Embryo development quality evaluation method and device, computer equipment and storage medium

Abstract

The invention relates to an embryo quality evaluation method and device, computer equipment and a storage medium. The method is applied to a processor and comprises the following steps: acquiring Raman spectrum data; wherein the Raman spectrum data is obtained by performing Raman spectrum detection on a culture solution of an embryo; inputting the Raman spectrum data into an embryo quality evaluation model, and outputting a transplantation outcome corresponding to the development quality category of the embryo; wherein the embryo quality evaluation model is obtained by training an initial embryo quality evaluation model based on a corresponding relationship between sample Raman spectrum data and a transplantation outcome corresponding to a development quality category. According to the method, the development quality and transplantation outcome of the embryo can be efficiently and accurately predicted.

Description

Technical Field

[0001] The present application relates to the technical field of bioinformatics, and in particular to a method, apparatus, computer equipment, storage medium and computer program product for evaluating embryo development quality. Background Art

[0002] With the increasing infertility rate among couples of childbearing age, the demand for assisted reproductive technology for infertile patients is growing. In Vitro Fertilization and Embryo Transfer (IVF-ET) is an effective assisted reproductive technology. It artificially removes eggs from the female body and combines them with sperm obtained in vitro to complete fertilization. The early embryos are cultured in in vitro culture medium and transplanted into the female uterus to implant and achieve pregnancy. Assessing the quality of the embryos and selecting high-quality embryos for transplantation can improve the transplant pregnancy rate.

[0003] In the related technology, a method for detecting embryo culture fluid using gas chromatography-mass spectrometry is used to evaluate the quality and differences of human in vitro fertilization embryos and predict the quality and development of embryos. In another related technology, the collected embryo culture fluid samples are quickly dehydrated and dried, and the embryo quality is evaluated based on the infrared spectrum of the solid sample. However, high-efficiency gas and liquid chromatography-mass spectrometry detection technology requires sample pretreatment in advance, and the detection process is time-consuming, costly, and complicated to operate. Near-infrared absorption spectroscopy detection of liquid samples has great interference with water molecules, affecting the detection results. Summary of the invention

[0004] Based on this, it is necessary to provide an evaluation method, device, computer equipment, storage medium and computer program product that can accurately evaluate the quality of embryo development in response to the above technical problems.

[0005] In a first aspect, the present application provides a method for evaluating embryo development quality, which is applied to a processor, and the method comprises:

[0006] Acquiring Raman spectral data; wherein the Raman spectral data is obtained by performing Raman spectral detection on the embryo culture medium;

[0007] The Raman spectral data are input into an embryo quality assessment model, and the embryo development quality category corresponding to the transplantation outcome is output; wherein the embryo quality assessment model is obtained by training an initial embryo quality assessment model based on the correspondence between the sample Raman spectral data and the development quality category corresponding to the transplantation outcome.

[0008] In a specific embodiment, the culture medium includes a dried crystalline culture medium and / or a liquid culture medium, and the obtaining of Raman spectroscopy data includes:

[0009] The dried crystalline culture solution is placed in an aluminum-coated silica wafer container, and the dried crystalline culture solution is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectrum data;

[0010] And / or, the liquid culture solution is placed in a gold-plated copper-nickel alloy container, and the liquid culture solution is irradiated with an optical fiber Raman spectrometer to obtain Raman spectrum data.

[0011] In a specific embodiment, the dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectrum data, including:

[0012] The dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer, and the irradiation parameters include: selecting an incident excitation power of 4-6 mW, selecting an exposure time of 3-5 s, accumulating n times, and averaging the Raman spectrum data of multiple sampling points to obtain Raman spectrum data.

[0013] In a specific embodiment, the liquid culture fluid is irradiated with a fiber optic Raman spectrometer to obtain Raman spectrum data, including:

[0014] The liquid culture medium is irradiated with a fiber optic Raman spectrometer, and the irradiation parameters include: the incident excitation power is selected to be 300-350mW, the exposure time is selected to be 30-50s, and the Raman spectrum data of multiple sampling points are averaged for m times to obtain Raman spectrum data.

[0015] In a specific embodiment, before acquiring Raman spectrum data, the method further includes:

[0016] Obtaining the initial culture medium containing embryos at the cleavage stage;

[0017] Mixing and centrifuging the initial culture solution to obtain a treated culture solution;

[0018] The supernatant of the treated culture medium is extracted as the culture medium for the embryos.

[0019] In a specific embodiment, the embryo quality assessment model is obtained by training the initial embryo quality assessment model based on the correspondence between the sample Raman spectral data and the transplantation outcomes corresponding to the developmental quality categories, and includes:

[0020] Acquire a sample set, wherein the sample set includes Raman spectral data samples, and the Raman spectral data samples are marked with developmental quality categories corresponding to transplant outcomes;

[0021] Inputting the Raman spectroscopy data sample into an initial embryo quality assessment model and outputting a prediction result;

[0022] Based on the difference between the prediction result and the transplantation outcome corresponding to the labeled developmental quality category, the training parameters in the initial embryo quality assessment model are iteratively adjusted to obtain the embryo quality assessment model.

[0023] In a specific embodiment, the developmental quality category corresponding to the transplantation outcome includes at least one of the following:

[0024] Live birth, chemical pregnancy or early miscarriage, no pregnancy potential.

[0025] In a specific embodiment, the obtaining of Raman spectroscopy data includes:

[0026] Acquiring initial Raman spectrum data;

[0027] The initial Raman spectrum data is preprocessed to obtain Raman spectrum data; wherein the preprocessing operation includes at least one of the following: quality control, peak removal, spectrum truncation, smoothing filtering, baseline correction and normalization.

[0028] In a specific embodiment, the initial Raman spectrum data is preprocessed to obtain the Raman spectrum data, including:

[0029] Performing a preprocessing operation on the initial Raman spectrum data to obtain intermediate Raman spectrum data;

[0030] Based on the principal component analysis algorithm, the intermediate Raman spectrum data is subjected to dimensionality reduction processing to obtain Raman spectrum data.

[0031] In a second aspect, the present application also provides a device for evaluating embryo development quality, the device comprising:

[0032] An acquisition module is used to acquire Raman spectrum data; wherein the Raman spectrum data is obtained by performing Raman spectrum detection on the embryo culture medium;

[0033] An evaluation module is used to input the Raman spectral data into an embryo quality evaluation model and output the embryo's developmental quality category corresponding to the transplant outcome; wherein the embryo quality evaluation model is obtained by training an initial embryo quality evaluation model based on the correspondence between the sample Raman spectral data and the developmental quality category corresponding to the transplant outcome.

[0034] In a specific embodiment, the acquisition module is further used for:

[0035] placing the dried crystalline culture fluid in a container, and irradiating the dried crystalline culture fluid with a microscopic confocal Raman spectrometer to obtain Raman spectrum data;

[0036] And / or, the liquid culture solution is placed in a container, and an optical fiber Raman spectrometer is used to irradiate the liquid culture solution to obtain Raman spectrum data.

[0037] In a specific embodiment, the acquisition module is further used for:

[0038] The dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer, and the irradiation parameters include: selecting an incident excitation power of 4-6 mW, selecting an exposure time of 3-5 s, accumulating n times, and averaging the Raman spectrum data of multiple sampling points to obtain Raman spectrum data.

[0039] In a specific embodiment, the acquisition module is further used for:

[0040] The liquid culture medium is irradiated with a fiber optic Raman spectrometer, and the irradiation parameters include: the incident excitation power is selected to be 300-350mW, the exposure time is selected to be 30-50s, and the Raman spectrum data of multiple sampling points are averaged for m times to obtain Raman spectrum data.

[0041] In a specific embodiment, the device further includes a generating module, wherein the generating module is configured to:

[0042] Obtaining the initial culture medium containing embryos at the cleavage stage;

[0043] Mixing and centrifuging the initial culture solution to obtain a treated culture solution;

[0044] The supernatant of the treated culture medium is extracted as the culture medium for the embryos.

[0045] In a specific embodiment, the device further comprises a training module, wherein the training module is used to:

[0046] Acquire a sample set, wherein the sample set includes Raman spectral data samples, and the Raman spectral data samples are marked with developmental quality categories corresponding to transplant outcomes;

[0047] Inputting the Raman spectroscopy data sample into an initial embryo quality assessment model and outputting a prediction result;

[0048] Based on the difference between the prediction result and the transplantation outcome corresponding to the labeled developmental quality category, the training parameters in the initial embryo quality assessment model are iteratively adjusted to obtain the embryo quality assessment model.

[0049] In a specific embodiment, the developmental quality category corresponding to the transplantation outcome includes at least one of the following:

[0050] Live birth, chemical pregnancy or early miscarriage, no pregnancy potential.

[0051] In a specific embodiment, the acquisition module is further used for:

[0052] Acquiring initial Raman spectrum data;

[0053] The initial Raman spectrum data is preprocessed to obtain Raman spectrum data; wherein the preprocessing operation includes at least one of the following: quality control, peak removal, spectrum truncation, smoothing filtering, baseline correction and normalization.

[0054] In a specific embodiment, the acquisition module is further used for:

[0055] Performing a preprocessing operation on the initial Raman spectrum data to obtain intermediate Raman spectrum data;

[0056] Based on the principal component analysis algorithm, the intermediate Raman spectrum data is subjected to dimensionality reduction processing to obtain Raman spectrum data.

[0057] In a third aspect, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, an embodiment of the method described in any one of the embodiments of the present disclosure is implemented.

[0058] In a fourth aspect, the present application further provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, an embodiment of the method described in any one of the embodiments of the present disclosure is implemented.

[0059] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product includes a computer program, and when the computer program is executed by a processor, an embodiment of the method described in any one of the embodiments of the present disclosure is implemented.

[0060] The above-mentioned embryo development quality assessment method, device, computer equipment, storage medium and computer program product, by obtaining Raman spectral data of embryo culture fluid, wherein the Raman spectral data records the scattering displacement and intensity of the change in the scattered light frequency, can reflect the property information of the material molecules in the culture fluid. In this embodiment, the material molecules of the culture fluid are mainly metabolites, and metabolites are important indicators for assessing the quality of cell development. Therefore, based on the correspondence between the sample Raman spectral data and the development quality category, an embryo quality assessment model is constructed, and the embryo quality assessment model can efficiently and accurately predict the development quality and transplantation outcome of the embryo. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] FIG1 is a diagram showing an application environment of a method for evaluating embryo development quality in one embodiment;

[0062] Figure 2 A schematic diagram of a process for evaluating embryo development quality in one embodiment;

[0063] Figure 3 is a schematic flow chart of a method for evaluating embryo development quality in another embodiment;

[0064] Figure 4 A schematic diagram of the result of preprocessing initial Raman spectrum data in one embodiment;

[0065] Figure 5 Schematic diagram of the structure of a convolutional neural network in one embodiment;

[0066] Figure 6 is a structural block diagram of an apparatus for evaluating embryo development quality in one embodiment;

[0067] FIG7 is a diagram showing the internal structure of a computer device in one embodiment;

[0068] FIG. 8 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0070] In order to facilitate those skilled in the art to understand the technical solution provided by the embodiments of the present disclosure, the technical environment in which the technical solution is implemented is described below.

[0071] At present, the main methods for evaluating embryo quality are morphological evaluation, dynamic time-lapse imaging analysis, and preimplantation genetic testing. Morphological evaluation generally involves embryologists observing the morphological characteristics, structure, and development rate of embryos at a specific developmental stage and giving scores. Dynamic time-lapse imaging analysis technology can continuously obtain images of early embryonic development, providing a comprehensive and detailed developmental picture. It does not require the embryo to be removed from the incubator, avoiding the interference of repeated human observation on embryonic development, but the equipment is expensive and there are currently no clear and stable clinical results. Preimplantation genetic testing (PGT) usually biopsies the blastocyst cytotrophoblast on the 5th to 6th day of embryonic development, and uses genomic analysis to assess whether there are genetic diseases caused by gene defects or chromosomal abnormalities in embryonic cells, thereby achieving the purpose of eugenics. However, mosaic embryos may cause misjudgment of the test results, and biopsy is invasive, complicated and time-consuming, and improper operation may cause potential damage to the embryo. The above methods actually have technical drawbacks.

[0072] Based on actual technical needs similar to those described above, the present application provides a method, apparatus, computer device, storage medium and computer program product for evaluating embryo development quality.

[0073] The embryo quality assessment method provided in the embodiment of the present application can be applied in the application environment as shown in Figure 1. Among them, the terminal 102 communicates with the server 104 via the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other servers. Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0074] In one embodiment, Figure 2 As shown, a method for evaluating embryo development quality is provided, which is described by taking the method applied to the server or terminal in FIG. 1 as an example, and includes the following steps:

[0075] Step S201, obtaining Raman spectral data; wherein the Raman spectral data is obtained by performing Raman spectral detection on embryo culture medium.

[0076] In IVF-ET technology, after the fertilization operation is completed on the day of egg retrieval, the fertilized egg is transferred to a culture dish containing a culture medium. In an exemplary embodiment, the culture medium corresponding to the embryo of the cleavage stage of the fertilized egg that develops normally to the third day can be selected. In another exemplary embodiment, the culture medium corresponding to the embryo of the blastocyst stage of the fertilized egg that develops normally to the fifth to sixth day can also be selected. It should be noted that the extraction of non-culture substances such as mineral oil should be avoided as much as possible.

[0077] Among them, Raman spectroscopy is a scattering spectrum based on molecular vibration. Specifically, based on the interaction between incident excitation photons and material molecules in the culture medium, energy exchange occurs between material molecules and excitation photons, molecular bond vibrations and generated frequency scattered photons comply with the law of conservation of energy, and recording the scattering displacement and intensity of the change in scattered light frequency can reflect the property information of material molecules in the culture medium.

[0078] Step S203, inputting the Raman spectrum data into an embryo quality assessment model, and outputting the embryo development quality category corresponding to the transplantation outcome.

[0079] The embryo quality assessment model is obtained by training an initial embryo quality assessment model based on the correspondence between sample Raman spectral data and developmental quality categories.

[0080] Specifically, the initial embryo quality assessment model can include a convolutional neural network model (CNN), a long short-term memory model (LSTM), a transformer network model (Transformer), a support vector machine (SVM), a random forest (RF), an extreme gradient boosting (XGBoost), etc.

[0081] Among them, the transplantation outcomes corresponding to the developmental quality categories can be set according to specific application scenarios, such as live birth, biochemical pregnancy or early miscarriage, and lack of pregnancy potential.

[0082] Specifically, live birth means that the patient can finally give birth to a healthy baby. In the specific implementation process, human chorionic gonadotropin (HCG) is usually tested on the 14th day after embryo transfer. If the HCG concentration exceeds 25IU / L, it is a positive biochemical pregnancy. Six weeks after embryo transfer, the presence of a gestational sac and fetal heartbeat in the uterus is confirmed by B-ultrasound examination, and clinical pregnancy is successful at this time. Pregnancy exceeding 12 weeks is considered to be ongoing pregnancy. Delivery at 28 weeks but less than 37 weeks of pregnancy is premature delivery, and delivery at 37 weeks but less than 42 weeks of pregnancy is called full-term delivery. Both full-term delivery and premature delivery are live birth outcomes. Biochemical pregnancy or early miscarriage can be a positive HCG test result of the patient 14 days after implantation, no gestational sac and fetal heartbeat is seen on B-ultrasound at 6 weeks, or miscarriage before 12 weeks of pregnancy. The clinical outcome after implantation of embryos without pregnancy potential can be a negative HCG test result of the patient 14 days after implantation.

[0083] In a specific implementation process, the embryo quality assessment model is obtained by training the initial embryo quality assessment model based on the correspondence between the sample Raman spectrum data and the development quality category corresponding to the transplantation outcome, including: obtaining a sample set, the sample set including a Raman spectrum data sample, and the Raman spectrum data sample is marked with a development quality category. The Raman spectrum data sample is input into the initial embryo quality assessment model, and a prediction result is output. Based on the difference between the prediction result and the marked development quality category, the training parameters in the initial embryo quality assessment model are iteratively adjusted to obtain an embryo quality assessment model.

[0084] In the above embodiment, by obtaining the Raman spectral data of the embryo culture medium, wherein the Raman spectral data records the scattering displacement and intensity of the change in the scattered light frequency, it is possible to reflect the property information of the material molecules in the culture medium. In this embodiment, the material molecules of the culture medium are mainly metabolites, and metabolites are important indicators for evaluating the quality of cell development. Therefore, based on the correspondence between the sample Raman spectral data and the development quality category, an embryo quality assessment model is constructed, and the embryo quality assessment model can efficiently and accurately predict the development quality and transplantation outcome of the embryo.

[0085] In one embodiment, the culture fluid includes a dried crystalline culture fluid and / or a liquid culture fluid, and the acquiring of Raman spectroscopy data includes:

[0086] The dried crystalline culture fluid is placed in a container, and the dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectrum data.

[0087] And / or, the liquid culture solution is placed in a container, and an optical fiber Raman spectrometer is used to irradiate the liquid culture solution to obtain Raman spectrum data.

[0088] Specifically, the original sample can be directly tested without treatment or the crystalline sample after drying can be tested, and the Raman spectral data obtained can be used to determine the prediction result alone or comprehensively. In an exemplary embodiment, the dried crystalline culture solution can be dropped into an aluminum-coated silica wafer container to better obtain the material detection signal, and a microscopic confocal Raman spectrometer can be used, for example, with a laser wavelength of 532nm, to irradiate the dried crystalline culture solution. Among them, the dried crystalline culture solution can be obtained by: using a constant temperature blast drying oven set at 35°C, drying the original culture solution for 25 to 30 minutes to completely dry it into a crystalline state. The sample volume required for a single test can be 2μL.

[0089] In another exemplary embodiment, the liquid culture solution can also be dropped into a gold-plated copper-nickel alloy container. Similarly, the material detection signal can be better obtained, and a fiber Raman spectrometer, for example, with a laser wavelength of 785nm, is used to irradiate the liquid culture solution to obtain Raman spectrum data. The sample volume required for a single test can be 7μL.

[0090] In the above embodiment, the microscopic confocal Raman spectrometer is used to detect the Raman signal of the dried crystalline culture medium, and the fiber optic Raman spectrometer is used to detect the Raman signal of the original liquid culture medium. The signal detected by the microscopic confocal Raman spectrometer is stronger, and it has a stronger ability to identify the molecular species in the culture medium; the fiber optic Raman spectrometer detects the original sample, and when the concentration of the substance is low, the signal is weak, but it can measure the relative concentration information between different molecules, and can avoid the interference of fluorescence signals. Through two different ways to complement each other in terms of molecular species and concentration, it can obtain more accurate and comprehensive retention of metabolic information and obtain accurate embryo quality assessment results.

[0091] In one embodiment, the dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectral data, including:

[0092] The dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer, and the irradiation parameters include: selecting an incident excitation power of 4-6 mW, selecting an exposure time of 3-5 s, accumulating n times, and averaging the Raman spectrum data of multiple sampling points to obtain Raman spectrum data.

[0093] Specifically, the microscope confocal Raman spectrometer can select a microscope objective of ×50 (NA=0.9), a grating of 1200l / mm, and an incident excitation power of 4-6mW, such as 4mW, 5mW and 6mW. The exposure time is selected to be 3-5s, such as 3s, 4s and 5s. The cumulative n times includes the cumulative exposure times n times, such as 15 times. The Raman spectrum data of multiple sampling points are averaged, for example, the Raman spectrum data of 5 samples are averaged to obtain the Raman spectrum data. It should be noted that the above parameters can ensure the quality of the Raman spectrum data, that is, the peak spectrum signal is stable and has good repeatability, which is conducive to accurately predicting the development quality of the embryo. Those skilled in the art can also make adaptive parameter changes in specific application scenarios, and this application does not limit this.

[0094] Optionally, before each sample test, the center peak of the silicon wafer is 520.5cm -1 The instrument is calibrated at the position, and the deviation is required to be ±0.3cm -1 .

[0095] In one embodiment, the liquid culture fluid is irradiated with a fiber optic Raman spectrometer to obtain Raman spectrum data, including:

[0096] The liquid culture medium is irradiated with a fiber optic Raman spectrometer, and the irradiation parameters include: the incident excitation power is selected to be 300-350mW, the exposure time is selected to be 30-50s, and the Raman spectrum data of multiple sampling points are averaged for m times to obtain Raman spectrum data.

[0097] Specifically, the fiber Raman spectrometer can select an incident wavelength of 785nm, and the incident excitation power can be selected from 300-350mW, such as 300mW, 320mW, and 350mW. The exposure time can be selected from 30-50s, such as 30s, 40s, and 50s, and the cumulative m times include the cumulative exposure times n times, such as 3 times. The Raman spectrum data of multiple sampling points are averaged. It should be noted that the above parameters can guarantee the quality of the Raman spectrum data, that is, the peak spectrum signal is stable and has good repeatability, which is conducive to accurately predicting the development quality of the embryo. Those skilled in the art can also make adaptive parameter changes in specific application scenarios, and this application does not limit this.

[0098] Optionally, before each sample test, the center peak of the silicon wafer is 520.5cm -1 The instrument is calibrated at the position, and the deviation is required to be ±0.3cm -1 .

[0099] In one embodiment, before acquiring Raman spectrum data, the method further includes:

[0100] Obtaining the initial culture medium containing embryos at the cleavage stage;

[0101] Mixing and centrifuging the initial culture solution to obtain a treated culture solution;

[0102] The supernatant of the treated culture medium is extracted as the culture medium for the embryos.

[0103] Specifically, the fertilized egg is transferred to a culture dish, or optionally, to a micro-droplet culture dish. The culture dish is placed in a cleavage stage when the fertilized egg develops to the third day.

[0104] Among them, the cleavage stage refers to the embryos that normally develop to the third day of embryonic cells in the culture cycle of clinical in vitro fertilization embryonic cells, which are called cleavage stage embryos. In the specific implementation process, after the fertilization operation is completed on the day of egg retrieval, the fertilized eggs can be transferred to a 25μL / drop micro-droplet culture dish, which can be placed in a 37°C, 6% CO2 incubator and cultured continuously until the third day of embryonic age when the cleavage stage is transplanted. After the transfer, the collection culture dish can contain 20μL and can be placed in a -80°C environment for cryopreservation.

[0105] In an exemplary embodiment, the culture medium is taken out from the -80°C environment, thawed at room temperature of 20-25°C for 20-30 minutes, mixed after thawing, for example, vortexed for 60 seconds, and then centrifuged, for example, at a centrifuge with a centrifugal parameter of 5000 rpm for about 30 seconds, and the supernatant is taken as the culture medium for the embryo. Optionally, 1.5 μL of the supernatant is dropped on an aluminized silica substrate and dried in a constant temperature forced air drying oven at 35°C for 25-30 minutes to prepare a crystal sample.

[0106] In the above embodiment, by mixing and centrifuging the initial culture medium, the molecular distribution in the culture medium can be made more uniform, thereby facilitating the accuracy of subsequent development quality assessment.

[0107] In one embodiment, reference Figure 3 As shown, the embryo quality assessment model is obtained by training the initial embryo quality assessment model based on the correspondence between the sample Raman spectrum data and the development quality category, including:

[0108] Step S301: Acquire a sample set, wherein the sample set includes Raman spectrum data samples, and the Raman spectrum data samples are marked with development quality categories.

[0109] Specifically, the sample set includes Raman spectroscopy data samples, and the Raman spectroscopy data samples are annotated with developmental quality categories corresponding to transplantation outcomes, such as live birth, biochemical pregnancy or early miscarriage, and no pregnancy potential. Among them, the live birth group can include 100 samples, and the biochemical pregnancy group and the early miscarriage group can include 100 samples. In an exemplary embodiment, the Raman spectroscopy data samples in the sample set can be used to establish a training set and a test set in a ratio of 4:1.

[0110] During the sample collection process, the culture medium of fertilized egg embryos from older pregnant women, volunteers with decreased ovarian reserve function or repeated test tube transplantation failures can be selected to increase the number of positive samples.

[0111] Step S303: input the Raman spectrum data sample into the initial embryo quality assessment model and output the prediction result.

[0112] In the specific implementation process, the Raman spectrum data samples can be directly input into the initial embryo quality assessment model to output the prediction results. In another exemplary embodiment, the Raman spectrum data samples can also be subjected to principal component analysis and information screening.

[0113] In the disclosed embodiment, the basic network of the initial embryo quality assessment model can be a convolutional neural network. Figure 5As shown in the figure, the convolutional neural network includes a data input layer (Input layer), a convolution calculation layer (CONV layer), a pooling layer (Pooling layer), an activation function layer (ReLU layer) and a fully connected layer (FC layer). The data input layer processes multiple input spectral data. The core of the convolutional neural network is the convolution calculation layer. The two key operations of the convolution calculation layer are local correlation and window sliding, which extract features from the input layer input data and perform convolution operations. The pooling layer is sandwiched between continuous convolution calculation layers to compress the amount of data and parameters and reduce overfitting. The activation function layer performs nonlinear mapping on the output results of the convolution calculation layer. After multiple rounds of convolution calculation layers and pooling layers, all features are generally connected by 1 to 2 fully connected layers to give the final classification results.

[0114] Step S305, based on the difference between the prediction result and the labeled development quality category, iteratively adjust the training parameters in the initial embryo quality assessment model to obtain an embryo quality assessment model.

[0115] Specifically, in an exemplary embodiment, after the training parameters of the initial embryo quality assessment model are adjusted to a preset number of times, the embryo quality assessment model is obtained. In another exemplary embodiment, the training parameters of the initial embryo quality assessment model are iteratively adjusted until the accuracy of the model meets the preset requirements, thereby obtaining the embryo quality assessment model.

[0116] Among them, the accuracy assessment can be achieved by selecting the receiver operating characteristic curve (ROC curve). Specifically, accuracy refers to the proportion of correct results predicted by the classification model in the whole. It is created by plotting the true positive rate (TP) and the false positive rate (FP) under various threshold settings, and is the most core indicator of the performance of the prediction model. An important feature of the ROC curve is its area under the curve (AUC). AUC=0.5 is random classification and the recognition ability is 0. The closer the area is to 1, the stronger the recognition ability is. The area equals 1 for complete recognition. The performance of the PC-CNN classification model is evaluated using prediction accuracy, ROC and AUC, and compared with several other classification models.

[0117] In one embodiment, the developmental quality category corresponding to the transplantation outcome includes at least one of the following:

[0118] Live birth, chemical pregnancy or early miscarriage, no pregnancy potential.

[0119] Specifically, live birth means that the patient can finally give birth to a healthy baby. In the specific implementation process, human chorionic gonadotropin (HCG) is usually tested on the 14th day after embryo transfer. If the HCG concentration exceeds 25IU / L, it is a positive biochemical pregnancy. Six weeks after embryo transfer, the presence of a gestational sac and fetal heartbeat in the uterus is confirmed by B-ultrasound examination, and clinical pregnancy is successful at this time. Pregnancy exceeding 12 weeks is considered to be ongoing pregnancy. Delivery at 28 weeks but less than 37 weeks of pregnancy is premature delivery, and delivery at 37 weeks but less than 42 weeks of pregnancy is called full-term delivery. Both full-term delivery and premature delivery are live birth outcomes. Biochemical pregnancy or early miscarriage can be a positive HCG test result of the patient 14 days after implantation, no gestational sac and fetal heartbeat is seen on B-ultrasound at 6 weeks, or miscarriage before 12 weeks of pregnancy. The clinical outcome after implantation of embryos without pregnancy potential can be a negative HCG test result of the patient 14 days after implantation.

[0120] It should be noted that the setting method of the embryo development quality corresponding to the transplantation outcome is not limited to the above examples, for example, premature birth, full-term birth, early miscarriage, non-pregnancy, etc. Technical personnel in the relevant field may make other changes under the inspiration of the technical essence of this application, but as long as the functions and effects achieved are the same or similar to those of this application, they should be covered within the scope of protection of this application.

[0121] In one embodiment, the obtaining of Raman spectroscopy data includes:

[0122] Acquire initial Raman spectral data.

[0123] The initial Raman spectrum data is preprocessed to obtain Raman spectrum data; wherein the preprocessing operation includes at least one of the following: quality control, peak removal, spectrum truncation, smoothing filtering, baseline correction and normalization.

[0124] Specifically, quality control can be performed by performing biochemical analysis on blank group samples as basic peak spectrum analysis, determining that the peak spectrum of the experimental group samples that do not conform to the basic peak spectrum analysis is an abnormal sample, and removing the abnormal sample from the initial Raman spectrum data. The blank group is the culture medium of the same culture batch, but no embryos are placed in the culture, and the culture medium sample is obtained at the same time.

[0125] Removing spikes: Random and accidental peaks with very narrow spectral bands and very high peak intensities are generally caused by cosmic ray interference. The general treatment method is to replace and remove the baseline of the spectral band.

[0126] Spectral interception: Based on the vibration of most biological samples, the molecular vibration will be between 600-1800cm -1 The Raman shift range is responsive, therefore, the embodiment of the present disclosure intercepts 600-1800cm -1Raman shift range as a bioinformatics "fingerprint" region. Figure 4 As shown, in Figure 4 In Figure A, the 600-1800cm -1 Bioinformation fingerprint region within the Raman shift range.

[0127] Smoothing filter: Reference Figure 4 As shown, in Figure 4 In Figure B, smoothing filtering is performed to reduce spectral noise and improve the signal-to-noise ratio.

[0128] The Savitzky-Golay high-pass filter is used for smoothing. The Savitzky-Golay high-pass filter is a finite digital filter based on local polynomial least squares fitting. The core idea is to fit a k-order polynomial to the data points within a window of a certain length, and obtain the fitting result while ensuring that the shape and width of the signal remain unchanged.

[0129] (1)

[0130] use The polynomial is fitted to the data points in the window, where x is the data to be fitted, y is the output data after fitting, and a is the parameter to be solved.

[0131] Baseline Correction: Reference Figure 4 As shown, in Figure 4 In Figure C, baseline correction is performed to fit the background and fluorescence signals to the original signal and remove them. Combined with the statistically sensitive nonlinear iterative peak-clipping algorithm (SNIP), baseline correction is completed by removing the Raman signals such as the background, fluorescence and the noise of the instrument itself. The SNIP algorithm fits and subtracts the background signal through statistical analysis of the spectral data. First, the linear least squares (LLS) fitting method is used to fit the data, and the formula is as follows:

[0132] (2)

[0133] Where i is the wave number of the original spectrum, y(i) is the intensity of the original spectrum at the i-th wave number, and V(i) is the data after LLS fitting.

[0134] Then V(i) is iterated and calculated in sequence. In the pth iteration, the average values ​​of Vp-1(i), Vp-1(i+p) and Vp-1(ip) are compared, and the smaller value of the two is determined as the value of Vp(i). The iterative formula is:

[0135] (3)

[0136] After the iteration is completed, the base spectrum data obtained by the inverse LLS transform is obtained by performing two exponential operations with base e on the transformed value y(i) and then squaring it. The formula is:

[0137] (4)

[0138] Then, the base spectrum data is subtracted from the original spectrum data to obtain the processed spectrum data. The formula is:

[0139] (5)

[0140] Normalization: Reference Figure 4 As shown, in Figure 4 In Figure D, the Raman intensity of each spectral data is normalized to the range of [0,1].

[0141] In the above embodiment, by performing a preprocessing operation on the initial Raman spectrum data, redundant information such as noise and background signals contained in the original Raman spectrum data can be removed, thereby improving the accuracy of model prediction.

[0142] In one embodiment, the initial Raman spectrum data is preprocessed to obtain Raman spectrum data, including:

[0143] Performing a preprocessing operation on the initial Raman spectrum data to obtain intermediate Raman spectrum data;

[0144] Based on the principal component analysis algorithm, the intermediate Raman spectrum data is subjected to dimensionality reduction processing to obtain Raman spectrum data.

[0145] Specifically, the principal component analysis method PCA is selected to reduce the dimension of the high-dimensional Raman spectral data samples through linear transformation, remove noise and unimportant features, retain important features, and extract the feature vector of the data. For example: the principal component analysis (PCA) algorithm is used to reduce the dimension of the pre-processed Raman spectral data, extract the Raman spectral features, and extract the first 100 principal components after dimensionality reduction.

[0146] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0147] Based on the same inventive concept, the present application embodiment also provides an embryo development quality assessment device for implementing the embryo development quality assessment method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the one or more embryo development quality assessment device embodiments provided below can refer to the limitations of the embryo development quality assessment method above, and will not be repeated here.

[0148] In one embodiment, Figure 6 As shown, a device for evaluating embryo development quality is provided, comprising:

[0149] An acquisition module 601 is used to acquire Raman spectrum data; wherein the Raman spectrum data is obtained by performing Raman spectrum detection on the embryo culture medium;

[0150] The evaluation module 603 is used to input the Raman spectral data into the embryo quality evaluation model and output the embryo development quality category corresponding to the transplantation outcome; wherein the embryo quality evaluation model is obtained by training the initial embryo quality evaluation model based on the correspondence between the sample Raman spectral data and the development quality category corresponding to the transplantation outcome.

[0151] Each module in the above-mentioned embryo development quality assessment device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0152] In one embodiment, a computer device is provided, which can be a server, and its internal structure diagram can be shown in Figure 7. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Wherein, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store embryonic development quality assessment data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for evaluating the development quality of an embryo is implemented.

[0153] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG8. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for evaluating the development quality of an embryo is implemented. The display unit of the computer device is used to form a visually visible image, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse, etc.

[0154] Those skilled in the art will understand that the structure shown in FIG. 8 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0155] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0156] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0157] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0158] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for evaluating embryo development quality, characterized in that: Applied to a processor, the method comprises: Acquiring Raman spectral data; wherein the Raman spectral data is obtained by performing Raman spectral detection on the embryo culture medium; The Raman spectral data are input into an embryo quality assessment model, and the embryo development quality category corresponding to the transplantation outcome is output; wherein the embryo quality assessment model is obtained by training an initial embryo quality assessment model based on the correspondence between the sample Raman spectral data and the development quality category corresponding to the transplantation outcome.

2. The method according to claim 1, characterized in that The culture solution includes a dried crystalline culture solution and / or a liquid culture solution, and the obtaining of Raman spectroscopy data includes: The dried crystalline culture solution is placed in an aluminum-coated silica wafer container, and the dried crystalline culture solution is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectrum data; And / or, the liquid culture solution is placed in a gold-plated copper-nickel alloy container, and the liquid culture solution is irradiated with an optical fiber Raman spectrometer to obtain Raman spectrum data.

3. The method according to claim 2, characterized in that The dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer to obtain Raman spectrum data, including: The dried crystalline culture fluid is irradiated with a microscopic confocal Raman spectrometer, and the irradiation parameters include: selecting an incident excitation power of 4-6 mW, selecting an exposure time of 3-5 s, accumulating n times, and averaging the Raman spectrum data of multiple sampling points to obtain Raman spectrum data.

4. The method according to claim 2, characterized in that: The method of irradiating the liquid culture fluid with a fiber optic Raman spectrometer to obtain Raman spectrum data includes: The liquid culture medium is irradiated with a fiber optic Raman spectrometer, and the irradiation parameters include: the incident excitation power is selected to be 300-350mW, the exposure time is selected to be 30-50s, and the Raman spectrum data of multiple sampling points are averaged for m times to obtain Raman spectrum data.

5. The method according to claim 1, characterized in that Before obtaining the Raman spectrum data, the method further includes: Obtaining the initial culture medium containing embryos at the cleavage stage; Mixing and centrifuging the initial culture solution to obtain a treated culture solution; The supernatant of the treated culture medium is extracted as the culture medium for the embryos.

6. The method according to claim 1, characterized in that The embryo quality assessment model is obtained by training the initial embryo quality assessment model based on the correspondence between the sample Raman spectral data and the transplantation outcomes corresponding to the developmental quality categories, and includes: Acquire a sample set, wherein the sample set includes Raman spectral data samples, and the Raman spectral data samples are marked with developmental quality categories corresponding to transplant outcomes; Inputting the Raman spectroscopy data sample into an initial embryo quality assessment model and outputting a prediction result; Based on the difference between the prediction result and the labeled developmental quality category, the training parameters in the initial embryo quality assessment model are iteratively adjusted to obtain the embryo quality assessment model.

7. The method according to any one of claims 1 to 6, characterized in that The developmental quality category corresponds to a transplant outcome including at least one of the following: Live birth, chemical pregnancy or early miscarriage, no pregnancy potential.

8. The method according to any one of claims 1 to 6, characterized in that The obtaining of Raman spectrum data comprises: Acquiring initial Raman spectrum data; The initial Raman spectrum data is preprocessed to obtain Raman spectrum data; wherein the preprocessing operation includes at least one of the following: quality control, peak removal, spectrum truncation, smoothing filtering, baseline correction and normalization.

9. The method according to claim 8, characterized in that The initial Raman spectrum data is preprocessed to obtain Raman spectrum data, including: Performing a preprocessing operation on the initial Raman spectrum data to obtain intermediate Raman spectrum data; Based on the principal component analysis algorithm, the intermediate Raman spectrum data is subjected to dimensionality reduction processing to obtain Raman spectrum data.

10. An embryo development quality assessment device, characterized in that: The device comprises: An acquisition module is used to acquire Raman spectrum data; wherein the Raman spectrum data is obtained by performing Raman spectrum detection on the embryo culture medium; An evaluation module is used to input the Raman spectral data into an embryo quality evaluation model and output the embryo's developmental quality category corresponding to the transplant outcome; wherein the embryo quality evaluation model is obtained by training an initial embryo quality evaluation model based on the correspondence between the sample Raman spectral data and the developmental quality category corresponding to the transplant outcome.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images