Construction method and system for reproductive medicine database

Abstract

The invention relates to the technical field of database construction, in particular to a construction method and system for a reproductive medicine database. The method comprises the following steps: acquiring a biological tissue sample of a patient; acquiring reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data from the biological tissue sample of the patient; performing time sequence analysis processing on the hormone fluctuation mode according to the reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; performing image recognition processing based on coherent superposition characteristics of light on the embryo morphological parameter data, and performing embryo development potential evaluation to obtain embryo development potential data; and performing hierarchical clustering processing on the individual heterogeneity of the patient according to the immune cell function data and the genome variation data to obtain immune-genetic typing data of the patient. By means of the quantum dot sensing technology and the multi-channel micro-fluidic chip, standardization of data acquisition is achieved, and the heterogeneity problem of a data source is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of database construction, and in particular to a method and system for constructing a reproductive medicine database. Background Art

[0002] The reproductive medicine database is an important part of modern medical information technology. It can effectively manage and analyze clinical and research data related to reproductive health, including endocrine test data, embryo morphology and development data, immune function data, and genomics data. When constructing a reproductive medicine database, the users of the reproductive medicine database usually include clinicians, researchers, managers, and patients. Each type of user has different requirements for the database. For example, doctors need to quickly retrieve patient medical records, researchers need to analyze large amounts of anonymized data, and managers need statistical reporting functions.

[0003] Traditional methods for building reproductive medicine databases often have the following problems: Traditional reproductive medicine databases are limited by the different data collection and recording standards adopted by different medical institutions, which makes data integration difficult. In the field of reproductive medicine, the correspondence between samples (eggs, sperm, embryos) and patient identities must be absolutely accurate, which is directly related to ethical and legal risks. The traditional method of relying on manual entry and barcode labeling has potential risks of confusion. Reproductive medicine involves multi-level biological data, from genomes, transcriptomes to cell functions and organ morphology. The integration of these multi-scale, multi-modal data has always been a technical challenge. Traditional methods often use simple data splicing or feature connection, which cannot effectively capture biological associations across scales. Summary of the invention

[0004] Based on this, it is necessary for the present invention to provide a method and system for constructing a reproductive medicine database to solve at least one of the above technical problems.

[0005] To achieve the above object, a method for constructing a reproductive medicine database comprises the following steps: Step S1: Obtain a biological tissue sample from a patient; obtain reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data, and genome variation data from the biological tissue sample from the patient; Step S2: performing time series analysis on the hormone fluctuation pattern according to the reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; performing image recognition processing on the embryo morphological parameter data based on the coherent superposition characteristics of light, and performing embryo development potential assessment to obtain embryo development potential data; Step S3: hierarchical clustering is performed on the individual heterogeneity of patients according to the immune cell function data and genomic variation data to obtain the patient immune-genetic typing data, wherein the hierarchical clustering process is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; Step S4: Perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

[0006] The present invention obtains biological tissue samples from patients and extracts reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genomic variation data, thereby achieving accurate collection of multi-dimensional information and providing a solid foundation for subsequent data analysis and processing. For reproductive endocrine hormone data, a time series analysis method is used to parse hormone fluctuation patterns and obtain endocrine dynamic characteristic data, so that the endocrine change trend of individuals can be quantified, which helps to identify potential physiological abnormalities or hormone imbalance problems, thereby improving the precise regulation and control capabilities of assisted reproduction. At the same time, based on the coherent superposition characteristics of light, image recognition processing is performed on embryo morphological parameter data, and combined with the embryo development potential assessment method, the developmental potential of the embryo can be accurately judged, making the screening of high-quality embryos more scientific and reducing errors caused by subjective judgment. In terms of immune-genetic feature analysis, through The memristor array with a double-layer oxide structure is hierarchically clustered to achieve deep mining of the patient's immune cell function data and genomic variation data. Each memristor has 64 adjustable conductance states, which can finely adjust the data clustering process, improve the recognition accuracy of individual heterogeneity, and then achieve accurate division of immune-genetic typing, making personalized diagnosis and treatment strategies more targeted. Based on the acquired endocrine dynamic feature data, embryonic development potential data and patient immune-genetic typing data, multi-dimensional feature fusion technology is used for deep integration, and a distributed knowledge graph index with homomorphic encryption is constructed, so that reproductive medical data can be efficiently stored and retrieved under the premise of ensuring privacy, improving data security and availability. The overall method forms a complete set of analysis processes from data acquisition, feature extraction, individual heterogeneity analysis to multi-dimensional data fusion, which not only improves the intelligence level of data processing in the field of reproductive medicine, but also provides reliable technical support for precision medicine and personalized assisted reproduction.

[0007] The present invention also provides a system for constructing a reproductive medicine database, which is used to execute the above-mentioned method for constructing a reproductive medicine database. The system for constructing a reproductive medicine database comprises: The biological data acquisition module is used to obtain biological tissue samples from patients; reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data are obtained from biological tissue samples from patients; The potential assessment module is used to perform time series analysis and processing on hormone fluctuation patterns based on reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; perform image recognition processing on embryo morphological parameter data based on the coherent superposition characteristics of light, and perform embryo development potential assessment to obtain embryo development potential data; The memristor heterogeneity typing module is used to perform hierarchical clustering processing on the individual heterogeneity of patients based on the immune cell function data and genomic variation data to obtain the patient's immune-genetic typing data, wherein the hierarchical clustering processing is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; The encrypted storage module is used to perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and to construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

[0008] The present invention systematically acquires biological tissue samples of patients through the biological data acquisition module, and accurately extracts reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data, so that personalized medical information can be fully integrated. Based on these data, the potential assessment module analyzes the timing of reproductive endocrine hormones, analyzes hormone fluctuation patterns, and constructs dynamic feature data to reveal the regulation of the endocrine system on the reproductive process. The embryo morphological parameter data is analyzed through high-precision image recognition and analysis using the coherent superposition characteristics of light to achieve quantitative assessment of embryonic development potential, ensuring that the assessment results have high stability and credibility. The memristor array with a double-layer oxide structure performs hierarchical clustering on immune cell function data and genomic variation data, identifies immune-genetic heterogeneity between individuals, and builds a refined patient typing model. The 64 adjustable conductance states in the memristor array enable it to accurately depict the complex information flow patterns between different immune-genetic subtypes, improving the resolution and adaptability of typing. At the data management level, the encryption storage module performs multi-dimensional feature fusion on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, so that the cross-correlation patterns between data can be deeply mined and stored in a homomorphic encryption manner to ensure that the data is not leaked during the calculation process. Furthermore, a knowledge graph index is constructed based on distributed ledger technology, so that the structure of the reproductive medicine database has traceability and tamper-proof capabilities, providing reliable data support for individualized reproductive health assessment. This database not only improves the systematic nature of reproductive health analysis, but also enables precision medical strategies to be implemented under data security guarantees, providing efficient and reliable intelligent auxiliary support for clinical decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 A schematic diagram of the steps of a method for constructing a reproductive medicine database according to the present invention; Figure 2 for Figure 1 Detailed step flow diagram of step S1; Figure 3 for Figure 1 Detailed step flow chart of step S3 in FIG. DETAILED DESCRIPTION

[0010] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0011] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different network and / or processor methods and / or microcontroller methods.

[0012] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0013] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for constructing a reproductive medicine database, the method comprising the following steps: Step S1: Obtain a biological tissue sample from a patient; obtain reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data, and genome variation data from the biological tissue sample from the patient; In this embodiment, biological tissue samples are first extracted from the patient's peripheral blood, and plasma and mononuclear cells are separated by flow cytometry sorting technology to obtain reproductive endocrine hormone data and immune cell function data, respectively. At the same time, for the acquisition of embryo morphological parameter data, a high-resolution optical microscope is used to collect real-time images of fertilized eggs and embryonic development, and morphological changes are recorded every 5 minutes until blastocyst formation. In addition, in order to obtain genomic variation data, DNA is extracted from biological tissue samples of patients, and genome analysis is performed using whole exome sequencing technology to detect key genetic information such as single nucleotide polymorphisms, insertion and deletion variations, and copy number variations. The sequencing depth is set to 100× to ensure data reliability. Finally, all data are stored in a standardized database for subsequent analysis.

[0014] Step S2: performing time series analysis on the hormone fluctuation pattern according to the reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; performing image recognition processing on the embryo morphological parameter data based on the coherent superposition characteristics of light, and performing embryo development potential assessment to obtain embryo development potential data; In this embodiment, for the time series analysis and processing of reproductive endocrine hormone data, the Savitzky-Golay filtering method is first used to smooth the data to remove short-cycle noise signals. Subsequently, a hormone fluctuation prediction model is established based on a long short-term memory neural network (LSTM), and nonlinear time series modeling is performed on the dynamic changes in hormone levels such as estradiol, progesterone, and luteinizing hormone, and key inflection point information is extracted to identify the endocrine dynamic characteristics of the ovulation cycle. For the processing of embryo morphological parameter data, based on the coherent superposition characteristics of light, spectral interference microscopy technology is used, combined with an improved U-Net deep learning model, to automatically identify and quantify parameters such as embryonic cell division, zona pellucida thickness, and blastocyst cavity expansion rate, and the developmental potential of the embryo is evaluated based on a multivariate regression model. Finally, the output endocrine dynamic feature data and embryo development potential data are used to guide personalized ovulation induction programs and embryo transplantation decisions, respectively.

[0015] Step S3: hierarchical clustering is performed on the individual heterogeneity of patients according to the immune cell function data and genomic variation data to obtain the patient immune-genetic typing data, wherein the hierarchical clustering process is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; In this example, the immune cell function data was first standardized, and the immune cell subtype composition in peripheral blood was analyzed using single-cell RNA sequencing technology. The functional status of immune cells was evaluated by cytokine secretion detection, including the cytotoxicity level of T cells and the immunosuppressive ability of regulatory T cells. Then, combined with the genomic variation data, the improved hierarchical clustering algorithm was used to perform hierarchical clustering analysis on the immune-genetic characteristics of the patients. The double-layer oxide memristor array is implemented, where each memristor node corresponds to a gene-immune feature combination, and the differences in immune response and genetic background of different patients are simulated by adjusting the conductance state of the memristor (up to 64 adjustable conductance values). Ultimately, the clustering results divide the patients into different immune-genetic subtypes.

[0016] Step S4: Perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

[0017] In this embodiment, the acquired endocrine dynamic feature data, embryonic development potential data and patient immune-genetic typing data are firstly fused in multiple dimensions, and a self-attention mechanism based on the Transformer architecture is adopted to establish a global dependency relationship between different feature dimensions to enhance the interactive information between data. Subsequently, the fused data is encrypted using homomorphic encryption technology to ensure the security of patient privacy in a distributed computing environment. The encrypted data adopts the CKKS (Cheon-Kim-Kim-Song) homomorphic encryption scheme, which supports addition and multiplication operations without decryption, thereby ensuring the security of data in remote storage and calculation. Finally, based on the knowledge graph construction method, the data is organized in RDF (Resource Description Framework) format, and the graph database is used for index optimization to achieve efficient query and retrieval functions, and finally a reproductive medicine database structure is formed to provide a high-quality data foundation for clinical decision support and scientific research analysis.

[0018] The present invention obtains biological tissue samples from patients and extracts reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genomic variation data, thereby achieving accurate collection of multi-dimensional information and providing a solid foundation for subsequent data analysis and processing. For reproductive endocrine hormone data, a time series analysis method is used to parse hormone fluctuation patterns and obtain endocrine dynamic characteristic data, so that the endocrine change trend of individuals can be quantified, which helps to identify potential physiological abnormalities or hormone imbalance problems, thereby improving the precise regulation and control capabilities of assisted reproduction. At the same time, based on the coherent superposition characteristics of light, image recognition processing is performed on embryo morphological parameter data, and combined with the embryo development potential assessment method, the developmental potential of the embryo can be accurately judged, making the screening of high-quality embryos more scientific and reducing errors caused by subjective judgment. In terms of immune-genetic feature analysis, through The memristor array with a double-layer oxide structure is hierarchically clustered to achieve deep mining of the patient's immune cell function data and genomic variation data. Each memristor has 64 adjustable conductance states, which can finely adjust the data clustering process, improve the recognition accuracy of individual heterogeneity, and then achieve accurate division of immune-genetic typing, making personalized diagnosis and treatment strategies more targeted. Based on the acquired endocrine dynamic feature data, embryonic development potential data and patient immune-genetic typing data, multi-dimensional feature fusion technology is used for deep integration, and a distributed knowledge graph index with homomorphic encryption is constructed, so that reproductive medical data can be efficiently stored and retrieved under the premise of ensuring privacy, improving data security and availability. The overall method forms a complete set of analysis processes from data acquisition, feature extraction, individual heterogeneity analysis to multi-dimensional data fusion, which not only improves the intelligence level of data processing in the field of reproductive medicine, but also provides reliable technical support for precision medicine and personalized assisted reproduction.

[0019] Preferably, step S1 comprises the following steps: Step S11: obtaining a biological tissue sample of a patient; In this embodiment, peripheral blood or endometrial tissue is first sampled from the patient. Peripheral blood samples are collected using EDTA anticoagulation tubes to ensure the integrity of cell components, while endometrial tissue samples are obtained through hysteroscopic biopsy. Subsequently, the biological tissue samples are transported to the laboratory at 4°C and pre-treated in a biosafety cabinet, including washing the blood samples with PBS buffer, separating mononuclear cells by density gradient centrifugation (centrifugation conditions set at 400g for 15 minutes), or digesting endometrial tissue with tissue dissociation enzymes (such as 0.25% trypsin) to release single cell suspensions. Finally, the concentration of the resulting cell suspension is adjusted to approximately 1×10 per milliliter. 6 cells to ensure the stability of subsequent experiments.

[0020] Step S12: The patient's biological tissue sample is introduced into the PDMS microfluidic chip with an integrated CdSe / ZnS core-shell structure quantum dot sensor array to obtain the dispersed flow data of the biological sample, wherein the PDMS microfluidic chip has a multi-channel structure with a channel width of 30-50 m; In this embodiment, a PDMS microfluidic chip with an integrated CdSe / ZnS core-shell structure quantum dot sensor array is used for flow detection of biological samples. The chip has a multi-channel structure, and the width of each channel is 30-50μm to match the flow characteristics of a single cell. First, the microfluidic chip is placed in a CO2 incubator for pre-wetting for 30 minutes to enhance the affinity of the biological fluid. Then, a syringe pump is used to introduce the patient's biological tissue sample into the microfluidic channel at a constant flow rate (about 1μL / min), so that the cells disperse and flow under the action of the fluid shear force, and ensure that the cells do not adhere to the channel wall. The flow state of cells in the microfluidic channel is monitored in real time by a fluorescence microscopy imaging system, and the spatial distribution data and flow rate change data of the cells are recorded to analyze the dispersion characteristics of different cell types in a microfluidic environment.

[0021] Step S13: modifying specific antibodies on the dispersed flow data of the biological sample on the quantum dot sensor array, and sorting and capturing NK cells, T cell subsets, B cells and monocytes in the patient's biological tissue sample, thereby obtaining immune cell capture data, wherein the modified specific antibodies include CD56, CD3, CD4, CD8, CD19, CD14 and HLA-DR; In this embodiment, a quantum dot sensor array is used to sort and capture target immune cells in flowing biological samples. First, specific antibodies are modified on the surface of the quantum dot sensor array, including CD56 (for identifying NK cells), CD3 (for identifying T cells), CD4 (for identifying helper T cells), CD8 (for identifying cytotoxic T cells), CD19 (for identifying B cells), CD14 (for identifying monocytes) and HLA-DR (for identifying antigen presenting cells). The antibody modification adopts a biotin-streptavidin coupling strategy to ensure that the antibody is stably fixed on the surface of the quantum dots. Subsequently, the biological tissue sample flows through the sensor array, and after the specific antibody specifically binds to the corresponding cell surface antigen, the unbound cells are eluted by a weak fluid shear force, and the target cells are finally retained on the array. The captured immune cells are photographed using a fluorescence microscope, and the number of captured immune cells of each type is calculated using automatic image analysis software to obtain immune cell capture data.

[0022] Step S14: performing multi-spectral excitation in the wavelength range of 525-655 nm on the immune cell capture data, and recording changes in quantum dot fluorescence resonance energy transfer signals, thereby obtaining immune cell function data; In this embodiment, in order to analyze and capture the functional state of immune cells, multi-spectral excitation (525-655nm wavelength range) is used to record the changes in quantum dot fluorescence resonance energy transfer (FRET) signals. First, a single-cell laminar flow excitation device is used to sequentially irradiate quantum dot-labeled cells in the wavelength range of 525nm, 575nm and 655nm to measure the changes in fluorescence intensity at different excitation wavelengths. Secondly, based on the FRET mechanism, the fluorescence energy transfer effect caused by changes in the conformation of surface receptor molecules when immune cells are in an active state is observed. For example, when T cells are activated, the activation of the CD3 / CD28 signaling pathway will cause the enhancement or attenuation of specific fluorescence signals. The FRET signal is recorded by a high-throughput spectral imaging system, and the multi-channel fluorescence data is analyzed by principal component analysis (PCA) to reduce the dimensionality of the multi-channel fluorescence data to distinguish the functional states of different immune cell subsets, and finally obtain the immune cell function data.

[0023] Step S15: Acquire reproductive endocrine hormone data, embryo morphological parameter data, and genome variation data from the patient's biological tissue sample.

[0024] In this embodiment, reproductive endocrine hormone data of the patient's biological tissue samples are first obtained, and eluting hormone (estradiol) is detected by ELISA (enzyme-linked immunosorbent assay). )、Progesterone( ), luteinizing hormone (LH) and follicle-stimulating hormone (FSH) concentrations, and an automated chemiluminescence immunoassay was used for quantitative analysis to ensure the sensitivity and accuracy of hormone determination. Secondly, for the acquisition of embryo morphological parameter data, an embryo monitoring system based on coherent light imaging was used to collect images of embryo morphological changes every 5 minutes, and key parameters such as zona pellucida thickness and blastocyst cavity volume were recorded. Finally, to analyze genomic variation data, DNA was extracted from patient biological tissue samples, and whole exome sequencing was performed using high-throughput sequencing technology (such as Illumina NovaSeq 6000). The coverage depth of sequencing data was set to 100×, and mutation screening was performed using bioinformatics algorithms (such as the GATK variation detection tool) to obtain genomic variation information related to reproductive health. Ultimately, the obtained reproductive endocrine hormone data, embryo morphological parameter data, and genomic variation data will be used for personalized reproductive medicine decision support.

[0025] The present invention obtains biological tissue samples of patients and introduces a PDMS microfluidic chip with an integrated CdSe / ZnS core-shell structure quantum dot sensor array, so that biological samples can achieve efficient dispersed flow in a multi-channel structure, ensuring the uniformity and stability of sample processing, while the 30-50μm channel width of the multi-channel structure is suitable for the flow behavior characteristics of different cell types, improving the accuracy of cell screening. On this basis, by modifying specific antibodies on the quantum dot sensor array, the sorting and capture of NK cells, T cell subsets, B cells and monocytes in the patient's biological tissue samples is achieved, and different immune cell types are accurately identified and separated, making subsequent analysis more targeted. The specific antibody markers CD56, CD3, CD4, CD8, CD19, CD14 and HLA-DR used cover the main immune cell types, ensuring the comprehensiveness of immune function evaluation. Further, through the multi-spectral excitation technology in the wavelength range of 525-655nm, combined with the changes in quantum dot fluorescence resonance energy transfer signals, the fine detection of immune cell functions is achieved, making the dynamic monitoring of immune response status more sensitive and accurate. In addition, the acquisition of reproductive endocrine hormone data, embryo morphological parameter data, and genomic variation data from patient biological tissue samples enables comprehensive integration of core biological information related to reproductive medicine, providing multi-dimensional data support for precision medicine analysis. The overall solution relies on microfluidics technology, quantum dot labeling, and multi-spectral detection to achieve efficient integration of immune cell screening, functional evaluation, and reproductive-related data acquisition, providing a precise technical foundation for personalized diagnosis and treatment and optimization of assisted reproductive strategies.

[0026] Preferably, step S15 comprises the following steps: Step S151: processing the follicular fluid in the patient's biological tissue sample by a microfluidic chip electrochemical sensor array to obtain a primary endocrine data stream, wherein the microfluidic chip electrochemical sensor array comprises 64 independent electrode units, and the surface of each unit is modified with specific aptamers for estradiol, progesterone, FSH and LH; In this embodiment, follicular fluid samples are collected from the patient's body to ensure that the temperature of the sample is maintained at 37°C to reduce the loss of biological activity. The sample is introduced into the microfluidic chip electrochemical sensor array through a microfluidic channel made of PDMS material. The array contains 64 independent electrode units, and the surface of each unit is modified with specific aptamers by chemical self-assembly method, including aptamers for estradiol, progesterone, follicle stimulating hormone (FSH) and luteinizing hormone (LH). After the sample enters the channel, an external micro-electric field (voltage range 0.1V-0.5V) is applied to promote the binding of hormone molecules to the aptamer, and the current signal change of each electrode unit is measured by differential pulse voltammetry (DPV), and converted into the corresponding primary endocrine data stream, and the real-time concentration change is recorded with nanoampere current resolution.

[0027] Step S152: Use the diamond quantum sensor based on nitrogen vacancy defects prepared by 28nm process to detect the hormone concentration with pT level accuracy on the primary endocrine data stream, so as to obtain a high-precision endocrine concentration matrix, wherein the detection accuracy of the diamond quantum sensor reaches Molar concentration, sampling frequency is 300Hz; In this embodiment, the obtained primary endocrine data stream is input into a diamond quantum sensor based on nitrogen vacancy defects for ultra-high precision hormone concentration determination. The diamond quantum sensor used is prepared using a 28nm process and uses optical detection magnetic resonance (ODMR) technology to detect the spin resonance signal of the nitrogen vacancy center through microwave excitation and 532nm laser pumping, thereby achieving pT-level concentration detection of estradiol, progesterone, FSH and LH. During the measurement process, the sampling frequency of the sensor is set to 300Hz to ensure that it can capture rapidly changing hormone secretion events. At the same time, the detection accuracy is further improved by Fourier transforming the optical readout signal to reach the level of 10^(-12) molar concentration. Ultimately, all measured values ​​constitute a high-precision endocrine concentration matrix, providing accurate hormone concentration information for subsequent analysis.

[0028] Step S153: extracting pulse signals from the high-precision endocrine concentration matrix to obtain hormone secretion pulse characteristic data; In this embodiment, a pulse signal extraction algorithm is applied to the high-precision endocrine concentration matrix to identify the pulse-like change characteristics of hormone concentrations. The time-frequency analysis method based on continuous wavelet transform (CWT) is used to normalize the hormone concentrations at different time points, and a threshold (such as 1.5 times the standard deviation) is set to screen significant pulse signals. This method can effectively distinguish between basal secretion and pulse secretion patterns, thereby obtaining hormone secretion pulse characteristic data, including important parameters such as the amplitude, frequency, peak interval and deceleration rate of the secretion peak, providing data support for evaluating the patient's endocrine function.

[0029] Step S154: combining the hormone secretion pulse characteristic data with the time series mark to obtain reproductive endocrine hormone data; In this embodiment, the acquired hormone secretion pulse characteristic data is combined with the time series mark to construct a reproductive endocrine hormone data set. The acquisition of time series marks is based on factors such as the patient's circadian rhythm, cyclical hormone secretion pattern, and ovulation cycle stage. The pulse data is automatically aligned using a time synchronization algorithm, and the long short-term memory (LSTM) network is used to model the data in time series to predict the hormone fluctuation trend in the future. The results of this step can be used to analyze the individualized reproductive endocrine status and provide a reference for subsequent embryo culture and transplantation decisions.

[0030] Step S155: continuously monitoring the patient's biological tissue sample through an integrated microfluidic culture chamber and a phase contrast microscopy imaging system to obtain an embryo dynamic development image sequence, wherein the phase contrast microscopy imaging system is composed of a frequency-adjustable optical resonant cavity and a high-speed CMOS image sensor; In this embodiment, the patient's embryonic sample is placed in a microfluidic culture chamber for continuous in vitro culture, and an integrated phase contrast microscopy system is used for real-time monitoring. The system consists of a tunable optical resonant cavity and a high-speed CMOS image sensor, and uses optical coherence phase shifting technology (PCI) to record label-free imaging data of embryonic cells. The resonant frequency of the optical resonant cavity is set at 560THz, which can complete single-frame imaging within 1ms, ensuring that the key dynamic changes in the embryonic development process are fully recorded. The imaging data is continuously collected at a frame rate of 50fps, and the clarity of cell boundaries is improved by an automatic image enhancement algorithm, thereby obtaining a high-resolution embryonic dynamic development image sequence.

[0031] Step S156: performing optical coherence processing on the embryo dynamic development image sequence to obtain a multi-parameter morphological feature set of the embryo, wherein the optical coherence processing is implemented using a photon interference calculation unit including a 256×256 array of integrated microring resonators; In this embodiment, the acquired embryo dynamic development image sequence is subjected to optical coherence processing to extract multi-parameter morphological features. Optical coherence processing is implemented using a photon interference calculation unit, which contains a 256×256 array of integrated microring resonators, which can perform light field information demodulation and extract key features such as embryonic cell growth rate, volume change, and refractive index distribution. First, the photon interference calculation unit performs spatiotemporal coherence analysis on embryo images at different time points to extract the refractive index change pattern between cells; then, a deep convolutional neural network (CNN) is used to extract high-dimensional features from the interference pattern to construct an embryo multi-parameter morphological feature set to provide data support for subsequent embryo quality assessment.

[0032] Step S157: performing optical calculation processing on the embryo multi-parameter morphological feature set to obtain embryo morphological parameter data, wherein the embryo morphological parameter data includes cleavage rate, cleavage symmetry, blastocyst cavity formation dynamics and chromosome euploidy prediction value; In the present embodiment, optical calculation processing is carried out to embryo multi-parameter morphological feature set to obtain embryo morphological parameter data. In the process, firstly, the cleavage rate is calculated using the morphological analysis algorithm based on curve fitting, and the uniformity of cleavage cell division is evaluated using the morphological symmetry measurement function. Secondly, the time characteristics of blastocyst cavity formation are detected using dynamic image processing technology, including blastocyst cavity expansion rate and cell mass contraction frequency. In addition, the chromosome euploidy prediction model trained by machine learning algorithm, combined with cell division dynamics and blastocyst morphological characteristics, is used to predict the genetic stability of embryo, thereby obtaining complete embryo morphological parameter data, for assisted reproduction clinical decision-making provides basis.

[0033] Step S158: Perform reproductive-related variation site analysis on the nucleic acid extracted from the patient's biological tissue sample to obtain genomic variation data.

[0034] In this embodiment, nucleic acids are extracted from biological tissue samples of patients, and reproductive-related variant site analysis is performed. The magnetic bead method is used for nucleic acid extraction to ensure efficient enrichment and purification of DNA and RNA. Next, the target gene region is sequenced using a high-throughput sequencing (NGS) platform, with special attention to genes related to reproductive health, such as FOXL2, AMH, FSHR, and ESR1. The variation analysis uses a bioinformatics pipeline, including base quality filtering, alignment to the reference genome, single nucleotide polymorphism (SNP) and insertion and deletion (INDEL) detection, and combines machine learning methods to predict the functional impact of the variation. Finally, genomic variation data is generated to provide data support for the patient's reproductive health assessment.

[0035] The present invention processes follicular fluid through a microfluidic chip electrochemical sensor array, so that key reproductive hormones such as estradiol, progesterone, FSH and LH can be accurately quantified under the efficient detection of 64 independent electrode units, and the specific aptamer modified on the surface of each electrode ensures the high sensitivity and specificity of the detection, thereby obtaining a stable primary endocrine data stream. Subsequently, the nitrogen vacancy defect diamond quantum sensor based on the 28nm process is used to perform pT-level high-precision detection of hormone concentrations, achieving ultra-sensitive measurement of 10^(-12) molar concentrations, and the high sampling frequency of 300Hz ensures the timing capture capability of hormone secretion fluctuations, making the establishment of a high-precision endocrine concentration matrix more comprehensive and accurate. On this basis, the high-precision data is converted into hormone secretion pulse characteristic data through pulse signal extraction technology, and combined with time series markers, complete reproductive endocrine hormone data is formed, providing complete timing information support for subsequent analysis. In addition, the dynamic development process of the embryo is continuously monitored by integrating the microfluidic culture chamber and phase contrast microscopy system to ensure the acquisition of data for the entire process of cleavage and blastocyst development. The phase contrast microscopy system combines a frequency-adjustable optical resonant cavity and a high-speed CMOS image sensor, which enables the embryo image acquisition to have high resolution and high frame rate characteristics, accurately restoring the details of embryo growth. The obtained embryo dynamic development image sequence is subjected to optical coherence processing using a photon interference calculation unit of an integrated microring resonator to extract a multi-parameter morphological feature set including cleavage rate, cleavage symmetry, blastocyst cavity formation dynamics, and chromosome euploidy prediction values, and optical calculations are used to further improve the accuracy of embryo morphological analysis, making it more valuable for reference. At the same time, the nucleic acids in biological tissue samples are analyzed to accurately screen for reproductive-related variation sites, thereby obtaining high-quality genomic variation data to ensure comprehensive information support at the genetic level. The overall solution combines a variety of advanced technologies such as electrochemical sensing, quantum precision measurement, microfluidic imaging monitoring and optical coherence computing, achieving a deep integration of reproductive endocrine analysis, embryo morphology assessment and genetic background analysis, providing accurate and systematic data support for reproductive health research and assisted reproductive medicine.

[0036] Preferably, step S158 includes the following steps: Step S1581: performing single-molecule sequencing based on a graphene nanopore array on nucleic acids extracted from a patient's biological tissue sample, thereby obtaining a DNA passing signal with a single-base resolution; In this embodiment, the cells in the patient's biological tissue sample are lysed and DNA is extracted, the target DNA fragments are enriched by the magnetic bead method, and the sequencing library is constructed by the method of end modification and connection adapter. Then, the library is loaded into a single-molecule sequencing platform based on a graphene nanopore array, and the DNA single strands are passed through the graphene nanopores one by one by applying a transmembrane electric field, while the current changes are monitored in real time. The graphene nanopore has a sub-nanometer pore size and is combined with surface functional group modification to improve the selectivity and stability of DNA molecules. When DNA passes through the nanopore, different base genes have different steric hindrances and charge distributions, and characteristic current blocking signals are generated when passing through the pores. The system records these signals through a highly sensitive current measurement module, thereby obtaining DNA passing signals with single-base resolution. The process uses a 100kHz sampling frequency for data acquisition, and combines a machine learning model to correct noise interference in real time to ensure signal accuracy.

[0037] Step S1582: The DNA passing signals are collected and processed in parallel by a multi-channel graphene transistor array, thereby obtaining original genome sequence data, wherein the sensitivity of the multi-channel graphene transistor array is sufficient to detect the current change caused by the passage of a single DNA molecule; In this embodiment, the obtained DNA is input into a multi-channel graphene transistor array through a signal, each channel is composed of an independent graphene field effect transistor (GFET), and its surface is modified with a specific probe to enhance the response to the DNA signal. Through external voltage control, the charge change induced by the passage of DNA molecules can modulate the conductivity of the graphene transistor to form a current signal, which is directly related to the type of DNA base. The array adopts a parallel architecture of 128×128 channels, and the detection limit of each channel can reach the single molecule level, and has a current sensitivity of 1pA. The collected signal is amplified by a high-precision low-noise amplifier, and then converted into a digital signal in parallel by a high-speed analog-to-digital converter (ADC), and transmitted to the central processing unit (CPU) in real time for preliminary data integration, and finally the original genome read sequence data is obtained.

[0038] Step S1583: performing base accuracy verification on the original genome read sequence to obtain high-precision genome sequencing data; In this embodiment, the base accuracy of the original genome read data is checked. First, the noise model based on the Markov Chain Monte Carlo (MCMC) method is used to evaluate the sequencing errors in the data and remove abnormal signals. Then, a recurrent neural network (RNN) model is constructed using a deep learning algorithm to compare the DNA signals collected by multiple channels, and error correction is performed in combination with a statistical probability model. In this process, a reference genome is used as a correction template, and the Smith-Waterman alignment algorithm is used for local alignment to detect base substitutions, insertions, and deletions. Finally, based on multiple rounds of iterative confidence weighted processing, high-confidence sequencing fragments are screened out to obtain high-precision genome sequencing data with a base accuracy of 99.99%.

[0039] Step S1584: Use the memristor neuromorphic computing unit to perform reproductive-related variation site analysis on the high-precision genome sequencing data to obtain genome variation data, wherein the genome variation data includes single nucleotide polymorphism, copy number variation and structural variation information of reproductive-related genes.

[0040] In this embodiment, high-precision genome sequencing data is input into a memristor neuromorphic computing unit, which is based on a cross-array memristor (RRAM) architecture and can simulate synaptic weighted computing, with low power consumption and high parallelism. First, a convolutional neural network (CNN) is used to extract features from the sequencing data, convert the DNA sequence into a numerical matrix, and perform pattern matching analysis. Subsequently, a long short-term memory (LSTM) network is used to identify the variation sites of reproductive-related genes, including single nucleotide polymorphisms (SNPs), copy number variations (CNVs), and structural variations (SVs). The system is trained using known variation data in the genome database and calculates the probability of occurrence of each variation through Bayesian inference. Finally, key gene variations that may affect reproductive function are extracted and classified according to their degree of influence to obtain the patient's genome variation data, which can be used to assess reproductive health risks or develop personalized medical plans.

[0041] The present invention processes follicular fluid through a microfluidic chip electrochemical sensor array, so that key reproductive hormones such as estradiol, progesterone, FSH and LH can be accurately quantified under the efficient detection of 64 independent electrode units, and the specific aptamer modified on the surface of each electrode ensures the high sensitivity and specificity of the detection, thereby obtaining a stable primary endocrine data stream. Subsequently, the nitrogen vacancy defect diamond quantum sensor based on the 28nm process is used to perform pT-level high-precision detection of hormone concentrations, achieving ultra-sensitive measurement of 10^(-12) molar concentrations, and the high sampling frequency of 300Hz ensures the timing capture capability of hormone secretion fluctuations, making the establishment of a high-precision endocrine concentration matrix more comprehensive and accurate. On this basis, the high-precision data is converted into hormone secretion pulse characteristic data through pulse signal extraction technology, and combined with time series markers, complete reproductive endocrine hormone data is formed, providing complete timing information support for subsequent analysis. In addition, the dynamic development process of the embryo is continuously monitored by integrating the microfluidic culture chamber and phase contrast microscopy system to ensure the acquisition of data for the entire process of cleavage and blastocyst development. The phase contrast microscopy system combines a frequency-adjustable optical resonant cavity and a high-speed CMOS image sensor, which enables the embryo image acquisition to have high resolution and high frame rate characteristics, accurately restoring the details of embryo growth. The obtained embryo dynamic development image sequence is subjected to optical coherence processing using a photon interference calculation unit of an integrated microring resonator to extract a multi-parameter morphological feature set including cleavage rate, cleavage symmetry, blastocyst cavity formation dynamics, and chromosome euploidy prediction values, and optical calculations are used to further improve the accuracy of embryo morphological analysis, making it more valuable for reference. At the same time, the nucleic acids in biological tissue samples are analyzed to accurately screen for reproductive-related variation sites, thereby obtaining high-quality genomic variation data to ensure comprehensive information support at the genetic level. The overall solution combines a variety of advanced technologies such as electrochemical sensing, quantum precision measurement, microfluidic imaging monitoring and optical coherence computing, achieving a deep integration of reproductive endocrine analysis, embryo morphology assessment and genetic background analysis, providing accurate and systematic data support for reproductive health research and assisted reproductive medicine.

[0042] Preferably, the time series analysis processing of the hormone fluctuation pattern in step S2 includes: The reproductive endocrine hormone data were normalized on the time axis, and a 400-hour unified time window was constructed with the first day of the menstrual cycle as the reference point to obtain standardized hormone data; The standardized hormone data were sampled at the femtosecond level with a time resolution of 10 femtoseconds to obtain a high time resolution hormone fluctuation curve; Extract intrinsic mode function sets from high time resolution hormone fluctuation curves; Correlation analysis was performed on the intrinsic mode function set to obtain the synergistic effect data between hormones; Extract nonlinear dynamic features from the synergistic effect data between hormones; The nonlinear dynamic characteristics were systematically identified to obtain the hormone regulation dynamics model; The hormone regulation dynamics model is simulated through forward evolution to obtain endocrine dynamic characteristic data.

[0043] In this embodiment, reproductive endocrine hormone data of subjects for several consecutive months, including luteinizing hormone (LH), follicle stimulating hormone (FSH), estradiol (E2), progesterone (P4), etc., are collected, and the starting time point of the menstrual cycle is recorded. First, according to the length of the subject's menstrual cycle, the hormone data of different cycles are converted to the same reference time axis, and the first day of the menstrual cycle is used as the time reference point to construct a unified time window of 400 hours, and the linear time scaling method is used for cycle normalization. For missing data points, the bicubic interpolation method is used to complete them to ensure the integrity of the time series. Finally, a hormone concentration data matrix represented by a standardized time axis is obtained, providing a unified time reference frame for subsequent analysis. High-precision optical detection technology is used to perform ultra-high resolution time sampling of standardized hormone data. Ultrafast dynamics measurement of hormone molecules in biological fluids is performed using femtosecond optical pulse lasers combined with Raman scattering spectroscopy analysis. First, a plasma or urine sample is placed in a nanofluidic channel, and a multispectral acquisition is performed using an ultrafast time-resolved optical detection system to obtain continuous data points at intervals of 10 femtoseconds (fs). Subsequently, a time interpolation reconstruction algorithm (such as spline interpolation) is used to reconstruct the discrete data to generate a high-time resolution hormone fluctuation curve, which can more accurately reflect the instantaneous dynamic changes in hormone secretion. The empirical mode decomposition (EMD) method is used to decompose the high-time resolution hormone fluctuation curve and extract its intrinsic mode function (IMF). First, the envelope fitting method is used to separate the different time-frequency components of the hormone fluctuation data, and the intrinsic mode functions containing different frequency characteristics are decomposed layer by layer. Then, the Hilbert transform is used to calculate the instantaneous frequency and instantaneous amplitude of each IMF component, and the Hilbert time-frequency spectrum is constructed to analyze the time-varying characteristics of hormone fluctuations. For example, in the process of hormone regulation, different IMF components may correspond to short-period pulse release, circadian rhythm changes, and longer-term regulatory trends, respectively. This information can be used to reveal the dynamic regulatory mechanism of reproductive endocrine hormones. The extracted IMF components were subjected to multidimensional correlation analysis by an adaptive time-frequency joint analysis method combining wavelet packet transform (WPT) and short-time Fourier transform (STFT). First, the time-frequency matrix of the IMF components of different hormones was constructed, and their cross-time-frequency spectrum was calculated to evaluate the time-frequency coupling relationship between different hormones. Then, the mutual information analysis method was used to quantify the degree of synergistic changes between hormone pairs, and to screen out hormone pairs with high temporal correlation and resonance modes. For example, some IMF components of LH and FSH may show high time-frequency coupling before ovulation, reflecting their synergistic regulatory effects. Finally, the synergistic effect data between hormones were output, including characteristic parameters such as synchronization, phase shift and coupling strength of different hormone pairs. The phase space trajectory of the hormone system was constructed by the delayed coordinate reconstruction method, and its nonlinear characteristics were analyzed. First, the optimal embedding dimension and time delay parameter were determined by Takens' theorem, and the synergistic effect data were mapped to a high-dimensional phase space.Then, the maximum Lyapunov exponent was calculated to evaluate the chaotic characteristics of the system, and the correlation dimension was used to measure the fractal structure of the system. In addition, recurrence plot analysis was applied to extract the periodic and non-periodic components of the hormone system and identify the endocrine regulation patterns in different physiological stages. For example, during the ovulation period, the hormone system may show low-dimensional chaotic characteristics, while during the luteal phase, it may show more stable quasi-periodic oscillations. The fractional derivative method was used to construct a hormone regulation dynamics model to describe the long-term memory effect of hormone secretion and metabolism. First, based on the experimental data, the fractional order of the hormone dynamics system was estimated by the least squares fitting method, and the state equation of the model was defined by the Caputo fractional differential operator. Then, the particle swarm optimization (PSO) algorithm was used to globally optimize the model parameters to fit the experimental data, and the model was cross-validated by multiple groups of subject data. For example, the fractional kinetic model can describe the time-delayed feedback regulation process of estrogen on luteinizing hormone and reveal the dynamic response pattern of the hormone system under different physiological states. Finally, the output hormone regulation dynamics model can be used to predict hormone level fluctuations and early identification of abnormal states. The numerical simulation framework based on the lattice Boltzmann method (LBM) is used to simulate the hormone regulation dynamics model. First, an LBM grid suitable for nonlinear physiological systems is constructed, and the hormone concentration distribution function is defined based on the Boltzmann transport equation. Then, the BGK (Bhatnagar-Gross-Krook) collision operator is used to calculate the diffusion and feedback regulation process of hormones to simulate the evolutionary behavior of hormones on a time scale. For example, in the simulation, the initial concentrations of LH, FSH, E2, and P4 can be set, and periodic perturbations can be applied to simulate physiological rhythm changes to observe the stability and self-organization behavior of the hormone system. The simulation results can be used to predict the potential mechanisms of abnormal hormone fluctuations, such as precise intervention plans for polycystic ovary syndrome (PCOS) or luteal insufficiency.

[0044] The present invention uses time axis standardization processing, takes the first day of the menstrual cycle as the reference point, and constructs a unified time window of 400 hours, so that the hormone data of different patients can be compared and analyzed in the same time frame, and the consistency and comparability of the data are improved. Subsequently, a femtosecond sampling technology with a time resolution of 10 femtoseconds is used to accurately capture the slight fluctuations of hormone levels, ensure the high time accuracy of the hormone fluctuation curve, and truly reflect the dynamic changes of the reproductive endocrine system. The high time resolution hormone fluctuation curve is decomposed by Hilbert-Huang transform, and the intrinsic mode function set under different time scales is extracted, which helps to identify periodic signals and short-time oscillation characteristics, and retains nonlinear information in the hormone regulation process. On this basis, the intrinsic mode function set is analyzed for correlation based on the adaptive time-frequency joint analysis technology, and the synergistic change pattern between hormones is accurately analyzed to reveal the coupling relationship between various hormones in physiological regulation. Further extract nonlinear dynamic characteristics, characterize the complex feedback mechanism in the hormone regulation process, and identify the system through fractional differential equation modeling, so that the hormone regulation dynamic model can accurately describe the regulation characteristics of the endocrine system. Finally, the lattice Boltzmann method was used to perform forward evolution simulation on the hormone regulation dynamics model to predict the changing trends of hormone levels at different time scales, provide a global perspective on the dynamic regulation of the reproductive endocrine system, and provide precise data support for the formulation of personalized intervention strategies.

[0045] Preferably, the image recognition processing based on the coherent superposition characteristic of light in step S2 includes: The embryo morphological parameter data were imported into a silicon-based photonic interferometer chip integrated with a 128×128 microring resonator array to obtain multi-angle embryo light field information; Perform phase modulation processing on multi-angle embryo light field information to obtain embryo characteristic light phase map; The coherence of light is used to perform parallel optical calculation on the characteristic light phase image of the embryo to obtain the embryo morphological characteristic vector; The embryo morphological feature vector is transmitted at low loss to obtain high-fidelity morphological feature data, where the waveguide loss of the low-loss transmission is less than 0.01dB / cm; Perform nonlinear transformation on high-fidelity morphological feature data to obtain embryonic structure-function correlation features; Dimensional optimization of embryonic structure-function association features to obtain optimized morphological feature representation; The optimized morphological feature representation is subjected to interferometric measurement-based classification processing to obtain embryo morphological data.

[0046] In this embodiment, embryo morphological parameter data, including key indicators such as embryo cell number, cytoplasm uniformity, fragmentation rate, and zona pellucida thickness, are first obtained, and converted into optical signals and input into a silicon-based photonic interferometer chip. The chip uses a 128×128 microring resonator array, and each microring resonator can regulate the interference state of light waves of a specific wavelength to form multi-angle optical information. In order to ensure accurate data input, a high-speed optical modulator is used to map the morphological parameters to the amplitude, phase, and polarization direction of the optical signal, and the optical signal is introduced into the chip by optical fiber coupling. During the experiment, the wavelength of the incident light was set to the near-infrared band of 1550nm to obtain a higher optical resolution, and a low-loss silicon optical waveguide was used to reduce signal attenuation. When the embryo morphological parameter optical signal enters the microring resonator array, different resonators selectively regulate the light of a specific wavelength, so that the multi-angle embryonic light field information forms a coherent superposition inside the array. In order to further extract effective information, a phase modulation technology based on the electro-optical effect is used to adjust the phase offset of the optical signal by applying an external modulated electric field to form an embryonic characteristic light phase map. Specifically, the intensity of the modulated electric field is controlled within the range of 10V / μm, and the modulation frequency is set to 10GHz to achieve fine control of the light field distribution inside the microring and ensure high-precision phase adjustment. After obtaining the characteristic light phase map of the embryo, optical coherence technology is used for information calculation. The specific method is to use a Mach-Zehnder interferometer array to interfere and superimpose the coherent light signals output by different microring resonators, and demodulate the morphological feature vector of the embryo through the light intensity distribution. During this calculation process, the optical path difference of each interference channel is controlled at the sub-wavelength level, with a typical value of less than 100nm, to ensure the high resolution of the interference fringes. The entire optical calculation process is executed in parallel on the chip without the need for traditional electronic computing resources, thereby achieving ultra-high-speed data processing and ensuring the real-time extraction of morphological features. After obtaining the embryo morphological feature vector, it needs to be transmitted over a long distance to ensure the integrity and accuracy of the data. To this end, a low-loss optical transmission solution based on silicon optical waveguide is adopted, and high-purity silicon materials with a loss of less than 0.01dB / cm are selected, and a passive mode matching structure is integrated in the waveguide to reduce scattering loss. In the specific implementation process, the waveguide width is optimized to 450nm, and the bending radius is designed to be not less than 5μm to reduce the mode mismatch loss. At the same time, in order to prevent environmental noise interference, the entire transmission channel adopts optical fiber packaging technology to improve the anti-interference ability, so as to obtain high-fidelity embryo morphological data. After obtaining high-fidelity embryo morphological feature data, the nonlinear transformation method is used to analyze the structure-function association characteristics of the embryo. Specifically, the nonlinear Fourier transform (NFT) method is used to map the light field data in the spatiotemporal domain to the nonlinear spectral domain, so as to extract the complex nonlinear relationship between embryo morphology and developmental potential.In the experiment, the calculation window of the nonlinear transformation was set to 5fs to ensure ultra-high-resolution data extraction. At the same time, the high-order mode decomposition (HOD) technology was combined to further explore potential biophysical features from the embryonic morphological features, such as the coupling effect between cell division rate and metabolic activity. After the nonlinear transformation, the extracted structure-function association features usually have high-dimensional characteristics, so they need to be optimized and reduced in dimension. The specific implementation method is to use principal component analysis (PCA) combined with manifold learning (LLE) technology to perform nonlinear dimensionality reduction on the data. In the experiment, the PCA feature selection threshold was set to 95%, that is, 95% of the information was retained to reduce information loss. At the same time, the LLE method was used to further optimize the feature space so that the morphological features can still maintain their key feature information in the low-dimensional space. The final output of this process is the optimized morphological feature representation, which can be used for subsequent classification and analysis. Finally, the optimized morphological feature representation was classified using optical interferometry to identify the characteristics of embryos at different developmental stages. Specifically, a classification algorithm based on optical Fourier transform was used to project the light field features into the frequency domain through fast Fourier transform (FFT), and a support vector machine (SVM) classifier was used for pattern recognition. During the experiment, the kernel function of SVM was set to Gaussian radial basis function (RBF), and the parameter C value was set to 10 to improve the classification accuracy. Ultimately, the interferometric measurement-based classification method can effectively distinguish high-quality embryos from low-quality embryos and output the final embryo morphology data.

[0047] The present invention realizes efficient collection of multi-angle light field information by importing embryo morphological parameter data into a silicon-based photonic interferometer chip integrated with a 128×128 microring resonator array, so that the optical characteristics of the embryo's spatial structural characteristics in different directions are fully recorded. Phase modulation processing is performed on the multi-angle embryo light field information to ensure high-resolution reconstruction of the light field data, generate embryo characteristic light phase map, and provide accurate coherent information for subsequent morphological analysis. The coherence of light is used to perform parallel optical calculations on the embryo characteristic light phase map, giving full play to the high-speed parallel computing capability of optical processing to obtain efficient and high-precision embryo morphological feature vectors. On this basis, low-loss optical waveguides are used for high-fidelity data transmission, effectively reducing the loss of morphological information during transmission and ensuring the integrity and accuracy of the data. Nonlinear transformation of high-fidelity morphological feature data is performed to deeply explore the relationship between the embryo's spatial structural information and biological functional characteristics, revealing the intrinsic relationship between morphology and function during embryonic development. Subsequently, the complexity of the data is reduced through dimensional optimization technology, the expression ability of morphological features is improved, and subsequent pattern recognition is more efficient. Finally, an interferometry-based classification method was used to analyze the optimized morphological feature representation, so that the embryo morphological data has good distinguishability based on high precision and high resolution, providing accurate data support for embryo development potential assessment and personalized reproductive medicine decision-making.

[0048] Preferably, the embryo development potential assessment in step S2 comprises: Extract the temporal features in the embryo morphology data to obtain the dynamic feature map of embryo development; Energy efficiency is calculated for the dynamic characteristic map of embryonic development, and the energy distribution pattern during cell division is used to obtain the metabolic efficiency index; According to the dynamic characteristic map of embryonic development, the critical turning points in the embryonic development process are analyzed to identify the stability and plasticity of the developmental trajectory, thereby obtaining the developmental elasticity score data; Quantitative assessment of the dynamic characteristics of embryonic development based on cell arrangement and polarization patterns to obtain morphological tissue integrity data; Based on the metabolic efficiency index, developmental elasticity score data, and morphological tissue integrity data, probabilistic reasoning was achieved through photon interferometry to obtain a preliminary prediction of implantation potential. Calibration processing is performed on the preliminary implantation potential prediction value based on the complex nonlinear relationship encoded by the coherent optical path to obtain calibrated implantation probability data; The calibrated implantation probability data were used to evaluate the embryonic developmental potential based on chromosome euploidy estimation, mitochondrial function score and epigenetic status to obtain the embryonic developmental potential data.

[0049] In this embodiment, the whole process of embryo from fertilization to blastocyst formation is first recorded by a high-resolution time-lapse imaging system, and the sampling interval is set to 10 minutes to capture the key morphological changes. Subsequently, the key timing characteristics such as cell division time point, cytoplasmic reorganization dynamics, cell symmetry and spatial arrangement during embryonic development are extracted by using a time series analysis method based on light field reconstruction. By constructing an optical time series matrix with a time resolution of not less than 100ms, the short-time Fourier transform (STFT) is used to analyze the frequency domain characteristics during development, and the multi-scale recurrent neural network (MS-RNN) is combined to deeply model the morphological change trend, and finally a dynamic feature map of embryonic development is formed, which can intuitively show the evolution pattern of embryonic morphological characteristics at different stages over time. After obtaining the dynamic feature map of embryonic development, the energy utilization during cell division is further analyzed by optical metabolic imaging technology. The specific implementation method is to use two-photon fluorescence lifetime imaging (FLIM) combined with phosphorylation and dephosphorylation characteristic ratio measurement to monitor the fluorescence lifetime changes of nicotinamide adenine dinucleotide (NADH) and flavin adenine dinucleotide (FAD) in embryonic cells to quantify cell metabolic activity. In the experiment, the excitation wavelength was set to 750nm and 920nm, and the fluorescence lifetime measurement range was set between 100ps-10ns. The energy consumption rate at different stages was calculated in combination with the fluorescence lifetime decay fitting model, and the energy distribution heat map during cell division was constructed. Finally, the metabolic efficiency index was calculated based on the time-energy integral, which can be used to evaluate the stability and efficiency of energy utilization during embryonic development. In order to identify the key turning points in embryonic development, the phase space reconstruction method was used to map the embryonic morphological evolution process into a multidimensional dynamic system to analyze the cell division rhythm, morphological change rate and developmental rhythm volatility. In the specific experiment, the delayed embedding method was used to construct the phase space trajectory, the time delay τ=30min, the embedding dimension d=5, and the fractal dimension (Fractal Dimension) and the maximum Lyapunov exponent (Lyapunov Exponent) of the developmental trajectory were calculated. If the embryonic developmental trajectory has a high fractal dimension and the Lyapunov exponent is lower than 0.1, it means that the developmental process is relatively stable, otherwise it means that the developmental trajectory has greater plasticity. The developmental elasticity score data calculated by this method can be used to evaluate the embryo's ability to adapt to changes in the external environment. The acquisition of morphological tissue integrity data relies on optical coherence tomography (OCT) combined with polarized light microscopy (PSM) to measure the embryonic cell arrangement structure and polarization direction with high precision. In the specific experiment, the axial resolution of OCT was set to 1μm and the lateral resolution was set to 5μm, and the spectral domain scanning method was used to obtain the hierarchical structure of cell arrangement; at the same time, the PSM technology evaluated the cell polarization pattern by measuring the birefringence characteristics of the cell membrane lipid bilayer.During the experiment, the cell packing density (CPD) and polarization deviation angle (PDA) of the embryos were calculated. Embryos with CPD greater than 0.85 and PDA less than 10° were considered to have high morphological and tissue integrity. The morphological and tissue integrity data finally obtained can be used to evaluate the stability of embryonic structure and the coordination of intercellular connections. After obtaining the metabolic efficiency index, developmental elasticity score data and morphological and tissue integrity data, photon interference computing technology was used for multi-factor fusion analysis to achieve a preliminary assessment of embryo implantation potential. The specific method is to use an integrated Mach-Zehnder interferometer (MZI) array to achieve high-dimensional quantum state mapping between data through phase modulation of different interference paths, and use the interference results of coherent light waves to derive probability distribution. In the experiment, the wavelength of the incident light was set to 1310nm, and a silicon-based optical waveguide chip was used for calculation. The correlation between the relative phase offset of the optical interference output and the characteristic data of the embryo was greater than 0.95, thereby ensuring the reliability of the calculation results. Finally, the initial implantation potential prediction value is obtained through coherent interference calculation, which can be used to evaluate the possibility of successful implantation of the embryo after implantation. Since the prediction of embryo implantation potential involves complex nonlinear relationships, it is necessary to calibrate the initial implantation potential prediction value. The specific implementation method is to adopt a deep learning optimization strategy based on optical path encoding, construct a photonic neural network (PNN), map the embryo feature data to a high-dimensional optical computing space, and realize nonlinear feature learning by modulating the optical interference path. During the experiment, 16-channel phased optical path modulation technology was used, and the phase shift control accuracy of each optical path was within 0.01π to achieve high-precision fitting of complex nonlinear relationships. The calibrated implantation probability data finally obtained can provide more accurate embryo implantation prediction results, reduce the misjudgment rate, and improve the reliability of clinical decision-making. On the basis of the calibrated implantation probability data, the embryo genome and metabolic characteristics are further integrated to achieve a comprehensive assessment of the embryonic development potential. The specific implementation method is to use single-cell whole genome amplification (scWGA) technology to determine the euploidy of embryonic chromosomes, and analyze the stability of embryonic karyotype by fluorescence in situ hybridization (FISH); use fluorescence resonance energy transfer (FRET) imaging technology to evaluate mitochondrial oxidative phosphorylation (OXPHOS) function and measure mitochondrial membrane potential (ΔΨm); at the same time, use single-molecule epigenetic sequencing technology to analyze DNA methylation patterns and histone modification status. In the experiment, the excitation wavelength of FRET determination was set to 405nm, the emission wavelength was set to 525nm, and the time-dependent fluorescence lifetime determination method was used to ensure the measurement accuracy.Finally, combining the above data, a photon probability calculation model based on Bayesian reasoning was used to derive embryo development potential data, which can provide an accurate basis for clinical embryo selection and improve the success rate of in vitro fertilization.

[0050] The present invention constructs a dynamic characteristic map of embryonic development by extracting the temporal features in the embryo morphological data, so that the morphological changes of the embryo at different developmental stages are systematically quantified, thereby providing accurate data support for analyzing its growth pattern. Based on the map, the energy utilization during cell division is calculated, the energy distribution pattern during cell division is identified, and then the metabolic efficiency of the embryo at different developmental stages is quantified, and a metabolic efficiency index is generated to characterize the balance and stability of embryonic energy utilization. Combined with the dynamic characteristic map, the key turning points in the embryonic development process are analyzed, the stability and plasticity of the developmental trajectory are evaluated, and finally the developmental elasticity scoring data is obtained to quantify the adaptability of the embryo under different environmental influences. Further, a quantitative analysis method based on cell arrangement and polarization pattern is adopted to evaluate the integrity of the embryonic morphological organization, so that the developmental coordination of the embryo in the spatial structure is accurately portrayed. Subsequently, the metabolic efficiency index, the developmental elasticity scoring data and the morphological organization integrity data are integrated, and probabilistic reasoning is performed using photon interference technology to generate a preliminary implantation potential prediction value, providing a preliminary quantitative index for embryo selection. On this basis, the implantation potential prediction value is further calibrated by encoding complex nonlinear relationships through coherent light paths, making the calculation of embryo development potential more robust and ensuring more reliable prediction results. Finally, combined with multidimensional biological indicators such as chromosome euploidy estimation, mitochondrial function score, and epigenetic status, the calibrated implantation probability data is deeply analyzed to comprehensively evaluate the developmental potential of the embryo, providing a high-precision reference basis for clinical decision-making, making embryo screening and personalized assisted reproductive strategies more scientific and accurate.

[0051] Preferably, step S3 comprises the following steps: Step S31: Utilize The double-layer oxide structured memristor array performs feature dimensionality reduction processing on immune cell function data and genomic variation data, thereby obtaining a high-dimensional feature compression representation, where each memristor has 64 adjustable conductance states; In this embodiment, using The double-layer oxide structured memristor array performs feature dimensionality reduction processing on immune cell function data and genomic variation data to reduce data dimensions and retain the main information. Specifically, immune cell function data include multiple parameters such as cell proliferation rate, cytokine secretion level, cytotoxicity score, etc., while genomic variation data involve single nucleotide polymorphism (SNP), copy number variation (CNV), methylation level, etc. First, the input data is standardized to ensure that the numerical range of the data is adapted to the conductivity adjustment range of the memristor. Then, the data matrix is ​​input into the memristor array, and the conductivity state of each memristor can be adjusted in 64 discrete levels, representing different feature weights. The training pulse is applied to the memristor by pulse programming, so that its conductive state is adjusted according to the target dimensionality reduction mapping, and finally a high-dimensional feature compression representation is obtained, which can effectively reduce data redundancy and improve computational efficiency.

[0052] Step S32: performing nonlinear transformation on the high-dimensional feature compression representation, and applying a pulse sequence of ±0.8V to ±1.2V to regulate the conductive state of the memristor, thereby obtaining a nonlinear feature mapping matrix; In this embodiment, the obtained high-dimensional feature compression representation is nonlinearly transformed to enhance the discriminative ability of the data. Specifically, a pulse sequence of ±0.8V to ±1.2V is first applied to the memristor array, and the change of the conductance value of the memristor is controlled by the change of the pulse amplitude and pulse width, so that it produces a nonlinear response. In this process, a nonlinear transformation method based on exponential mapping is adopted to expand the feature dimension of the input data to a more complex feature space, while enhancing the separability between data. Utilizing the programmable characteristics of the memristor, the mapping results of different data categories are adaptively adjusted to ensure that the nonlinear feature mapping matrix can correctly express the distribution characteristics of the data and provide support for subsequent clustering analysis.

[0053] Step S33: identifying natural clustering structures on the nonlinear feature mapping matrix, thereby obtaining preliminary clustering grouping data; In this embodiment, the natural clustering structure is identified on the nonlinear feature mapping matrix to obtain preliminary clustering grouping information of the data. First, a similarity matrix based on the conductive state of the memristor is constructed using the spectral clustering method, which reflects the similarity relationship between different data points. Then, a random perturbation signal is applied to the memristor array to detect the inherent data structure in the feature mapping matrix. Unsupervised learning methods such as k-means or density clustering (DBSCAN) are used to classify the change trend of the conductive state of the memristor to generate preliminary clustering grouping data. In this process, the Silhouette coefficient or Davies-Bouldin index can be used to evaluate the clustering effect to ensure the rationality of the clustering structure.

[0054] Step S34: performing feature vector similarity calculation based on the memristor network in the memristor array on the preliminary clustering grouping data, thereby obtaining immune-genetic subtype data; In this embodiment, in order to further refine the clustering results, the immune-genetic subtype data is calculated based on the feature vector similarity of the memristor array. Specifically, the representative feature vectors in each cluster grouping are first extracted, and the cosine similarity, Euclidean distance, and dynamic time warping (DTW) similarity between different groups are calculated on the memristor network to measure the cohesion and separation of the data categories. Then, the preliminary cluster grouping data is adjusted according to the calculation results so that the data points within the same subtype have a higher similarity, while the similarity between different subtypes is lower. Finally, the immune-genetic subtype data is generated to provide a basis for subsequent information flow simulation.

[0055] Step S35: simulating complex information flow according to the immune-genetic subtype data to obtain a typing stability index; In this embodiment, complex information flow is simulated according to immune-genetic subtype data to obtain a typing stability index. Specifically, the random walk method based on the memristor array is first used to simulate the dynamic information transfer process between different subtypes. Based on the adjustable conductance characteristics of the memristor, this method maps the data flow path to the topological structure of the memristor array, and calculates the information propagation efficiency and steady-state distribution. Then, the Markov chain steady-state probability calculation method is used to evaluate the stability of different subtypes, calculate the subtype internal information retention rate and the cross-subtype information exchange rate, and thus obtain the typing stability index. Subtypes with higher typing stability indexes usually have stable gene regulatory networks, while subtypes with lower indexes may be greatly affected by the external environment or genetic drift.

[0056] Step S36: identifying the key driving factors of each subtype in the immune-genetic subtype data according to the typing stability index, and ranking the importance of high-dimensional features, thereby obtaining subtype characteristic spectrum data; In this embodiment, the typing stability index is used to identify the key driving factors of each subtype in the immune-genetic subtype data, and the high-dimensional feature importance is sorted to obtain the subtype characteristic spectrum data. First, the characteristic variables with significant differences within different subtypes are extracted, and the contribution of each feature to the classification result is calculated using the SHAP (Shapley AdditiveExplanations) method based on the memristor array. Then, the feature contribution is normalized and sorted according to methods such as information gain and mutual information entropy to screen out the key driving factors. For the top-ranked characteristic variables, their distribution patterns among different subtypes are further analyzed to construct subtype characteristic spectrum data and provide a basis for subsequent typing tree construction.

[0057] Step S37: construct a hierarchical classification tree based on the subtype characteristic spectrum data and the typing stability index calculation unit to obtain the patient's immune-genetic typing data.

[0058] In this embodiment, a hierarchical classification tree is constructed according to the subtype characteristic spectrum data and the typing stability index calculation unit to obtain the patient's immune-genetic typing data. Specifically, a decision tree algorithm (Decision Tree) or a random forest method (Random Forest) is first used to construct an initial classification tree with the subtype characteristic spectrum data as the input variable and the immune-genetic subtype label as the target variable. Then, the energy optimal pruning algorithm (OptimalPruning) based on the memristor is used to remove redundant feature branches and improve the generalization ability of the classification tree. Finally, the confidence of the hierarchical classification tree is calculated in combination with the typing stability index, and the classification accuracy is evaluated by the cross-validation method to ensure the reliability and interpretability of the immune-genetic typing data.

[0059] The present invention adopts The double-layer oxide structured memristor array performs feature dimensionality reduction processing on immune cell function data and genomic variation data, so that high-dimensional data can be effectively compressed, and the accuracy of feature representation and data storage efficiency are improved through 64 adjustable conductance states. The high-dimensional features after dimensionality reduction are nonlinearly transformed, and a pulse sequence of ±0.8V to ±1.2V is applied to regulate the conduction state of the memristor, so that the feature mapping process can adapt to complex nonlinear relationships, generate a nonlinear feature mapping matrix, and enhance the separability of the data. On the basis of this matrix, by identifying the natural clustering structure and extracting the intrinsic grouping pattern of the data, preliminary clustering grouping data is obtained to improve the adaptive ability of classification. Subsequently, the feature vector similarity is calculated based on the memristor network to further optimize the clustering accuracy, obtain immune-genetic subtype data, and more accurately characterize the individual differences of different immune genetic characteristics. The immune-genetic subtype data is used to simulate complex information flow and calculate the typing stability index, thereby quantifying the robustness and evolution trend of different subtypes. The key driving factors of each subtype are further identified, and the high-dimensional features are ranked by importance, and the subtype characteristic spectrum data is extracted, so that the core characteristics of the immune genetic pattern are effectively revealed. Finally, through the comprehensive calculation of the subtype characteristic spectrum data and the typing stability index, a hierarchical classification tree is constructed to accurately divide the patient's immune-genetic typing.

[0060] Preferably, step S4 comprises the following steps: Step S41: performing a unified mathematical representation conversion on the endocrine dynamic characteristic data, the embryonic development potential data, and the patient immune-genetic typing data, thereby obtaining a multi-dimensional joint tensor representation; In this embodiment, for endocrine dynamic feature data, embryonic development potential data and patient immune-genetic typing data, the data are normalized by standardized methods, wherein the endocrine dynamic feature data uses a sliding window method to extract time series features, the embryonic development potential data constructs feature vectors based on morphological scores and metabolic dynamic parameters, and the immune-genetic typing data uses principal component analysis for dimensionality reduction. Then, the three types of data are converted into a unified mathematical representation using the Kronecker product method to construct a third-order tensor structure, each tensor dimension corresponds to time series features, genetic features and embryonic features, and finally forms a multidimensional joint tensor representation of N×M×K scale, where N is the number of samples, M is the comprehensive feature dimension, and K is the time step.

[0061] Step S42: performing cross-modal feature alignment on the multi-dimensional joint tensor representation to obtain aligned feature space data; In this embodiment, for the multi-dimensional joint tensor representation, the mutual information maximization method is used to calculate the correlation matrix between different modal features, and the low-correlation features are reduced in dimension based on the orthogonal projection method to reduce noise interference. Then, a generator-discriminator framework is constructed using a cross-modal adversarial learning network, in which the generator uses a convolutional neural network (CNN) to learn the common representation of different modal data, and the discriminator judges whether the data comes from the same modality based on a recurrent neural network (RNN), thereby optimizing the alignment effect of the feature space. Finally, the feature distribution between different modalities is constrained by the maximum mean difference (MMD) loss, so that it is mapped to the same feature space, and the aligned feature space data is obtained, whose structural dimension is the same as the multi-dimensional joint tensor representation, but the feature distribution between modalities tends to be consistent.

[0062] Step S43: constructing a heterogeneous graph neural network based on a silicon photonic chip, using the heterogeneous graph neural network to perform association modeling on the aligned feature space data to obtain multimodal association graph structure data, wherein the heterogeneous graph neural network integrates 1024 programmable microring resonators as graph nodes; In this embodiment, 1024 programmable microring resonators are integrated on a silicon photonic chip, and each resonator corresponds to a graph node for storing single sample features of aligned feature space data. Then, an optical beam splitter (BeamSplitter) and a tunable filter (Tunable Filter) are used to realize light intensity modulation, so as to build edge weights between different resonators, and thus construct the adjacency matrix of the heterogeneous graph neural network. Next, the GraphAttentionMechanism is used to calculate the information propagation weights between different nodes, and the photon matrix multiplication unit is combined to realize multi-layer heterogeneous graph convolution operations, extract nonlinear correlation features between data, and finally output multimodal correlation graph structure data, in which the embedded representation of each node contains the cross-modal features of the individual patient.

[0063] Step S44: extracting persistent homology features on the multimodal association graph structure data and identifying stable patterns across data types, thereby obtaining multi-level clinical association pattern data; In this embodiment, for multimodal association graph structure data, the persistent homology method in topological data analysis (TDA) is first used to extract the high-order topological structure of cross-modal features, specifically including the calculation of the Betti number, persistence barcode and persistence diagram in the graph to quantify the stable patterns between different modal data. Then, a feature clustering model is constructed based on the graph neural network, and the persistent homology features are divided using the spectral clustering algorithm to identify stable patterns across data types and construct multi-level clinical association pattern data, which can be used to predict the clinical risk characteristics of different patients in reproductive medicine.

[0064] Step S45: performing homomorphic encryption processing on the multi-level clinical association pattern data to obtain encrypted clinical association data; In this embodiment, in order to ensure the privacy security of multi-level clinical association model data, the Paillier homomorphic encryption algorithm is first used to encrypt the data, in which a 1024-bit key is used to encrypt the data, and the data privacy protection capability is enhanced by a random noise perturbation strategy. Then, a distributed encryption calculation method is used to implement matrix operations on encrypted data without decrypting the original data, including basic operations such as addition and multiplication. Finally, encrypted clinical association data is obtained, which can be securely shared between different institutions while ensuring the confidentiality of the data.

[0065] Step S46: constructing a knowledge graph index structure based on the encrypted clinical related data through distributed ledger technology, thereby obtaining primary knowledge graph framework data; In this embodiment, a decentralized storage architecture is built based on distributed ledger technology, and a smart contract is created under the Hyperledger Fabric framework to achieve storage and access control of encrypted clinical-related data. Then, for different clinical data sources, a knowledge graph construction method based on RDF (Resource Description Framework) is used to convert encrypted data into triples (subject-predicate-object) representation, and an index structure is established using the SPARQL query language. Finally, a primary knowledge graph framework data is formed, which contains the relationship between different data entities and supports distributed access and query.

[0066] Step S47: Deploy the temporal logic reasoning engine on the primary knowledge graph framework and establish a security access control mechanism to obtain the reproductive medicine database structure.

[0067] In this embodiment, a temporal logic reasoning engine is deployed on the primary knowledge graph framework, wherein temporal reasoning rules are first defined based on temporal extended description logic (TDL), and ontology constraint modeling is performed using OWL 2 (Web Ontology Language). Then, the causal relationship between clinical data is automatically derived through inductive reasoning methods in combination with the entity relationships in the knowledge graph. Finally, a secure access control mechanism is established based on the Zero Trust Architecture, including role-based access control (RBAC) and attribute-based access control (ABAC), to ensure the security and traceability of the data, and finally obtain a complete reproductive medicine database structure to realize intelligent storage and analysis of data.

[0068] The present invention performs a unified mathematical representation conversion on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, so that data from different sources can be described under the same mathematical framework, and the integrity and relevance of the data are ensured through a multidimensional joint tensor representation. Subsequently, cross-modal feature alignment is performed for the multidimensional joint tensor representation to eliminate the modal differences between the data, construct a unified feature space, and improve the effectiveness of information fusion. A heterogeneous graph neural network based on a silicon photonic chip is used for association modeling of aligned feature space data. The network integrates 1024 programmable microring resonators as graph nodes, so that the complex nonlinear relationship between the data can be efficiently modeled, and the potential interaction patterns between endocrine, embryonic development, and immune-genetic factors are fully explored. Persistent homology features are extracted through a multimodal association graph structure, and stable patterns across data types are identified, thereby revealing multi-level clinical association information and improving the stability and robustness of medical data analysis. Homomorphic encryption processing is implemented for multi-level clinical association pattern data to ensure the privacy and security of the data during the calculation process, so that sensitive medical information can be analyzed and calculated in an encrypted state without decryption. Distributed ledger technology is used to build a knowledge graph index structure based on encrypted clinical-related data, making data storage and management more transparent and traceable, and ensuring the integrity and tamper-proof capabilities of the knowledge graph framework. Based on the primary knowledge graph framework, a temporal logic reasoning engine is deployed to enable the system to perform temporal reasoning on dynamic medical data, and combined with a secure access control mechanism to ensure the security and controllability of data query and reasoning results, ultimately forming a reproductive medicine database structure. This database structure can integrate and manage multi-source medical data, support precision medicine analysis, and provide a solid data foundation for individualized reproductive health assessment, embryo selection, and clinical decision-making.

[0069] The present invention also provides a system for constructing a reproductive medicine database, which is used to execute the above-mentioned method for constructing a reproductive medicine database. The system for constructing a reproductive medicine database comprises: The biological data acquisition module is used to obtain biological tissue samples from patients; reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data are obtained from biological tissue samples from patients; The potential assessment module is used to perform time series analysis and processing on hormone fluctuation patterns based on reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; perform image recognition processing on embryo morphological parameter data based on the coherent superposition characteristics of light, and perform embryo development potential assessment to obtain embryo development potential data; The memristor heterogeneity typing module is used to perform hierarchical clustering processing on the individual heterogeneity of patients based on the immune cell function data and genomic variation data to obtain the patient's immune-genetic typing data, wherein the hierarchical clustering processing is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; The encrypted storage module is used to perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and to construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

[0070] The present invention systematically acquires biological tissue samples of patients through the biological data acquisition module, and accurately extracts reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data, so that personalized medical information can be fully integrated. Based on these data, the potential assessment module analyzes the timing of reproductive endocrine hormones, analyzes hormone fluctuation patterns, and constructs dynamic feature data to reveal the regulation of the endocrine system on the reproductive process. The embryo morphological parameter data is analyzed through high-precision image recognition and analysis using the coherent superposition characteristics of light to achieve quantitative assessment of embryonic development potential, ensuring that the assessment results have high stability and credibility. The memristor array with a double-layer oxide structure performs hierarchical clustering on immune cell function data and genomic variation data, identifies immune-genetic heterogeneity between individuals, and builds a refined patient typing model. The 64 adjustable conductance states in the memristor array enable it to accurately depict the complex information flow patterns between different immune-genetic subtypes, improving the resolution and adaptability of typing. At the data management level, the encryption storage module performs multi-dimensional feature fusion on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, so that the cross-correlation patterns between data can be deeply mined and stored in a homomorphic encryption manner to ensure that the data is not leaked during the calculation process. Furthermore, a knowledge graph index is constructed based on distributed ledger technology, so that the structure of the reproductive medicine database has traceability and tamper-proof capabilities, providing reliable data support for individualized reproductive health assessment. This database not only improves the systematic nature of reproductive health analysis, but also enables precision medical strategies to be implemented under data security guarantees, providing efficient and reliable intelligent auxiliary support for clinical decision-making.

[0071] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0072] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for constructing a reproductive medicine database, characterized in that: The following steps are involved: Step S1: Obtain a biological tissue sample from a patient; obtain reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data, and genome variation data from the biological tissue sample from the patient; Step S2: performing time series analysis on the hormone fluctuation pattern according to the reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; performing image recognition processing on the embryo morphological parameter data based on the coherent superposition characteristics of light, and performing embryo development potential assessment to obtain embryo development potential data; Step S3: hierarchical clustering is performed on the individual heterogeneity of patients according to the immune cell function data and genomic variation data to obtain the patient immune-genetic typing data, wherein the hierarchical clustering process is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; Step S4: Perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

2. The method for constructing a reproductive medicine database according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: obtaining a biological tissue sample of a patient; Step S12: The patient's biological tissue sample is introduced into the PDMS microfluidic chip with an integrated CdSe / ZnS core-shell structure quantum dot sensor array to obtain the dispersed flow data of the biological sample, wherein the PDMS microfluidic chip has a multi-channel structure with a channel width of 30-50 m; Step S13: modifying specific antibodies on the dispersed flow data of the biological sample on the quantum dot sensor array, and sorting and capturing NK cells, T cell subsets, B cells and monocytes in the patient's biological tissue sample, thereby obtaining immune cell capture data, wherein the modified specific antibodies include CD56, CD3, CD4, CD8, CD19, CD14 and HLA-DR; Step S14: performing multi-spectral excitation in the wavelength range of 525-655 nm on the immune cell capture data, and recording changes in quantum dot fluorescence resonance energy transfer signals, thereby obtaining immune cell function data; Step S15: Acquire reproductive endocrine hormone data, embryo morphological parameter data, and genome variation data from the patient's biological tissue sample.

3. The method for constructing a reproductive medicine database according to claim 2, characterized in that: Step S15 includes the following steps: Step S151: processing the follicular fluid in the patient's biological tissue sample by a microfluidic chip electrochemical sensor array to obtain a primary endocrine data stream, wherein the microfluidic chip electrochemical sensor array comprises 64 independent electrode units, and the surface of each unit is modified with specific aptamers for estradiol, progesterone, FSH and LH; Step S152: Use the diamond quantum sensor based on nitrogen vacancy defects prepared by 28nm process to detect the hormone concentration with pT level accuracy on the primary endocrine data stream, so as to obtain a high-precision endocrine concentration matrix, wherein the detection accuracy of the diamond quantum sensor reaches Molar concentration, sampling frequency is 300Hz; Step S153: extracting pulse signals from the high-precision endocrine concentration matrix to obtain hormone secretion pulse characteristic data; Step S154: combining the hormone secretion pulse characteristic data with the time series mark to obtain reproductive endocrine hormone data; Step S155: continuously monitoring the patient's biological tissue sample through an integrated microfluidic culture chamber and a phase contrast microscopy imaging system to obtain an embryo dynamic development image sequence, wherein the phase contrast microscopy imaging system is composed of a frequency-adjustable optical resonant cavity and a high-speed CMOS image sensor; Step S156: performing optical coherence processing on the embryo dynamic development image sequence to obtain a multi-parameter morphological feature set of the embryo, wherein the optical coherence processing is implemented using a photon interference calculation unit including a 256×256 array of integrated microring resonators; Step S157: performing optical calculation processing on the embryo multi-parameter morphological feature set to obtain embryo morphological parameter data, wherein the embryo morphological parameter data includes cleavage rate, cleavage symmetry, blastocyst cavity formation dynamics and chromosome euploidy prediction value; Step S158: Perform reproductive-related variation site analysis on the nucleic acid extracted from the patient's biological tissue sample to obtain genomic variation data.

4. The method for constructing a reproductive medicine database according to claim 3, characterized in that: Step S158 includes the following steps: Step S1581: performing single-molecule sequencing based on a graphene nanopore array on nucleic acids extracted from a patient's biological tissue sample, thereby obtaining a DNA passing signal with a single-base resolution; Step S1582: The DNA passing signals are collected and processed in parallel by a multi-channel graphene transistor array, thereby obtaining original genome sequence data, wherein the sensitivity of the multi-channel graphene transistor array is sufficient to detect the current change caused by the passage of a single DNA molecule; Step S1583: performing base accuracy verification on the original genome read sequence to obtain high-precision genome sequencing data; Step S1584: Use the memristor neuromorphic computing unit to perform reproductive-related variation site analysis on the high-precision genome sequencing data to obtain genome variation data, wherein the genome variation data includes single nucleotide polymorphism, copy number variation and structural variation information of reproductive-related genes.

5. The method for constructing a reproductive medicine database according to claim 1, characterized in that: The time series analysis processing of the hormone fluctuation pattern described in step S2 includes: The reproductive endocrine hormone data were normalized on the time axis, and a 400-hour unified time window was constructed with the first day of the menstrual cycle as the reference point to obtain standardized hormone data; The standardized hormone data were sampled at the femtosecond level with a time resolution of 10 femtoseconds to obtain a high time resolution hormone fluctuation curve; Extract intrinsic mode function sets from high time resolution hormone fluctuation curves; Correlation analysis was performed on the intrinsic mode function set to obtain the synergistic effect data between hormones; Extract nonlinear dynamic features from the synergistic effect data between hormones; The nonlinear dynamic characteristics were systematically identified to obtain the hormone regulation dynamics model; The hormone regulation dynamics model is simulated through forward evolution to obtain endocrine dynamic characteristic data.

6. The method for constructing a reproductive medicine database according to claim 1, characterized in that: The image recognition process based on the coherent superposition characteristic of light described in step S2 includes: The embryo morphological parameter data were imported into a silicon-based photonic interferometer chip integrated with a 128×128 microring resonator array to obtain multi-angle embryo light field information; Perform phase modulation processing on multi-angle embryo light field information to obtain embryo characteristic light phase map; The coherence of light is used to perform parallel optical calculation on the characteristic light phase image of the embryo to obtain the embryo morphological characteristic vector; The embryo morphological feature vector is transmitted at low loss to obtain high-fidelity morphological feature data, where the waveguide loss of the low-loss transmission is less than 0.01dB / cm; Perform nonlinear transformation on high-fidelity morphological feature data to obtain embryonic structure-function correlation features; Dimensional optimization of embryonic structure-function association features to obtain optimized morphological feature representation; The optimized morphological feature representation is subjected to interferometric measurement-based classification processing to obtain embryo morphological data.

7. The method for constructing a reproductive medicine database according to claim 6, characterized in that: The embryonic developmental potential assessment described in step S2 includes: Extract the temporal features in the embryo morphology data to obtain the dynamic feature map of embryo development; Energy efficiency is calculated for the dynamic characteristic map of embryonic development, and the energy distribution pattern during cell division is used to obtain the metabolic efficiency index; According to the dynamic characteristic map of embryonic development, the critical turning points in the embryonic development process are analyzed to identify the stability and plasticity of the developmental trajectory, thereby obtaining the developmental elasticity score data; Quantitative assessment of the dynamic characteristics of embryonic development based on cell arrangement and polarization patterns to obtain morphological tissue integrity data; Based on the metabolic efficiency index, developmental elasticity score data, and morphological tissue integrity data, probabilistic reasoning was achieved through photon interferometry to obtain a preliminary prediction of implantation potential. Calibration processing is performed on the preliminary implantation potential prediction value based on the complex nonlinear relationship encoded by the coherent optical path to obtain calibrated implantation probability data; The calibrated implantation probability data were used to evaluate the embryonic developmental potential based on chromosome euploidy estimation, mitochondrial function score and epigenetic status to obtain the embryonic developmental potential data.

8. The method for constructing a reproductive medicine database according to claim 7, characterized in that: Step S3 includes the following steps: Step S31: Utilize The double-layer oxide structured memristor array performs feature dimensionality reduction processing on immune cell function data and genomic variation data, thereby obtaining a high-dimensional feature compression representation, where each memristor has 64 adjustable conductance states; Step S32: performing nonlinear transformation on the high-dimensional feature compression representation, and applying a pulse sequence of ±0.8V to ±1.2V to regulate the conductive state of the memristor, thereby obtaining a nonlinear feature mapping matrix; Step S33: identifying natural clustering structures on the nonlinear feature mapping matrix, thereby obtaining preliminary clustering grouping data; Step S34: performing feature vector similarity calculation based on the memristor network in the memristor array on the preliminary clustering grouping data, thereby obtaining immune-genetic subtype data; Step S35: simulating complex information flow according to the immune-genetic subtype data to obtain a typing stability index; Step S36: identifying the key driving factors of each subtype in the immune-genetic subtype data according to the typing stability index, and ranking the importance of high-dimensional features, thereby obtaining subtype characteristic spectrum data; Step S37: construct a hierarchical classification tree based on the subtype characteristic spectrum data and the typing stability index calculation unit to obtain the patient's immune-genetic typing data.

9. The method for constructing a reproductive medicine database according to claim 8, characterized in that: Step S4 includes the following steps: Step S41: performing a unified mathematical representation conversion on the endocrine dynamic characteristic data, the embryonic development potential data, and the patient immune-genetic typing data, thereby obtaining a multi-dimensional joint tensor representation; Step S42: performing cross-modal feature alignment on the multi-dimensional joint tensor representation to obtain aligned feature space data; Step S43: constructing a heterogeneous graph neural network based on a silicon photonic chip, using the heterogeneous graph neural network to perform association modeling on the aligned feature space data to obtain multimodal association graph structure data, wherein the heterogeneous graph neural network integrates 1024 programmable microring resonators as graph nodes; Step S44: extracting persistent homology features on the multimodal association graph structure data and identifying stable patterns across data types, thereby obtaining multi-level clinical association pattern data; Step S45: performing homomorphic encryption processing on the multi-level clinical association pattern data to obtain encrypted clinical association data; Step S46: constructing a knowledge graph index structure based on the encrypted clinical related data through distributed ledger technology, thereby obtaining primary knowledge graph framework data; Step S47: Deploy the temporal logic reasoning engine on the primary knowledge graph framework and establish a security access control mechanism to obtain the reproductive medicine database structure.

10. A system for constructing a reproductive medicine database, characterized in that: Used to execute the method for constructing a reproductive medicine database according to claim 1, the system for constructing a reproductive medicine database comprises: The biological data acquisition module is used to obtain biological tissue samples from patients; reproductive endocrine hormone data, embryo morphological parameter data, immune cell function data and genome variation data are obtained from biological tissue samples from patients; The potential assessment module is used to perform time series analysis and processing on hormone fluctuation patterns based on reproductive endocrine hormone data to obtain endocrine dynamic characteristic data; perform image recognition processing on embryo morphological parameter data based on the coherent superposition characteristics of light, and perform embryo development potential assessment to obtain embryo development potential data; The memristor heterogeneity typing module is used to perform hierarchical clustering processing on the individual heterogeneity of patients based on the immune cell function data and genomic variation data to obtain the patient's immune-genetic typing data, wherein the hierarchical clustering processing is performed by A memristor array with a double-layer oxide structure is realized, and each memristor has 64 adjustable conductance states; The encrypted storage module is used to perform multi-dimensional feature fusion processing on endocrine dynamic feature data, embryonic development potential data, and patient immune-genetic typing data, and to construct a homomorphically encrypted distributed knowledge graph index to obtain the reproductive medicine database structure.

Images