of the formula The central radius of curvature of the cornea κ avg and the average radius R of the eyeball eyeTogether, they describe the effect of the curvature of the cornea on the path of light. Since the cornea is the first major refractive surface after light enters the eyeball, its curvature directly affects the refraction angle of light, thereby affecting the propagation path of light in the eye. Through this term, the formula corrects the path deviation of light caused by the curvature of the cornea, making the calculation of optical path difference more accurate. Next, the other part of the formula It reflects the propagation characteristics of light in the intraocular medium. The speed of light c is the speed of light in a vacuum, while the average refractive index of the intraocular medium n is avg This determines the actual propagation speed of light in these media. Unlike propagation in a vacuum, light propagates in a medium and its speed is slowed down by the refractive index. Therefore, the formula To calculate the effective propagation speed of light in the medium, and by multiplying it by the depth of the vitreous cavity D, the total time the light travels in the eye is obtained. This part of the calculation is based on the overall change of the light propagation path in the eye, rather than simply considering a certain section of the optical path. In this way, the formula can more comprehensively reflect the actual propagation of light in the complex medium in the eyeball, thus providing an accurate basis for the subsequent optical path difference calculation. It is worth noting that the argmax in the formula τ This part represents the optimization and matching of the time delay of the optical signal. OCT technology relies on the intensity of the interference signal to determine the propagation path of light, and the intensity of the interference signal is closely related to the optical path difference. Since light may have phase differences on different paths, resulting in changes in the strength of the interference signal, it is necessary to find the time delay τ that makes the interference signal the strongest through optimization matching. This optimization process is actually looking for the optimal path for light to propagate in the eye, thereby ensuring that the calculated axial length is the most accurate. By maximizing τ, the formula can eliminate the deviation caused by measurement errors or system noise, making the final axial length calculation result more accurate.
[0082] In addition, the double integral form ∫∫I is used in the formula gray (x,y)·I gray (x+Δx,y+Δy)dxdy is used to analyze the correlation of grayscale images. Grayscale images are obtained by processing OCT data, which reflects the reflective properties and structural characteristics of eye tissues. By calculating the correlation between different pixels in the image, the formula can identify the key features of the eye structure, which is crucial for accurately calculating the axial length of the eye. Each pixel in the grayscale image represents the reflection intensity of light at a specific location, and changes in these intensities often correspond to the boundaries or other structural features of the tissue. Therefore, the formula captures these changes through integral operations to ensure accurate identification and utilization of key structures during the calculation process. The Gaussian weighted term The introduction of further improves the accuracy of correlation analysis. The Gaussian function has the characteristic of giving a larger weight to the central area and a smaller weight to the area far from the center. Through this weighted processing, the formula can effectively focus on the most diagnostically valuable part of the image, rather than averaging the correlation of all pixels. This processing method not only reduces the interference caused by image noise or edge effects, but also ensures that the key structures in the image can be more accurately identified and utilized, thereby improving the accuracy of axial length calculation. Other parameters in the formula, such as the anterior chamber depth H and the vitreous cavity depth D, further improve the description of the eye structure. The anterior chamber depth H reflects the distance between the cornea and the lens, while the vitreous cavity depth D describes the distance between the lens and the retina. These parameters are all anatomical parameters that are crucial in the calculation of axial length. By integrating these parameters into the formula, the present invention can more accurately describe the propagation path of light in the eye, especially when there are large physiological differences between different individuals, this detailed anatomical description is particularly important. The mean length σ of each point in the range of 5mm from the center of the anterior surface of the cornea to the macular area c It is introduced to describe the length changes of different parts of the eye axis. Since the structure of the eyeball is not completely symmetrical, different parts of the eye axis may have different lengths. By statistically analyzing these lengths, an average value σ is obtained. c , which can better reflect the actual situation of the eye axis. This parameter plays a corrective role in the formula, so that the final calculated eye axis length can more accurately represent the optical path of the entire eyeball. In general, this formula is not a simple geometric or optical calculation, but a complex but highly accurate way to determine the eye axis length by combining multiple factors. Each part of the formula represents a deep understanding of the structure and optical properties of the eye, and through precise mathematical modeling, it ensures that accurate measurement results can still be obtained under different physiological conditions. Compared with the traditional eye axis length measurement method, this complex calculation method can significantly improve the accuracy and reliability of the measurement, especially when facing complex or diseased eyeballs, it can provide more accurate and meaningful measurement results. In short, the eye axis length calculation formula of the present invention uses optical coherence tomography technology, combined with detailed geometric and statistical analysis, to comprehensively model the propagation path of light in the eye. Through reasonable parameter setting and optimized calculation, the present invention can still maintain the accuracy and stability of the measurement results under complex physiological conditions. This not only has important application value in the diagnosis of conventional refractive errors, but also can provide strong support and help in the diagnosis of more complex eye diseases.
[0083] The preferred embodiments of the present disclosure are described above with reference to the accompanying drawings, but the scope of the present disclosure is not limited thereby. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the present disclosure shall be within the scope of the present disclosure.
Claims
1. A multi-point measurement method of eye axis based on OCT, characterized in that: The method comprises: Step 1: The light source emits laser, and the fiber coupler receives and splits the laser to obtain a first split light and a second split light; the first split light enters the reference light path, and the second split light enters the sample light path; Step 2: The light in the reference light path is reflected by the reference mirror, and the light in the sample light path is reflected in the human eye tissue; both reflected lights return to the fiber coupler, where they interfere with each other and generate interference signals; Step 3: The detector receives the interference signal and transmits the interference signal to the computer. At the same time, a whole-eye cross-sectional image of the human eye tissue is obtained and sent to the computer. After the computer receives the interference signal, it uses the interference signal to generate a grayscale image based on the whole-eye cross-sectional image, analyzes the generated grayscale image, and calculates the axial length of the eye.
2. The multi-point measurement method of eye axis based on OCT according to claim 1, characterized in that: When acquiring the whole-eye tomographic image, an SLD light source or a long coherence swept-frequency light source is used; when using the SLD light source, the reference light path reflector is moved to image the anterior segment of the eye and the posterior segment of the eye respectively, and then they are spliced to obtain the whole-eye tomographic image; when using the long coherence swept-frequency light source, the whole-eye tomographic image is directly acquired; after obtaining the whole-eye tomographic image, the position information of the center of the anterior surface of the cornea and the 5mm range of the macular area is extracted, and after data processing, the length information of the position of each point within the 5mm range from the center of the anterior surface of the cornea to the macular area is obtained, and the maximum value is selected as the best reference value of the eye axis.
3. The multi-point measurement method of eye axis based on OCT as claimed in claim 2, characterized in that: A measuring probe is arranged in the sample optical path; the measuring probe is used to control and guide the second split light to enter the measured human eye tissue; the measuring probe can move in the X-axis direction and the Y-axis direction; by moving the measuring probe, the center of the second split light is aligned with the pupil center in the human eye tissue, ensuring that the second split light passes through the corneal reflection point and makes the corneal reflection point appear in the center of the window.
4. The multi-point measurement method of eye axis based on OCT as claimed in claim 3, characterized in that: The light source in step 1 is a multimodal light source; when emitting laser light, the light source increases the depth resolution by wavelength modulation, and the formula is as follows: Wherein, λ(t) is the time-varying wavelength of the laser obtained by wavelength modulation; λ0 = 850nm, which is the central wavelength; A n =10 - n nm is the amplitude of each mode; n is the number of modes; ω n =2πn·10 6 rad / s, is the angular frequency of each mode; is the initial phase; σ n =10 -3-n s is the standard deviation of the Gaussian envelope corresponding to each mode; t is time in seconds; s represents seconds.
5. The multi-point measurement method of eye axis based on OCT as claimed in claim 4, characterized in that: In step 2, the intensity of the interference signal generated in the fiber coupler is expressed using the following formula: Where, I(t) is the intensity of the interference signal; I0 = 1mW, which is the reference intensity; OPD(t) is the time-varying optical path difference; is the phase modulation term; is the coherence length; Δλ is the light source bandwidth.
6. The multi-point measurement method of eye axis based on OCT as claimed in claim 5, characterized in that: The time-varying optical path difference OPD(t) is expressed using the following formula: Among them, L0 = 23mm, which is the reference axial length; Δn is the amplitude of the refractive index change; n0 = 1.336, which is the average refractive index in the eye; ω T is the temperature change frequency; φ T is the initial phase of temperature change; τ T is the temperature change time constant; θ x (t) = 0.02 sin(2π 0.1t) is the eye rotation angle function in the X-axis direction; θ y (t) = 0.015·sin(2π·0.15t+π / 4) is the eyeball rotation angle function in the Y-axis direction; C IOP is the intraocular pressure influence coefficient; τ IOP is the time constant of intraocular pressure change; ω IOP is the frequency of intraocular pressure changes.
7. The multi-point measurement method of eye axis based on OCT as claimed in claim 5, characterized in that: In step 3, the grayscale image I is generated by the following formula gray (x,y): Among them, x is the X-axis coordinate; y is the Y-axis coordinate; and represent Fourier transform and inverse Fourier transform respectively; H(ω) is a filter; W(ω) is a weighted window function; K(x,y) is a spatial kernel function; V(x,y) is a tissue weight function based on the whole-eye tomographic image; is the convolution operation.
8. The multi-point measurement method of eye axis based on OCT as claimed in claim 7, characterized in that: The filter H(ω) is expressed using the following formula: Where p is the filter order; α is the smoothing parameter; ω0 is the center frequency; σ ω is the frequency bandwidth; β is the edge suppression coefficient; ω max is the maximum frequency; the weighted window function W(ω) is expressed using the following formula: Where q is the window function index; γ is the window function modulation coefficient; the spatial kernel function K(x,y) is expressed using the following formula: Among them, σ x and σ y are the spatial resolution parameters in the X-axis direction and the Y-axis direction respectively; J0 is the zero-order Bessel function; R is the characteristic radius; the tissue weight function V(x,y) based on the whole eye tomographic image is expressed using the following formula: Among them, I OCT (x, y, z) is the whole eye tomographic image; z is the image depth; θ is the preset threshold parameter.
9. The multi-point measurement method of eye axis based on OCT according to claim 7, characterized in that: The axial length L is calculated by the following formula: axis : Among them, κ avg is the central radius of curvature of the cornea; R eye is the average radius of the eyeball; c is the speed of light; n avg is the average reflectivity of the eyeball; τ is the time delay; σ c It is the mean value of the length from the center of the anterior surface of the cornea to each point within 5mm of the macular area; H is the anterior chamber depth; D is the vitreous cavity depth; Δx is the X-axis offset; Δy is the Y-axis offset.
Citation Information
Cited By
Wide-range and high-precision instantaneous rotation speed sensor and measurement method
CN119510804A
Sweep frequency OCT imaging system phase correction method based on GPU acceleration
CN120436573A
Broadband noise generation method based on predistortion calibration
CN120474878A
Anterior segment optical coherence tomography image measurement method and system
CN122004738A
