Eye ball biological parameter measurement method, system, electronic device and readable storage medium

Abstract

The application discloses an eyeball biological parameter measurement method and system, an electronic device and a readable storage medium, relates to the technical field of signal processing and measurement, and comprises the following steps: in response to collecting a plurality of groups of first interference signals corresponding to a plurality of groups of reflecting lenses, respectively, preprocessing each first interference signal to obtain a plurality of groups of second interference signals; packaging and splicing each second interference signal to obtain a plurality of groups of third interference signals; fitting each third interference signal and then aligning to obtain a plurality of groups of fourth interference signals; and extracting a target eyeball biological parameter from each fourth interference signal. The application first fits each interference signal and then aligns, improves the accuracy of the obtained fourth interference signal, eliminates signal drift and coordinate system offset caused by mechanical jitter of a rotating mechanism and fixation micro-motion of a patient, and improves the accuracy of eyeball biological parameter measurement.

Description

Technical Field

[0001] This application relates to the field of signal processing and measurement technology, and in particular to a method, system, electronic device, and readable storage medium for measuring ocular bioparameters. Background Technology

[0002] With the development of ophthalmic medical technology, ocular bioparameter measurement techniques based on the principle of low-coherence optical reflection have been widely used in fields such as preoperative cataract assessment and myopia control. The measurement accuracy of this type of technology depends on accurately identifying the position of the reflection peaks at the interfaces of various tissues from the interference signal waveform.

[0003] When locating the reflection peaks at tissue interfaces in the interference signal waveform from the reflecting lens, the mainstream approach uses a simple threshold method or direct peak search. However, when faced with weak reflection signals, it is difficult to reliably determine their precise locations, resulting in insufficient accuracy of the interference signal and affecting the measurement accuracy of the final extracted ocular biological parameters.

[0004] Therefore, overcoming the shortcomings of traditional methods in measuring ocular bioparameters with insufficient accuracy is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] The main objective of this application is to provide a method, system, electronic device, and readable storage medium for measuring ocular bioparameters, aiming to solve the technical problem of insufficient measurement accuracy of ocular bioparameters.

[0006] To achieve the above objectives, this application proposes a method for measuring ocular bioparameters, the method comprising: In response to the acquisition of first interference signals corresponding to multiple sets of reflective mirrors, each of the first interference signals is preprocessed to obtain multiple sets of second interference signals; Each of the second interference signals is encapsulated and spliced ​​together to obtain multiple sets of third interference signals; The third interference signals are first fitted and then aligned to obtain multiple sets of fourth interference signals; Biological parameters of the target eyeball are extracted from each of the fourth interference signals.

[0007] In one embodiment, the step of preprocessing each of the first interference signals includes: By removing the DC component of each of the first interference signals, multiple sets of zero-mean interference signals are obtained; Demodulate each of the zero-mean interference signals to obtain multiple sets of first filter matrices; By filtering out isolated outliers in each of the first filter matrices, multiple sets of second filter matrices are obtained. Each of the second filter matrices is downsampled to obtain the second interference signal.

[0008] In one embodiment, the step of fitting and aligning each of the third interference signals to obtain multiple sets of fourth interference signals includes: Acquire each of the third interference signals within a preset threshold range, obtain significant signal points, and determine multiple sets of intermediate peak indices; By fitting the signal segments corresponding to each of the intermediate peak indices, multiple sets of precise peak positions are obtained; Each precise peak position is captured into a window, and the signal within each window is accumulated to obtain multiple sets of intra-group accumulated signals.

[0009] In one embodiment, the step of determining multiple sets of intermediate peak indices includes: Determine whether the spacing between the significant signal points is within a preset spacing range; When the spacing between the significant signal points is within a preset spacing range, the two candidate peak indices within each significant signal point are respectively used as the intermediate peak index; If the spacing between the significant signal points is not within a preset spacing range, the first consecutive signal block within each significant signal point is used as the intermediate peak index.

[0010] In one embodiment, the intra-group accumulated signal includes at least a reference group signal and multiple compensated non-reference group signals; after the step of extracting windows corresponding to each of the precise peak positions and accumulating the signals within each window to obtain multiple intra-group accumulated signals, the method further includes: Based on the signal-to-noise ratio of the accumulated signals within each group, a reference group signal and multiple non-reference group signals are determined, wherein the signal-to-noise ratio of the reference group signal is higher than that of each of the non-reference group signals; The reference group signal and each of the non-reference group signals are cross-correlated to obtain multiple sets of offsets. Each of the non-reference group signals is compensated based on the offset to obtain multiple sets of compensated non-reference group signals.

[0011] In one embodiment, the step of extracting target eye biological parameters from each of the fourth interference signals includes: Multiple sets of original ocular biological parameters were extracted from the fourth interference signal; Calculate the minimum variance or the average value of each of the target eye biological parameters compared to the preset data; By combining the original ocular biological parameters corresponding to the average value or the minimum variance of each of the above, intermediate ocular biological parameters are obtained; The intermediate ocular bioparameters are converted to obtain the target ocular bioparameters.

[0012] Furthermore, to achieve the above objectives, this application also proposes an ocular bioparameter measurement system, which includes: The preprocessing module is used to preprocess each of the first interference signals corresponding to the multiple sets of reflective mirrors to obtain multiple sets of second interference signals in response to the acquisition of the first interference signals. The encapsulation module is used to encapsulate and splice each of the second interference signals to obtain multiple sets of third interference signals; The alignment module is used to first fit and then align each of the third interference signals to obtain multiple sets of fourth interference signals; An extraction module is used to extract target eye biological parameters from each of the fourth interference signals.

[0013] In addition, to achieve the above objectives, this application also proposes an electronic device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the ocular bioparameter measurement method as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a readable storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the ocular bioparameter measurement method described above.

[0015] This application provides a method for measuring ocular biological parameters, which includes: in response to acquiring first interference signals corresponding to multiple sets of reflecting lenses respectively, preprocessing each of the first interference signals to obtain multiple sets of second interference signals; encapsulating and stitching each of the second interference signals to obtain multiple sets of third interference signals; fitting and aligning each of the third interference signals to obtain multiple sets of fourth interference signals; and finally extracting target ocular biological parameters from each of the fourth interference signals.

[0016] This application improves the standardization of interference signals by encapsulating and stitching the interference signals collected from each group of reflective lenses, thus avoiding the second interference signal output after preprocessing from directly entering the subsequent alignment process in a scattered form. Then, it first fits each interference signal and then aligns it, which improves the accuracy of the obtained fourth interference signal and eliminates signal drift and coordinate system offset caused by mechanical jitter of the rotating mechanism and micro-movement of the patient's fixation, thereby improving the accuracy of ocular biological parameter measurement. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an embodiment of the ocular bioparameter measurement method of this application. Figure 2 This is a schematic diagram of the overall system architecture provided in Embodiment 1 of the ocular bioparameter measurement method of this application; Figure 3 This is a schematic diagram of the multi-frequency orthogonal matched filter structure of the ocular bioparameter measurement method provided in Embodiment 3 of this application; Figure 4 This is a schematic diagram of the intragroup signal alignment and accumulation process of the ocular bioparameter measurement method provided in Embodiment 2 of this application; Figure 5 This is a schematic diagram of the intergroup offset alignment process of the ocular bioparameter measurement method provided in Embodiment 2 of this application; Figure 6 This is a schematic diagram of the parameter detection process for the first half of the ocular bioparameter measurement method provided in Embodiment 4 of this application. Figure 7 This is a schematic diagram of the retinal parameter detection process in the posterior region of the ocular bioparameter measurement method provided in Embodiment 4 of this application. Figure 8 This is a schematic diagram of subsampling point-level signal alignment based on a cross-correlation algorithm for the ocular bioparameter measurement method provided in Embodiment 2 of this application. Figure 9 This is a schematic diagram of the module structure of the ocular bioparameter measurement system according to an embodiment of this application; Figure 10 This is a schematic diagram of the device structure of the hardware operating environment involved in the ocular bioparameter measurement method in the embodiments of this application. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0021] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0022] The main solution of this application embodiment is as follows: in response to the acquisition of first interference signals corresponding to multiple sets of reflective lenses, each first interference signal is preprocessed to obtain multiple sets of second interference signals; each second interference signal is encapsulated and spliced ​​to obtain multiple sets of third interference signals; each third interference signal is first fitted and then aligned to obtain multiple sets of fourth interference signals; and the target eye biological parameters are extracted from each fourth interference signal.

[0023] Because existing technologies often use simple thresholding or direct peak search with integer indexing to locate reflection peaks at tissue interfaces in interference signal waveforms, it is difficult to reliably determine the precise location of weak reflection signals. Furthermore, there is a lack of effective compensation for the relative offset between multiple lens signals caused by mechanical jitter and micromotions in patient fixation, resulting in insufficient peak positioning accuracy at tissue interfaces and affecting the measurement accuracy and consistency of the final extracted ocular biological parameters.

[0024] This application provides a solution that determines the peak position of the signal within each group by continuously fitting during alignment processing, and performs translation compensation on the signals of each group based on the relative offset between groups, so that the peak position positioning accuracy breaks through the integer level limitation of sampling points, solves the technical problem of insufficient signal alignment accuracy in the prior art, and improves the measurement accuracy and repeatability of ocular biological parameters.

[0025] Based on this, embodiments of this application provide a method for measuring ocular bioparameters, referring to... Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the ocular bioparameter measurement method of this application.

[0026] In this embodiment, the method for measuring ocular biological parameters includes the following steps: Step S110: In response to the acquisition of first interference signals corresponding to multiple sets of reflective mirrors, preprocess each first interference signal to obtain multiple sets of second interference signals; It should be noted that the first interference signal refers to the original interference signal collected by multiple sets of reflecting mirrors. This signal carries optical information reflected back from the interfaces of various tissues such as the cornea, lens, and retina, but it is also mixed with DC offset introduced by the dark current of the detector and the average optical power of the light source, noise components other than the modulation frequency of each tissue, and isolated pulse interference caused by detector defects or transient jitter.

[0027] For example, taking a set of reflecting mirrors as an example, the corresponding first interference signal is a time-domain waveform distributed along the sampling point axis. The waveform contains multiple reflection peaks corresponding to different tissue interfaces. The amplitude and width of each reflection peak are different due to the difference in reflectivity of the tissue interface. Among them, the amplitude of the reflection peaks of deep tissues such as the vitreoretinal interface is usually weak and easily submerged by noise.

[0028] It should be noted that the second interference signal is an intermediate signal obtained by preprocessing the first interference signal. Compared with the first interference signal, the DC offset in the second interference signal has been eliminated, the modulation components corresponding to each tissue interface have been separated and extracted from the aliasing signal, isolated pulse interference has been suppressed, and the data sampling rate has been adapted to the bandwidth requirements of subsequent processing stages.

[0029] Preprocessing includes sequential DC removal, frequency domain demodulation, nonlinear filtering, and downsampling. DC removal eliminates the DC offset component in the first interference signal; frequency domain demodulation separates and extracts the modulation frequency points corresponding to each tissue interface from the aliased signal to obtain multiple envelope matrices; nonlinear filtering suppresses isolated outliers in the envelope matrix; and downsampling reduces the sampling rate of the envelope matrix to a level suitable for subsequent transmission and processing.

[0030] Understandably, this step preprocesses the original interference signal, converting the original noisy and aliased first interference signal into a second interference signal with improved signal-to-noise ratio, compressed data volume, and regular shape, thus providing a high-quality input foundation for subsequent encapsulation, splicing, and alignment processing.

[0031] Step S120: Encapsulate and splice each of the second interference signals to obtain multiple sets of third interference signals; It should be noted that the third interferometric signal is a structured signal obtained by encapsulating and splicing the preprocessed second interferometric signal. Although the second interferometric signal already possesses a high signal-to-noise ratio and a regular shape, it is still fragmented in its data organization. Multiple sets of reflectors each correspond to multiple sets of frequency points, and each set of frequency points contains multiple periodic signal segments. Without organization, subsequent alignment processing would be difficult to perform direct addressing operations by group and frequency point. Through encapsulation and splicing, these fragmented data are grouped together according to group identifiers and frequency point identifiers to form continuous signal segments that can be addressed by group and frequency point.

[0032] For example, suppose there are three sets of reflective mirrors, each set corresponding to four downsampled signals at four frequencies, with each frequency containing multiple cycles of data. The encapsulation step adds group number and frequency number identifiers to the data at each frequency in each set; the splicing step connects multiple cycles of data at the same frequency in the same set sequentially to form a continuous signal segment. Thus, the originally scattered twelve sets of data are organized into twelve continuous signal segments, each uniquely identified by its group number and frequency number.

[0033] Additionally, it's important to note that after stitching, each signal group needs to be normalized. Normalization involves scaling the amplitude of each signal group to the same order of magnitude to eliminate amplitude inconsistencies caused by differences in optical efficiency of different reflectors, gain differences at different frequencies, and power fluctuations of the light source at different periods. Normalization typically uses the maximum amplitude of each signal group as a benchmark, dividing the amplitude of each sampling point by the maximum amplitude of that group, ensuring the normalized signal amplitude falls between 0 and 1. Through normalization, subsequent fitting and alignment steps can more accurately focus on the phase and position information of the signal, rather than being affected by amplitude differences, thereby further improving the processing accuracy and stability.

[0034] Understandably, this step organizes the fragmented data output from the preprocessing into structured continuous signals through encapsulation and splicing. This allows the subsequent alignment step to directly perform independent fitting and inter-group compensation operations on each group of signals without the need for additional data filtering and recombination during the alignment process, thereby improving the coherence and efficiency of the overall processing flow.

[0035] Step S130: Fit and align each third interference signal to obtain multiple sets of fourth interference signals; It should be noted that the fourth interferometric signal is a high-precision signal obtained by fitting and aligning the third interferometric signal. Although the third interferometric signal already has a structured shape and a consistent amplitude range, its peak position remains at the integer level of sampling point precision, and there are coordinate system offsets between multiple signal groups due to mechanical jitter of the rotating mechanism and micro-movements of the patient's fixation. Through the process of fitting and then aligning, the fourth interferometric signal breaks through the integer level limitation of sampling point precision within a group and eliminates the coordinate system offsets between multiple signal groups in terms of inter-group consistency, becoming a final signal form that combines high precision and high consistency.

[0036] For example, taking a signal at a certain frequency corresponding to a set of reflecting mirrors, the fitting process determines the precise position of the reflection peaks of each tissue interface in the signal by continuous fitting, improving the positioning accuracy from the integer level of sampling points to the sub-sampling point level. The alignment process uses the signal with the highest signal-to-noise ratio as a reference, and calculates the relative offset between the other signal groups and the reference signal group to perform translation compensation on the other signal groups. After the above processing, the positions of the reflection peaks of each tissue interface in the fourth interference signal are aligned to the same coordinate system with sub-sampling point accuracy.

[0037] Understandably, this step, through a two-layer process of fitting followed by alignment, addresses two independent technical issues: insufficient intra-group positioning accuracy and inter-group coordinate system offset. This makes the fourth interferometric signal superior to the third interferometric signal in both accuracy and consistency, providing a reliable, high-quality signal foundation for subsequent parameter extraction steps.

[0038] Step S140: Extract the target eyeball biological parameters from each of the fourth interference signals.

[0039] It should be noted that this step is the final step in the entire measurement method, used to extract the final desired ocular biological parameters from the aligned high-precision fourth interferometric signal. Target ocular biological parameters include, but are not limited to, commonly used clinical parameters such as axial length, central corneal thickness, anterior chamber depth, lens thickness, vitreous thickness, and retinal thickness.

[0040] For example, multiple sets of original ocular biological parameters can be extracted from the fourth interference signal, and then filtered and merged by calculating the average value or the minimum variance with the preset reference data, and finally the target ocular biological parameters are obtained by physical unit conversion.

[0041] Understandably, this step, as the endpoint of the measurement process, transforms the signal processing advantages accumulated in the preceding steps into the final measurement results. Thanks to the quality assurance in each stage of preprocessing, encapsulation and splicing, and fitting and alignment in the preceding steps, this step can stably extract various biological parameters from high-quality signals, thereby achieving efficient and integrated measurement of all ocular biological parameters.

[0042] In some embodiments, please refer to Figure 2 , Figure 2 This is a schematic diagram of the overall system architecture provided for Embodiment 1 of the ocular bioparameter measurement method of this application. The overall system architecture is divided into three core processing stages: signal acquisition and preprocessing, UDP (User Datagram Protocol) communication protocol, and host computer parameter calculation.

[0043] In the signal acquisition and preprocessing stage, the interference signal is first acquired through an optical hardware system. Specifically, this involves using three sets of reflecting mirrors rotated in an angular array to acquire the signal, followed by synchronous acquisition using a high-speed ADC (Analog-to-Digital Converter). The acquired raw signal enters the software processing pipeline, where it undergoes a series of processes: DC component removal to eliminate baseline drift, multi-viewpoint orthogonal matched filtering to enhance signal characteristics, median filtering to remove impulse noise interference, downsampling to reduce data throughput and increase processing speed, and finally encapsulation into UDP packets for network transmission.

[0044] The system then enters the UDP communication protocol phase, which is responsible for efficient data transmission and integrity verification. The host computer receives UDP data packets over the network and sequentially performs UDP physical layer reception, frame header parsing to extract the payload, and data packet verification to ensure that no data corruption or packet loss occurs during transmission.

[0045] Finally, the host computer parameter calculation stage is the core of data processing. The received signal first undergoes signal-to-noise ratio normalization to eliminate amplitude differences between different lens groups. Then, signal alignment and coefficient calibration are performed, which involves calculating the offset of each signal group relative to the reference group using a cross-correlation algorithm and performing translation compensation to achieve precise subsampling point-level positioning and alignment. After signal alignment, the system enters the multi-tissue interface precision detection stage. Interfaces such as the anterior and posterior corneal surfaces and the anterior and posterior lens capsules are detected in the front half of the window, while the inner and outer retinal surfaces are detected in the rear half of the window. Subsampling point-level peak positions of each interface are extracted. Next, six parameters are jointly calculated, and based on the peak positions of each interface, the following six raw parameters are obtained: CCT (Central Corneal Thickness), ACD (Anterior Chamber Depth), LT (Lens Thickness), VT (Vitreous Thickness), RT (Retinal Thickness), and AL (Axial Length). Each of the three sets of reflecting lenses performs the above tests independently, obtaining three sets of six-parameter candidate values. Finally, a redundancy and arbitration mechanism is used to calculate the average or maximum value of the three sets of candidate values, outputting the final accurate and reliable ocular parameter measurement results.

[0046] In this embodiment, the present application encapsulates and splices the interference signals collected from each group of reflective lenses, avoiding the second interference signal output after preprocessing from directly entering the subsequent alignment process in a scattered form, thereby improving the standardization of the interference signals; then, by fitting each interference signal and then aligning it, the accuracy of the obtained fourth interference signal is improved, eliminating signal drift and coordinate system offset caused by mechanical jitter of the rotating mechanism and micro-movement of the patient's fixation, thereby improving the accuracy of ocular biological parameter measurement.

[0047] Based on Embodiment 1 of this application, in Embodiment 2 of this application, the content that is the same as or similar to that in Embodiment 1 can be referred to the above description, and will not be repeated hereafter. Based on this, step S130, the method for measuring ocular bioparameters includes the following steps: Step S210: Obtain each third interference signal within a preset threshold range, obtain significant signal points, and determine multiple sets of intermediate peak indices; A preset threshold is a threshold used to filter signal amplitudes, distinguishing effective reflection peaks from noise levels in the signal. In the third interference signal, the reflection peak amplitudes at different tissue interfaces vary; superficial tissues such as the cornea have stronger reflection peaks, while deeper tissues such as the vitreoretinal interface have weaker reflection peaks. By setting a preset threshold, noise signals with amplitudes below the threshold can be filtered out, retaining only candidate signal points with amplitudes exceeding the threshold.

[0048] For example, suppose the preset threshold is set to 10% of the maximum amplitude of the signal. For a set of third interference signals whose amplitude is distributed between 0 and 1, the threshold value is 0.1. Iterate through all the sampling points of this set of signals, and mark the sampling points with amplitudes greater than 0.1 as candidate signal points, and consider the sampling points with amplitudes less than 0.1 as noise and filter them out.

[0049] It should be noted that significant signal points refer to signal points whose amplitude exceeds a preset threshold. These signal points correspond to the approximate location of the reflection peaks at various tissue interfaces, reflecting the range within which the reflection peaks are located, but have not yet reached the precision at the sampling point level. Significant signal points may contain multiple consecutive sampling points, forming one or more signal blocks.

[0050] For example, in a set of third interference signals, sampling points with amplitudes exceeding the 0.1 threshold may be concentrated in several intervals, each interval corresponding to a reflection peak at a tissue interface. For instance, significant signal points corresponding to the anterior corneal surface reflection peak are concentrated between sampling point indices 100 and 120, while significant signal points corresponding to the posterior corneal surface reflection peak are concentrated between sampling point indices 140 and 160. These continuously distributed significant signal points constitute candidate regions for subsequent processing.

[0051] It should be noted that the intermediate peak index refers to the index value, which is further determined within the significant signal points and represents the approximate position of each reflection peak. This index value is on the order of integers of the sampling points and serves as input for subsequent fitting processing. The method for determining the intermediate peak index depends on the distribution pattern of the significant signal points. If there are two peak points with a spacing within a preset range among the significant signal points, then these two peak points are used as a set of intermediate peak indices; if there are no two peak points that meet the conditions, then the peak point of the first consecutive signal block is used as the intermediate peak index.

[0052] For example, if two adjacent peak points are detected within a significant signal point interval corresponding to the reflectance peak on the anterior surface of the cornea, and the interval between them is 15 sampling points, which falls within a preset interval range (10 to 20 sampling points), then these two peak points are respectively used as the intermediate peak indexes of the anterior and posterior surfaces of the tissue interface. In another reflectance peak interval, if only one continuous signal block is detected and no two peak points meeting the criteria are found, then the peak point of this continuous signal block is used as the intermediate peak index of the tissue interface.

[0053] Understandably, this step filters out significant signal points by setting a preset threshold, and then determines the intermediate peak index from the significant signal points, thus achieving the initial positioning from the original signal to the candidate reflection peak position.

[0054] In one feasible implementation, step S210 includes: Step S310: Determine whether the spacing between each significant signal point is within the preset spacing range; The preset spacing range refers to the range of sampling points corresponding to the physical distance between the anterior and posterior surfaces of the cornea. Since the cornea is the most anterior refractive medium in the anterior segment of the eye, its thickness physiologically has a defined range (typically between 450 μm and 550 μm). In signal processing, this physical thickness range corresponds to a specific range of sampling points, calculated using the speed of light and refractive index. Therefore, the preset spacing range is actually a threshold set based on the physiological range of CCT (Central Corneal Thickness), used to verify whether two detected significant signal points conform to the spatial arrangement pattern of the corneal bilayer structure.

[0055] For example, assuming a system sampling rate of 1200Hz, a corneal refractive index of approximately 1.376, and a round-trip time of light in the tissue corresponding to a certain sampling point delay, a CCT thickness of 450μm to 550μm corresponds to 15 to 18 sampling points on the signal waveform. When step S210 detects that the distance between two adjacent significant signal points is 16 sampling points, this distance is within the preset CCT range, and the system determines that these two signal points are the anterior and posterior surfaces of the cornea. Conversely, if the distance between two significant signal points is only 5 sampling points, significantly smaller than the physiological range corresponding to the CCT, then these two signal points are determined not to belong to the corneal interface (they may be noise or reflections from other tissues).

[0056] Understandably, this step pairs and filters significant signal points by using a preset spacing range, correctly grouping two reflection peaks belonging to the front and back surfaces of the same tissue interface into one group, while excluding interfering peaks that do not belong to the same interface. This filtering mechanism improves the accuracy of determining the intermediate peak index.

[0057] Step S320: When the spacing between each significant signal point is within a preset spacing range, the two candidate peak indices within each significant signal point are respectively used as the intermediate peak index. The two candidate peak indices refer to the sampling point positions corresponding to the two local amplitude maxima detected within the same set of significant signal point intervals, each representing the center point of a reflection peak. When the judgment interval falls within the preset range of CCT, the system identifies it as a bimodal mode, meaning that the interval corresponds to the two reflection interfaces of the anterior and posterior corneal surfaces. Among the two candidate peak indices, the one located on the anterior side corresponds to the center point of the reflection peak on the anterior corneal surface, and the one located on the posterior side corresponds to the center point of the reflection peak on the posterior corneal surface; both serve as the intermediate peak indices of their respective reflection peaks.

[0058] Understandably, this step, in bimodal mode, determines the center points of the two reflection peaks as intermediate peak indices, allowing each of the anterior and posterior corneal surfaces to obtain an independent, coarse location. These two indices will serve as the starting point for subsequent fitting processing, ensuring that the precise peak positions of the anterior and posterior surfaces can be calculated separately, thus laying the foundation for the accurate extraction of parameters such as corneal thickness.

[0059] Step S330: If the spacing between each significant signal point is not within the preset spacing range, the first consecutive signal block within each significant signal point is used as the intermediate peak index.

[0060] The first continuous signal block refers to the first interval of sampling points in the set of significant signal points, arranged from smallest to largest by sampling point index, whose amplitude continuously exceeds a preset threshold. When it is determined that the distance between adjacent significant signal points does not fall within the preset range of CCT, the system identifies it as a single-peak mode, meaning that this interval does not constitute a bimodal structure of the anterior and posterior surfaces of the cornea. It may correspond to a single-layer reflection interface such as the posterior surface of the lens or the retina, or a situation where the corneal signal can only be detected as a single peak due to attenuation. In this case, instead of attempting to pair the two peaks, the system takes the earliest continuous signal block in this interval and uses the sampling point index corresponding to its local amplitude maximum point or the center of the block as the intermediate peak index, representing the approximate center point of the reflection peak.

[0061] Understandably, this step uses the first continuous signal block as the output intermediate peak index in single-peak mode, avoiding the positioning errors caused by forcibly pairing two peaks when the corneal double-layer structure does not exist. Symmetrical processing is achieved with the bimodal output (two indices for the anterior and posterior surfaces): the bimodal mode retains corneal thickness information, while the single-peak mode focuses on the center positioning of the single-layer reflective surface. Although the number of intermediate peak indices output by the two modes differs, they are both coarse positions at the integer level of the sampling points, uniformly serving as the starting point for subsequent continuous fitting.

[0062] In this embodiment, by introducing a preset interval determination based on the physiological range of CCT, the significant signal point interval is adaptively divided into bimodal and unimodal modes, avoiding the problem of unimodal being misjudged as bimodal or bimodal being misjudged as unimodal under weak signals, and improving the positioning robustness of the intermediate peak index under different signals.

[0063] Step S220: Fit the signal segments corresponding to each intermediate peak index to obtain multiple sets of precise peak positions; Fitting refers to curve fitting of the signal segment surrounding the intermediate peak index using a Gaussian function. Since the reflection peaks at various tissue interfaces approximately follow a Gaussian distribution in the interference signal waveform, Gaussian fitting can approximate discrete sampling points as a continuous function. Therefore, by finding the extreme points of the fitted curve, the peak positions at the sub-sampling point level can be obtained. The signal segment used for fitting is formed by extending a predetermined number of sampling points to both sides of the intermediate peak index to ensure that the fitting window covers the main part of the reflection peak without introducing excessive sidelobe noise.

[0064] For example, consider a set of intermediate peak indices on the anterior corneal surface, located at sampling point 110 (a coarse integer position). Taking 5 sampling points forward and backward from 110, a signal segment of 11 points (sampling points 105 to 115) is formed. This signal segment is then fitted using a Gaussian function with least-squares to obtain fitting parameters. These fitting parameters represent the precise peak position on the anterior corneal surface, improving accuracy from the integer level to the sub-sampling point level (approximately an offset of 0.37 sampling points). Similarly, the intermediate peak index on the posterior corneal surface (e.g., sampling point 128) is fitted in the same way to obtain fitting parameters, which serve as the precise peak position on the posterior surface.

[0065] It should be noted that the precise peak position refers to the x-coordinate value corresponding to the extreme point of the fitted curve, which is in continuous floating-point form, breaking through the integer-level limitation of the intermediate peak index sampling points. After fitting, each set of intermediate peak indices outputs a precise peak position (one for the anterior and posterior corneal surfaces in bimodal mode, and one for the single-layer interface in unimodal mode), which serves as the coordinate reference for subsequent windowing and accumulation.

[0066] Understandably, this step uses Gaussian fitting to improve the intermediate peak index from the integer level to the precise peak position at the sub-sampling point level, thus solving the positioning accuracy bottleneck caused by the discreteness of sampling points.

[0067] Step S230: Extract the windows corresponding to each precise peak position and accumulate the signals within each window to obtain multiple sets of intra-group accumulated signals.

[0068] For example, taking a signal at a certain frequency corresponding to a certain reflective lens, the precise peak position μ1≈110.37 on the anterior corneal surface and the precise peak position μ2≈127.84 on the posterior corneal surface have been obtained in step S220 (both are sub-sampling-point floating-point numbers), and this frequency point contains a signal segment of 8 scan cycles. When windowing, 7 sampling points are taken on each side of μ1 and μ2 as the center to form a window (window width 15 sampling points); since μ is a floating-point number, the original signal is resampled by linear interpolation before windowing, so that the floating-point center is aligned with the integer index after resampling, and then the ±7 neighborhood is taken. For the signal of 8 cycles at this frequency point, the anterior surface window and the posterior surface window are cut out according to μ1 and μ2 respectively in each cycle, resulting in 8 anterior surface window segments and 8 posterior surface window segments. During accumulation, two strategies can be selected: for interfaces with strong reflections and low periodic jitter, such as the cornea and lens, mean accumulation is used; for deep interfaces with weak reflections, such as the vitreoretinal junction, maximum accumulation is used. The remaining frequencies of the lens are processed in the same way, and the lens ultimately outputs multiple sets of intra-group accumulated signals for subsequent inter-group alignment.

[0069] It should be noted that the mean accumulation uses the Welford (online mean algorithm), which can complete the multi-period mean update with O(1) extra space during a single-period traversal, and its numerical stability is better than the traditional method of "summing first and then dividing by N", making it suitable for conventional accumulation scenarios with a large number of periods (such as 8 periods or more). For example, taking 8 window segments on the front surface as an example, M1=x1 is initialized, and k iterates from 2 to 8. After the iteration is completed, M8 is the mean accumulation result at that position. The mean accumulation signal of the whole window is obtained by executing the algorithm independently at each sampling point position.

[0070] It should be noted that maximum value accumulation refers to taking the maximum value element-wise from multiple periodic window segments at the same sampling point location. That is, the value at each sampling point location of the output window is the maximum value of the eight periodic segments at that location. This strategy can preserve the morphological characteristics of weak reflection peaks and suppress peak collapse caused by single-period jitter, making it particularly suitable for deep weak reflection interfaces such as the vitreous-retinal interface. For example, taking eight window segments of the posterior surface (or retinal interface) as an example, for each sampling point location j (j=1~15), the maximum value of the output accumulated signal is taken, and this process is repeated point-by-point to obtain the maximum value accumulated signal for the entire window.

[0071] Understandably, this step achieves intra-group accumulation by truncating the window at the precise peak position at the sub-sampling point level and accumulating it over multiple cycles. This not only utilizes the high positioning accuracy of the precise peak position to ensure that the center of the truncated window is aligned with the true reflection peak, but also suppresses random noise in the single-cycle signal through multi-cycle accumulation, making the signal-to-noise ratio of the intra-group accumulated signal significantly higher than that of the single-cycle signal.

[0072] This step ultimately yields multiple sets of intra-group cumulative signals, as referenced. Figure 4 , Figure 4 This is a schematic diagram of the intragroup signal alignment and accumulation process of the ocular bioparameter measurement method provided in Embodiment 2 of this application.

[0073] The intra-group signal alignment and accumulation process first receives the normalized signal of the current frame as input. To capture the main energy concentration area of ​​the signal, the system first performs threshold detection within the first 33% of the signal sequence, searching and filtering out valid signal points with amplitudes greater than the high threshold. Subsequently, the system enters the core judgment node, which determines whether the detected double-peak spacing is within the CCT range. Depending on the judgment result, the system intelligently switches to different processing branches: if the judgment result is yes, it enters double-peak mode. In this mode, the system locates the center positions of the two distinct peaks, using them as the reference for signal alignment. If the judgment result is no, it enters single-peak mode. In this mode, the system locates the center position of the first continuous signal block, using it as the reference for signal alignment.

[0074] After establishing the alignment reference, regardless of the mode, the system performs a Gaussian precise fitting operation to calculate the sub-sampling point level *mu*. Using a Gaussian fitting algorithm, a center offset *mu* with sub-sampling point precision is calculated. Next, the system uses the calculated precise offset as an anchor point to truncate a 68% frame length window, ensuring that each truncated local signal segment accurately covers the target physiological characteristics. To eliminate spectral leakage and artifacts caused by boundary truncation, the truncated signal undergoes zero-padding boundary processing.

[0075] Finally, the processed signal segments enter the core accumulation stage. The system accelerates the accumulation into this set of buffers using SIMD (Single Instruction Multiple Data) technology, significantly improving accumulation efficiency. After processing all frames, the accumulated signal output is the final alignment and accumulation result, serving as the basis for subsequent multi-frequency joint calculations.

[0076] In one possible implementation, after step S230, the method further includes: Step S410: Based on the signal-to-noise ratio of the accumulated signals in each group, determine the reference group signal and multiple non-reference group signals, wherein the signal-to-noise ratio of the reference group signal is higher than that of each non-reference group signal. It should be noted that the reference group signal refers to the group with the highest signal-to-noise ratio among the cumulative signals output by each of the multiple groups of reflecting mirrors. This group is selected as the reference coordinate system for inter-group alignment, and subsequent cross-correlation and compensation are all based on the reference group as the alignment target, without further translation of the reference group itself. Non-reference group signals refer to the cumulative signals output by the other reflecting mirrors besides the reference group. These signals have coordinate system offsets from the reference group due to mechanical jitter, fixation micro-motion, or assembly discrepancies. These offsets need to be calculated and compensated in subsequent steps through cross-correlation to align with the reference group coordinate system.

[0077] For example, suppose three sets of reflective mirrors (mirror A, lens B, and lens C) each output an accumulated signal within their respective front surface groups. The calculated signal-to-noise ratios (SNRs) are as follows: lens A: 32 dB (Decibel), lens B: 27 dB, and lens C: 24 dB. Lens A has the highest SNR among the three groups, therefore its accumulated signal is determined as the reference group signal. The accumulated signals of lenses B and C have lower SNRs than the reference group and are therefore determined as non-reference group signals. The reference group (lens A) remains stationary during subsequent inter-group alignment processes, serving as a reference coordinate system. The non-reference groups (lens B and lens C) need to be cross-correlated with the reference group, and the resulting offset is used to compensate for their own deviations.

[0078] Understandably, this step selects a benchmark group based on signal-to-noise ratio (SNR) to align the groups using the signal with the best quality as the reference coordinate system. This avoids introducing additional alignment errors when using a low SNR signal as the benchmark, and provides a reliable reference benchmark for subsequent cross-correlation calculations.

[0079] Step S420: Perform cross-correlation calculations on the reference group signal and each non-reference group signal to obtain multiple sets of offsets; It should be noted that cross-correlation calculation is used to measure the change in similarity between two sets of signals during translation. The result is a one-dimensional correlation curve, and the translation amount corresponding to the peak position of the curve is the relative offset between the two sets of signals. In specific calculation, the reference group signal is fixed, and the non-reference group signal is translated point by point within a preset translation range (such as ±20 sampling points). After each translation, the cross-correlation coefficient between the two sets of signals is calculated, and the translation amount corresponding to the maximum correlation coefficient is the offset of the non-reference group relative to the reference group.

[0080] For example, taking the reference group (cumulative signal from the front surface of lens A, with a length of 15 sampling points) and the non-reference group (cumulative signal from the front surface of lens B) as examples, the lens B signal is shifted point by point within the range of [-5, +5] sampling points. After each shift of Δ, the cross-correlation coefficient with the lens A signal is calculated. When Δ=+2, the correlation coefficient is the largest (0.97), so the offset of lens B relative to lens A is +2 sampling points; similarly, the offset calculated by cross-correlation between lens C and lens A is -1 sampling point. The two sets of offsets constitute multiple sets of offsets.

[0081] It should be noted that the offset refers to the translational difference of the non-reference group signal relative to the reference group signal on the sampling point axis. Its physical sources include random errors from mechanical jitter of the rotating mechanism, beam incident point offset caused by micro-movements of the patient's fixation, and initial position differences during the assembly and adjustment of multiple lens groups. The offset is usually an integer or a sub-sampling point level floating-point number (sub-sampling accuracy can be obtained by quadratic fitting of the cross-correlation peaks).

[0082] Understandably, this step quantifies the coordinate system offset of each group of signals relative to the reference group through cross-correlation calculation, transforming the inconsistency between groups caused by "mechanical jitter + fixed micro-motion + assembly adjustment difference" into a calculable offset, providing direct correction parameters for subsequent compensation steps.

[0083] Specifically, the cross-correlation formula is: This formula describes the calculation logic of the cross-correlation function. In the formula, This represents the amplitude of the signal to be aligned at position x after a delay of τ. This represents the amplitude of the reference signal at position x; and These represent the mean (DC component) of the signal to be aligned and the reference signal, respectively. Additionally, the system employs a dual vertical axis design to simultaneously display the original signal characteristics and the algorithm's results: The left vertical axis (Y-axis) is labeled with amplitude, ranging from -1.0 to 1.0. This axis corresponds to the blue curve. and orange curve The figures represent the original fluctuation amplitudes of the reference signal and the signal to be aligned in the optical path domain, respectively. The right vertical axis (Y-axis) is labeled R, with values ​​ranging from -40 to 40. This axis corresponds to the green curve and is used to represent the calculated results of the cross-correlation function. The horizontal axis (X-axis) in the figure is labeled x, representing the sampling point position in the optical path domain.

[0084] The specific meanings and functions of each color curve and key annotation are as follows: Blue curve : Represents the reference signal, i.e., the original signal used as the alignment reference group. Orange curve : Represents the signal to be aligned, i.e., the target signal that has a time offset and needs to be aligned with the reference signal. Purple vertical lines: Located on the left and right sides of the chart, marked respectively. and These two lines define the boundary range of the effective offset search of the system, limiting the algorithm to scan within the interval of the maximum positive and negative sampling points to avoid invalid calculations and edge effects. The green curve represents the changing trend of the cross-correlation function R(τ). Here, the horizontal axis τ represents the delay of the signal on the time axis (in sampling points), indicating the displacement of the signal to be aligned relative to the reference signal; the vertical axis R represents the normalized cross-correlation value, reflecting the similarity between the two signals at different delays. When the two signals are perfectly aligned, the cross-correlation function reaches its global maximum. By subtracting the mean of each signal from the two signals and then summing their products, the algorithm can eliminate the DC offset effect of the signals and accurately calculate the similarity between the two signals at a specific delay τ. Red dots : Marks the highest point of the green cross-correlation curve, i.e., the location of the global maximum. The corresponding x-coordinate value of this point (as marked in the figure) =30) is the optimal offset obtained by the system. The red dashed line indicates that the peak value is finely fitted at the sub-pixel level using parabolic interpolation, thereby breaking through the limitation of discrete sampling points and obtaining a more accurate sub-sampling point level alignment result than a single sampling point, significantly improving the accuracy of timing synchronization.

[0085] By comparing the waveforms of the blue reference signal and the orange signal to be aligned, the offset relationship between the two in the optical path domain is visually demonstrated. The purple vertical line limits the effective range of the offset searched by the algorithm. The green cross-correlation function curve quantifies the similarity between the two signals under different delays, and the red dot corresponding to its global maximum value is the optimal offset. The red dashed line refines the peak value at the sub-sampling point level through parabolic interpolation, breaking through the limitation of discrete sampling points. Combining the amplitude ordinate on the left and the correlation coefficient ordinate on the right, as well as the cross-correlation formula, the system realizes a complete alignment process from coarse-grained integer offset detection to fine-grained sub-sampling point precise positioning, significantly improving signal alignment accuracy.

[0086] Step S430: Compensate each non-reference group signal based on each offset to obtain multiple sets of compensated non-reference group signals.

[0087] For example, if the offset of lens B is +2 sampling points, the cumulative signal within the group of lens B is shifted to the left by 2 sampling points (since a positive offset indicates that the signal of lens B has shifted to the right relative to the reference, it needs to be shifted back to the reference coordinate system during compensation); if the offset of lens C is -1 sampling point, the signal of lens C is shifted to the right by 1 sampling point. If the offset is a sub-sampling floating-point number (such as +2.37) during the shift, the original signal is first resampled to the sub-sampling grid using linear interpolation before shifting to ensure compensation accuracy. After compensation, the cumulative signals of the front surfaces of lens A (reference), lens B (after compensation), and lens C (after compensation) are aligned on the sampling point axis, which is the front surface signal of that frequency point in the fourth interference signal. The back surface and multiple frequency points are processed in the same way, finally obtaining multiple sets of fourth interference signals.

[0088] Understandably, this step compensates for the non-reference group signals by shifting them by an offset, thereby eliminating the coordinate system offset between multiple groups of signals and normalizing the signals output by multiple reflective mirrors to the same reference coordinate system on the sampling point axis.

[0089] This step achieves inter-group alignment of the accumulated signals within a group, referencing... Figure 5 , Figure 5 This is a schematic diagram of the inter-group offset alignment process for the ocular bioparameter measurement method provided in Embodiment 2 of this application. The inter-group offset alignment process aims to eliminate systematic time drift or phase deviation between three independently acquired cumulative signals. First, the system inputs three sets of cumulative signals. To ensure the reliability of subsequent alignment, the process enters the stage of evaluating the signal quality of each set. By scoring the signal-to-noise ratio, signal amplitude, or stability, the set with the best data quality is selected as the benchmark for the entire system. Subsequently, the system selects the set with the best signal quality as the benchmark set, and the other two sets are used as non-benchmark sets to be aligned. Next, the system calculates the peak offset of the correlation function between the non-benchmark set and the benchmark set, and uses a cross-correlation algorithm to accurately quantify the relative displacement of the two sets of signals on the time axis. Finally, based on the calculated offset, the system applies offset compensation to the non-benchmark set, adjusting the three sets of signals to the same time reference, and finally outputs the three aligned cumulative signals, providing spatiotemporal consistency for subsequent joint calculation of six parameters.

[0090] In this embodiment, by selecting a reference group based on signal-to-noise ratio, calculating the offset through cross-correlation, and performing translation compensation, the coordinate system offset between multiple groups of reflecting mirrors caused by mechanical jitter, fixed micro-motion, and assembly discrepancies is corrected one by one, so that multiple groups of signals are normalized to the same reference coordinate system, thus achieving high-precision inter-group alignment of multiple groups of interference signals.

[0091] Based on any of the above embodiments, Embodiment 3 of this application proposes a method for measuring ocular bioparameters. Step S110, the step of preprocessing each first interference signal, includes: Step S510: Remove the DC component of each first interference signal to obtain multiple sets of zero-mean interference signals; It should be noted that the DC component refers to the constant offset component in the first interference signal introduced by the detector's dark current and the average optical power of the light source. This offset manifests as an overall rise or fall in the signal waveform above zero level. This component does not carry any reflection information from tissue interfaces; if not removed, it will generate spurious low-frequency components during subsequent demodulation, interfering with the separation and extraction of modulation signals from various tissue interfaces. After removing the DC component, the signal fluctuates around zero level, forming a zero-mean interference signal.

[0092] For example, suppose the first interference signal acquired by a set of reflecting mirrors has a length of 1024 sampling points and an average amplitude of +0.25V. Subtracting this average of 0.25V from the amplitude value of each sampling point results in a new signal that fluctuates around 0V, i.e., a zero-mean interference signal.

[0093] The formula for removing the DC component is: ,in Let be the amplitude value of the original interference signal at the i-th sampling point, where i is the sampling point index, and its value range is 1≤i≤N; N is the total number of sampling points of the signal in this frame. This is the amplitude value of the i-th sampling point after removing the DC component, i.e., the zero-mean AC signal.

[0094] Understandably, this step eliminates constant offset interference in the signal by removing the DC component, enabling subsequent demodulation processing to accurately extract the modulation components of each tissue interface based on zero level, thus avoiding the contamination of frequency domain analysis by DC residue.

[0095] Step S520: Demodulate each zero-mean interference signal to obtain multiple sets of first filter matrices; It should be noted that demodulation refers to the process of separating and extracting the modulation components corresponding to each tissue interface from a zero-mean interference signal. Since the light signals reflected from each tissue interface have different modulation frequencies in the frequency domain, multiple modulation components aliased in a single interference signal are separated one by one through multi-frequency orthogonal matched filtering. Each frequency point corresponds to a set of envelope signals. The first filtering matrix represents the multiple sets of envelope signals output after demodulation; its row number corresponds to the number of frequency points, and its column number corresponds to the number of sampling points.

[0096] For example, assume the system uses four modulation frequencies (f1, f2, f3, f4), and the zero-mean interference signal has a length of 1024 sampling points. After multi-frequency orthogonal matched filtering, the output is a 4-row × 1024-column matrix, where the first row corresponds to the envelope signal of frequency f1, the second row corresponds to the envelope signal of frequency f2, and so on.

[0097] It should be noted that the first filtering matrix refers to the multiple envelope signals output after demodulation. The number of rows corresponds to the number of modulation frequency points, and the number of columns corresponds to the number of sampling points. Each row represents the envelope waveform at a frequency point, including the reflection peak morphology of each tissue interface. Although the first filtering matrix has separated the envelopes of each frequency point, isolated anomalies caused by detector defects, transient electrical pulses, or light source flicker may still remain.

[0098] For example, assume the system uses four modulation frequencies (f1, f2, f3, f4), and the zero-mean interference signal has a length of 1024 sampling points. After multi-frequency orthogonal matched filtering, a 4-row × 1024-column matrix is ​​output: the first row corresponds to the envelope signal of frequency f1, the second row to f2, the third row to f3, and the fourth row to f4. This 4 × 1024 matrix is ​​the first filtering matrix, and each row is an envelope waveform, with the bulges in the waveform corresponding to the reflection peaks at the interfaces of tissues such as the cornea, lens, and retina.

[0099] The orthogonal demodulation formula is: , , m is the frequency index, ranging from 1 to M, where M is the total number of preset frequency points (typically 8 to 16). Sampling rate; is the carrier frequency of the m-th frequency point, in Hz; i is the sampling point index, ranging from 1 to N (N is the frame length); I(m) is the in-phase component of the m-th frequency point; Q(m) is the quadrature component of the m-th frequency point; A(m) is the amplitude envelope of the m-th frequency point.

[0100] Understandably, this step separates the modulation components of each tissue interface in a single interference signal into independent envelope signals through multi-frequency demodulation, so that the information of different reflection interfaces is clearly presented in matrix form, providing a structured processing object for subsequent outlier filtering and downsampling.

[0101] In some embodiments, refer to Figure 3 , Figure 3 This diagram illustrates the multi-frequency orthogonal matched filter structure for the ocular bioparameter measurement method provided in Embodiment 3 of this application. The input to the multi-frequency orthogonal matched filter structure is the AC signal after DC removal. This signal is first fed in parallel into multiple DLA (Digital Lock-in Amplifier) ​​comb filter modules for processing. The diagram exemplarily shows processing branches for different frequency points, specifically including DLA comb point frequency f1, DLA comb point frequency f2, DLA comb point frequency f3 up to DLA comb point frequency fn, to achieve demodulation of multiple frequency component signals in a broadband light source.

[0102] Subsequently, the signals output from each frequency filter enter the corresponding amplitude envelope extraction stage. Specifically, the f1 frequency signal is processed to generate amplitude envelope A1, the f2 frequency signal to generate amplitude envelope A2, the f3 frequency signal to generate amplitude envelope A3, and the fn frequency signal to generate amplitude envelope An.

[0103] After extracting the amplitude envelopes of each frequency point in parallel, the system summarizes these results to construct a multi-frequency amplitude envelope matrix. Finally, to ensure signal quality and improve the stability of subsequent parameter calculations, the data in this matrix passes through a median filtering output module. The median filtering algorithm removes outliers or noise interference, resulting in a smooth and reliable filtered result, providing high-quality data support for subsequent accurate detection of multiple tissue interfaces.

[0104] Step S530: Filter out isolated outliers in each of the first filter matrices to obtain multiple sets of second filter matrices; It should be noted that isolated outliers refer to single-point noise in the first filtering matrix caused by detector defects, transient electrical pulse interference, or light source flicker, resulting in abrupt amplitude changes. These outliers manifest as one or more sampling points with amplitudes significantly higher than their neighbors within a continuous, smooth envelope signal. If not filtered out, these outliers will be misidentified as false reflection peaks in subsequent peak searches or amplified during downsampling due to sample hold. Filtering typically employs median filtering, replacing the original value with the median of the current point and its neighboring window, thereby suppressing isolated outliers without affecting the overall shape of the envelope signal.

[0105] For example, in the first filtering matrix, the amplitude of frequency point f1 at the 512th sampling point is 3.82, while the amplitudes of the five sampling points before and after it are all between 0.12 and 0.21, making this point a clear isolated outlier. A 3-point median filter (window [511, 512, 513]) is applied, resulting in a median of 0.17 after sorting. The amplitude at this point is then replaced with 0.17, while the amplitudes of the other sampling points remain unchanged. After performing median filtering independently on each of the four rows of frequency points, the second filtering matrix is ​​obtained.

[0106] It should be noted that the second filtering matrix refers to the envelope matrix output after performing median filtering on each row of the first filtering matrix independently. Its number of rows and columns is the same as that of the first filtering matrix (the number of rows is the number of frequency points, and the number of columns is the number of sampling points), but isolated outliers have been suppressed, and the main shape and edge sharpness of the envelope signal are preserved.

[0107] For example, after applying a 3-point median filter to each row of the first filter matrix (4 rows × 1024 columns), a second filter matrix (also 4 rows × 1024 columns) is obtained. In the second filter matrix, the amplitude at the 512th sampling point in the f1 frequency row has decreased from 3.82 to 0.17, while the amplitudes at the remaining sampling points remain unchanged. The reflection peak morphology of the interfaces of tissues such as the cornea, lens, and retina in the envelope signal is unaffected.

[0108] Understandably, this step suppresses isolated outliers in the envelope signal through median filtering, avoiding the generation of false reflection peaks, while preserving the main shape and edge sharpness of the envelope signal, providing a clean signal foundation for subsequent downsampling and peak localization.

[0109] Step S540: Downsample each of the second filter matrices to obtain each of the second interference signals.

[0110] It should be noted that downsampling refers to reducing the sampling rate of the second filtering matrix to reduce the amount of data and adapt to the bandwidth requirements of subsequent encapsulation and transmission. Downsampling requires first performing a low-pass filter on the signal to prevent spectral aliasing, and then extracting sampling points according to a preset downsampling factor. The second interference signal is the downsampled envelope matrix; its number of rows remains unchanged (still the number of frequency points), while the number of columns is reduced according to the downsampling factor. This significantly reduces the amount of data while preserving the main morphological characteristics of the reflection peaks at each tissue interface.

[0111] For example, the second filtering matrix is ​​4 rows × 1024 columns, and the downsampling factor is set to 4. First, each row of the signal is low-pass filtered with a cutoff frequency of 1 / 8 of the original sampling rate. Then, one sample is extracted every three sampling points to obtain a 4-row × 256-column downsampling matrix, which is the second interference signal. The data volume is compressed from 4096 points to 1024 points, a reduction of 75%.

[0112] Understandably, this step significantly reduces the amount of data by downsampling while preserving the main shape of the reflection peak, enabling subsequent encapsulation, splicing, and alignment processes to run with lower data bandwidth, thus improving the efficiency of the overall processing flow and avoiding transmission bottlenecks caused by excessive data volume.

[0113] In this embodiment, removing the DC component eliminates constant offset interference in the signal; multi-frequency demodulation separates the modulation components of each aliased tissue interface into independent envelope signals; median filtering suppresses isolated outliers in the envelope signals; and downsampling significantly compresses the data volume while preserving the main shape of the reflection peaks. The four-step preprocessing transforms the original noisy, aliased first interference signal into a second interference signal with improved signal-to-noise ratio, compressed data volume, and regularized shape, providing a high-quality input foundation for subsequent encapsulation, splicing, and fitting alignment processes.

[0114] Based on any of the above embodiments, Embodiment 4 of this application proposes a method for measuring ocular biological parameters. Step S140, the step of extracting target ocular biological parameters from each fourth interference signal, includes: Step S610: Extract multiple sets of original eyeball biological parameters from the fourth interference signal; For example, taking the fourth interference signal of a set of reflecting lenses as an example, two adjacent signal clusters are detected at the beginning of the anterior half window. The peak positions of the subsampling points on the anterior and posterior corneal surfaces are obtained by Gaussian fitting, and the difference between the two is converted into CCT. After the peak on the posterior corneal surface, the first signal rising edge above the low threshold is searched to locate the anterior lens capsule, and the difference between it and the posterior corneal surface is converted into ACD. In the 20% to 50% region of the anterior half window, the first signal rising edge above the high threshold is searched to locate the posterior lens capsule, and the difference between it and the anterior capsule is converted into LT. Two signal clusters above the high threshold are detected in the posterior half window, corresponding to the anterior retinal surface (internal limiting membrane) and the posterior retinal surface (pigmented epithelium), respectively, and the difference between them is converted into RT. The total difference from the anterior corneal surface to the posterior retinal surface is converted into AL. After this set of lenses detects 6 parameters, the other two sets of lenses independently perform the same detection, for a total of 18 sets of raw ocular biological parameters.

[0115] Understandably, this step relies on the aligned subsampling point-level peaks in the fourth interference signal, combined with anatomical prior partitioning and differential threshold strategies, to obtain 18 sets of original candidate values ​​based on the independent detection of the three sets of lenses, providing a statistical basis for subsequent screening and merging.

[0116] This step yielded multiple sets of raw ocular biological parameters, which were then referenced. Figure 6 , Figure 7 , Figure 6 This is a schematic diagram of the parameter detection process for the first half of the ocular bioparameter measurement method provided in Embodiment 4 of this application. Figure 7 This is a schematic diagram of the retinal parameter detection process in the posterior region of the ocular bioparameter measurement method provided in Embodiment 4 of this application.

[0117] The system divides the accumulated signal into two halves for targeted biological parameter calculation: parameter detection in the first half, referencing... Figure 6 First, the system processes the first half of the accumulated signal region (0% to 50%). It searches for bimodal peaks within the range of 0% to the margin plus CCT Max, and performs Gaussian fitting on each of the detected peaks. The fitting results determine the positions of the anterior and posterior corneal surfaces, and the difference between them yields the preliminary CCT. The CCT calculation formula is as follows: ,in Location of the anterior surface of the cornea. The location is the posterior surface of the cornea. Then, starting from this location, the first rising edge above the threshold is searched in subsequent signals to locate the anterior capsule of the lens, thus determining the ACD (Aspect Ratio). The ACD calculation formula is as follows: ,in The location of the anterior lens capsule is determined. Next, the first rising edge above the threshold is searched again in the 20% to 50% signal region to determine the location of the posterior lens capsule, thereby calculating the LT (Long Threshold). The formula for LT calculation is: ,in This is the location of the posterior capsule of the lens.

[0118] Retinal parameter detection in the posterior half region, refer to Figure 7 Simultaneously, processing is performed on the latter half of the accumulated signal region (50% to 100%). The system searches within signal clusters above a high threshold to precisely locate the anterior and posterior surfaces of the retina. The positional difference between these two locations is used to calculate the RT (Retardation Time), calculated using the following formula: ,in Location on the posterior surface of the retina. This represents the location of the anterior retinal surface. Finally, combining the locations of the posterior corneal surface obtained from the anterior half of the region and the posterior retinal surface obtained from the posterior half, the crucial axial length (AL) is calculated. The formula for AL is: The vitreous cavity length VT is calculated by combining the location of the posterior capsule of the lens. The formula for calculating VT is: .

[0119] Step S620: Calculate the minimum variance or the average value of each target eye biological parameter compared to the preset data; For example, taking CCT parameters as an example, step S610 has extracted three sets of original CCT candidate values ​​from three sets of lenses (e.g., CCT_A=532μm, CCT_B=528μm, CCT_C=535μm). If the minimum variance strategy is adopted, the variance of these three sets of candidate values ​​is calculated with the preset reference data (e.g., the CCT value of the standard eye mold is 530μm), and the candidate value with the smallest variance is selected; if the average value strategy is adopted, the arithmetic mean of the three sets of candidate values ​​is directly calculated. One of the two strategies can be used. The former is suitable for scenarios with clear standard reference values ​​(e.g., factory calibration), while the latter is suitable for routine measurement scenarios without reference values ​​but with multiple sets of redundant data.

[0120] Understandably, this step uses either the minimum variance or the average value strategy to select or synthesize a representative value from multiple sets of original candidate values, providing statistically processed input for subsequent merging steps and avoiding bias caused by local interference in a single set of candidate values.

[0121] Step S630: Combine the original ocular biological parameters corresponding to each mean or each minimum variance to obtain the intermediate ocular biological parameters; For example, if the average value strategy is used, the average values ​​of the six parameters CCT, ACD, LT, VT, RT, and AL are combined into a six-tuple (CCT≈531.7μm, ACD≈3.82mm, LT≈4.26mm, VT≈9.61mm, RT≈0.23mm, AL≈19.41mm), and this six-tuple is the intermediate ocular biological parameter. If the minimum variance strategy is used, the candidate values ​​with the smallest variance for each of the six parameters are combined into a six-tuple.

[0122] Understandably, this step merges the results of the six parameters after each filtering into a complete six-tuple, ensuring that all parameters originate from the same measurement process and the same set of alignment signals, thus guaranteeing the inherent consistency between the parameters and providing a structured input for subsequent physical unit conversion.

[0123] Step S640: Perform data conversion on the intermediate eyeball bioparameters to obtain the target eyeball bioparameters.

[0124] For example, the values ​​in the intermediate ocular bioparameters are currently optical path length values ​​(in mm), which need to be converted to clinically common physical length values. During conversion, each optical path length value is divided by the refractive index of the corresponding tissue interface (e.g., corneal refractive index 1.376, aqueous humor refractive index 1.336, lens refractive index 1.406, vitreous refractive index 1.337) to obtain the physical length value in μm. For example, the intermediate CCT optical path length value of 531.7 μm divided by the corneal refractive index 1.376 yields a physical CCT of approximately 386.4 μm. After the six parameters are converted sequentially, the final target ocular bioparameter six-tuple is output.

[0125] Understandably, this step, as the final step in the parameter extraction process, converts the intermediate parameters in the optical path domain into clinically usable physical length values, enabling the measurement results to be directly used in clinical applications such as preoperative cataract assessment and myopia control, thus completing the complete measurement link from the original interference signal to clinical parameters.

[0126] In this embodiment, by extracting 18 sets of original ocular biological parameters independently detected by three lenses from the fourth interference signal, and then filtering and merging them into intermediate parameters through the average value or minimum variance strategy, and finally converting them into physical length values ​​through refractive index conversion, an efficient process of extracting all six parameters at once from a single cumulative interference signal is realized.

[0127] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the ocular biological parameter measurement method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0128] This application also provides an ocular bioparameter measurement system; please refer to... Figure 9 The ocular bioparameter measurement system includes: The preprocessing module 10 is used to preprocess each first interference signal in response to the acquisition of multiple sets of first interference signals corresponding to the reflective mirrors to obtain multiple sets of second interference signals; The encapsulation module 20 is used to encapsulate and splice each of the second interference signals to obtain multiple sets of third interference signals; Alignment module 30 is used to fit and then align each third interference signal to obtain multiple sets of fourth interference signals; Extraction module 40 is used to extract target eye biological parameters from each fourth interference signal.

[0129] The ocular bioparameter measurement system provided in this application, employing the ocular bioparameter measurement method described in the above embodiments, can solve the technical problem of insufficient accuracy in ocular bioparameter measurement. Compared with the prior art, the beneficial effects of the ocular bioparameter measurement system provided in this application are the same as those of the ocular bioparameter measurement method provided in the above embodiments, and other technical features of the ocular bioparameter measurement system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0130] This application provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the ocular bioparameter measurement method of the above embodiment 1.

[0131] The following is for reference. Figure 10 The diagram illustrates a structural schematic of an electronic device suitable for implementing embodiments of this application. The electronic devices in these embodiments may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 10The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0132] like Figure 10 As shown, the electronic device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the electronic device. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. The communication device 1009 allows the electronic device to communicate wirelessly or wiredly with other devices to exchange data. Although the diagrams show electronic devices with various systems, it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems may be implemented alternatively.

[0133] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0134] The electronic device provided in this application, employing the ocular bioparameter measurement method described in the above embodiments, can solve the technical problem of insufficient accuracy in ocular bioparameter measurement. Compared with the prior art, the beneficial effects of the electronic device provided in this application are the same as those of the ocular bioparameter measurement method provided in the above embodiments, and other technical features of this electronic device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0135] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0136] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0137] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to perform the ocular bioparameter measurement method in the above embodiments.

[0138] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0139] The aforementioned computer-readable storage medium may be included in an electronic device or may exist independently without being assembled into an electronic device.

[0140] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by an electronic device, the electronic device: in response to acquiring first interference signals corresponding to multiple sets of reflecting lenses respectively, preprocesses each first interference signal to obtain multiple sets of second interference signals; encapsulates and splices each second interference signal to obtain multiple sets of third interference signals; fits and aligns each third interference signal to obtain multiple sets of fourth interference signals; and extracts target eyeball biological parameters from each fourth interference signal.

[0141] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0142] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0143] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0144] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described method for measuring ocular bioparameters, thereby solving the technical problem of insufficient accuracy in ocular bioparameter measurement. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the ocular bioparameter measurement method provided in the above embodiments, and will not be repeated here.

[0145] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for measuring ocular bioparameters, characterized in that, The methods for measuring ocular bioparameters include: In response to the acquisition of first interference signals corresponding to multiple sets of reflective mirrors, each of the first interference signals is preprocessed to obtain multiple sets of second interference signals; Each of the second interference signals is encapsulated and spliced ​​together to obtain multiple sets of third interference signals; The third interference signals are first fitted and then aligned to obtain multiple sets of fourth interference signals; Biological parameters of the target eyeball are extracted from each of the fourth interference signals.

2. The method for measuring ocular bioparameters as described in claim 1, characterized in that, The step of preprocessing each of the first interference signals includes: By removing the DC component of each of the first interference signals, multiple sets of zero-mean interference signals are obtained; Demodulate each of the zero-mean interference signals to obtain multiple sets of first filter matrices; By filtering out isolated outliers in each of the first filter matrices, multiple sets of second filter matrices are obtained. Each of the second filter matrices is downsampled to obtain the second interference signal.

3. The method for measuring ocular bioparameters as described in claim 1, characterized in that, The step of fitting and aligning each of the third interference signals to obtain multiple sets of fourth interference signals includes: Acquire each of the third interference signals within a preset threshold range, obtain significant signal points, and determine multiple sets of intermediate peak indices; By fitting the signal segments corresponding to each of the intermediate peak indices, multiple sets of precise peak positions are obtained; Each precise peak position is captured into a window, and the signal within each window is accumulated to obtain multiple sets of intra-group accumulated signals.

4. The method for measuring ocular bioparameters as described in claim 3, characterized in that, The step of determining multiple sets of intermediate peak indices includes: Determine whether the spacing between the significant signal points is within a preset spacing range; When the spacing between the significant signal points is within a preset spacing range, the two candidate peak indices within each significant signal point are respectively used as the intermediate peak index; If the spacing between the significant signal points is not within a preset spacing range, the first consecutive signal block within each significant signal point is used as the intermediate peak index.

5. The method for measuring ocular bioparameters as described in claim 3, characterized in that, The intra-group accumulated signal includes at least a reference group signal and multiple compensated non-reference group signals; after the step of extracting windows corresponding to each of the precise peak positions and accumulating the signals within each window to obtain multiple intra-group accumulated signals, the method further includes: Based on the signal-to-noise ratio of the accumulated signals within each group, a reference group signal and multiple non-reference group signals are determined, wherein the signal-to-noise ratio of the reference group signal is higher than that of each of the non-reference group signals; The reference group signal and each of the non-reference group signals are cross-correlated to obtain multiple sets of offsets. Each of the non-reference group signals is compensated based on the offset to obtain multiple sets of compensated non-reference group signals.

6. The method for measuring ocular bioparameters as described in claim 1, characterized in that, The step of extracting target eye biological parameters from each of the fourth interference signals includes: Multiple sets of original ocular biological parameters were extracted from the fourth interference signal; Calculate the minimum variance or the average value of each of the target eye biological parameters compared with the preset data; By combining the original ocular biological parameters corresponding to the average value or the minimum variance of each of the above, intermediate ocular biological parameters are obtained; The intermediate ocular bioparameters are converted to obtain the target ocular bioparameters.

7. A system for measuring ocular bioparameters, characterized in that, The ocular bioparameter measurement system includes: The preprocessing module is used to preprocess each of the first interference signals corresponding to the multiple sets of reflective mirrors to obtain multiple sets of second interference signals in response to the acquisition of the first interference signals. The encapsulation module is used to encapsulate and splice each of the second interference signals to obtain multiple sets of third interference signals; The alignment module is used to first fit and then align each of the third interference signals to obtain multiple sets of fourth interference signals; An extraction module is used to extract target eye biological parameters from each of the fourth interference signals.

8. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the ocular bioparameter measurement method as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the ocular bioparameter measurement method as described in any one of claims 1 to 6.

Images