Intelligent endometrial receptivity evaluation system based on multi-modal fusion

Abstract

The invention relates to the technical field of medical informatics, in particular to an intelligent endometrial receptivity evaluation system based on multi-modal fusion. The system comprises a memory and a processor, and the processor executes a computer program stored in the memory so as to realize the following steps: acquiring endometrial thickness, estrogen content, luteal hormone content and pulse index of a reference person and a person to be evaluated; determining the intimal change stability according to the endometrial thickness increase characteristics of the reference person; according to the fluctuation condition of the estrogen content of the reference personnel, the decrease condition of the estrogen content detected in two adjacent times and the change condition of the luteal hormone content, determining the category of the to-be-evaluated personnel; and determining the creep over-frequency expression degree by combining the decreasing condition of the pulse index of the to-be-evaluated person and the uterine contraction condition so as to obtain a comprehensive index. According to the method, the data of different dimensions of the endometrium of the to-be-evaluated person is synthesized to obtain the comprehensive index, and reference data is provided for subsequent evaluation of a doctor.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical informatics, and in particular to an endometrial receptivity intelligent evaluation system based on multi-modal fusion. BACKGROUND

[0002] Endometrial receptivity refers to the ability of the endometrium to accept a embryo at a specific period, allowing the embryo to locate, adhere, penetrate and ultimately implant. This ability is usually at its best 6-10 days after ovulation (i.e. implantation window) of the menstrual cycle. However, endometrial receptivity is influenced by a variety of factors, including hormone levels, endometrial thickness, blood flow, morphological characteristics, and uterine cavity environment. The core purpose of endometrial receptivity evaluation is to determine the optimal timing of embryo implantation by scientific means to judge the receptivity of the endometrium to the embryo, so as to improve the success rate of pregnancy.

[0003] In the evaluation of endometrial receptivity, it is often through ultrasonic detection and other conventional detection to collect monitoring data of different dimensions, and the doctor analyzes the monitoring data of each dimension respectively and evaluates the endometrial receptivity of the person to be evaluated combined with his own experience. Since there are too much monitoring data, a comprehensive index cannot be provided, so this method needs to spend a lot of time, resulting in low evaluation efficiency. SUMMARY

[0004] In order to solve the problem that the existing method cannot comprehensively evaluate the different dimensions of the monitoring data of the person to be evaluated of the endometrial receptivity, the purpose of the present application is to provide an endometrial receptivity intelligent evaluation system based on multi-modal fusion, and the technical solution is as follows:

[0005] The present application provides an endometrial receptivity intelligent evaluation system based on multi-modal fusion, which comprises a memory and a processor, and the processor executes the computer program stored in the memory to realize the following steps:

[0006] Obtaining the endometrial thickness, estrogen content, luteinizing hormone content and pulse index of the reference person and the person to be evaluated;

[0007] According to the endometrial thickness growth characteristics of the reference person, the endometrial change stability degree of the reference person is determined; according to the fluctuation of the estrogen content of the reference person, the decrease of the estrogen content of the adjacent two times of detection and the change of the luteinizing hormone content, the endometrial development influence degree of the reference person is determined; the all reference patients are divided into multiple categories combined with the endometrial change stability degree and the endometrial development influence degree, and the class where the person to be evaluated is located is determined;

[0008] determine a PI decline delay influence degree based on a decline of the pulse index of the to-be-evaluated person; and determine a peristalsis over-frequency performance degree based on a uterine contraction condition of the to-be-evaluated person;

[0009] obtain a comprehensive index of endometrial receptivity of the to-be-evaluated person by synthesizing the embryo transfer day, the PI decline delay influence degree and the peristalsis over-frequency performance degree of all persons in the category to which the to-be-evaluated person belongs.

[0010] Preferably, the determination of the endometrial change homeostasis degree of the reference person according to the endometrial thickness growth characteristic of the reference person comprises:

[0011] For any reference person:

[0012] calculate a growth value of the endometrial thickness of the reference person detected at two adjacent times within a preset time period; and calculate an average value of all growth values corresponding to the reference person;

[0013] determine a fluctuation value of each growth value corresponding to the reference person by the difference between the growth value and the average value of all growth values;

[0014] determine whether each growth value corresponding to the reference person is within a preset reference range, and if yes, set an abnormal growth coefficient of the corresponding growth value as 0; and if not, calculate a difference value between each growth value and two boundary values of the preset reference range, and take the smallest difference value as the abnormal growth coefficient of the corresponding growth value;

[0015] determine the endometrial change homeostasis degree of the reference person in combination with the fluctuation value and the abnormal growth coefficient.

[0016] Preferably, the determination of the endometrial change homeostasis degree of the reference person in combination with the fluctuation value and the abnormal growth coefficient comprises:

[0017] respectively take a sum value of the abnormal growth coefficient of each growth value corresponding to the reference person and a constant 1 as a first sum value of the corresponding growth value;

[0018] respectively calculate a first product of the fluctuation value of each growth value corresponding to the reference person and the first sum value of the corresponding growth value, and determine a reciprocal normalization result of an accumulated sum of all first products as the endometrial change homeostasis degree of the reference person.

[0019] Preferably, the determination of the endometrial development influence degree of the reference person according to the fluctuation of the estrogen content, the decline of the estrogen content detected at two adjacent times and the change of the luteinizing hormone content comprises:

[0020] For any reference person:

[0021] normalizing the inverse of the standard deviation of the estrogen content of all the detection results of the any reference person in the preset time period as the stable decline degree of the estrogen of the any reference person;

[0022] screening the suspected abnormal stage of the any reference person based on the decline of the estrogen content of the any reference person in the preset time period;

[0023] calculating the first ratio between the lutein hormone content of the latter detection and the lutein hormone content of the former detection of the any reference person in the adjacent two detections in the suspected abnormal stage; and normalizing the inverse of the mean value of all the first ratios of the any reference person as the lutein hormone deficiency performance degree of the any reference person;

[0024] comprehensively considering the difference between the lutein hormone contents of the adjacent two detections, the lutein hormone deficiency performance degree and the stable decline degree of the estrogen to obtain the endometrial development influence degree of the any reference person.

[0025] Preferably, the screening of the suspected abnormal stage of the any reference person based on the decline of the estrogen content of the any reference person in the preset time period comprises:

[0026] For any reference person:

[0027] respectively calculating the first decline value of the estrogen content of the any reference person in the adjacent two detections in the preset time period, and taking the time period before the detection time of the latter detection corresponding to the maximum first decline value as the suspected abnormal stage of the any reference person.

[0028] Preferably, the comprehensive consideration of the difference between the lutein hormone contents of the adjacent two detections, the lutein hormone deficiency performance degree and the stable decline degree of the estrogen to obtain the endometrial development influence degree of the any reference person comprises:

[0029] calculating the second ratio between the maximum value and the minimum value of the lutein hormone contents of the any reference person in the adjacent two detections; and taking the mean value of all the second ratios of the any reference person as the lutein hormone persistent deficiency influence coefficient of the any reference person;

[0030] calculating the second sum of the constant 1 and the lutein hormone persistent deficiency influence coefficient;

[0031] determining the product of the lutein hormone deficiency performance degree, the second sum and the stable decline degree of the estrogen as the endometrial development influence degree of the any reference person.

[0032] Preferably, the PI decline delay impact degree is determined based on the decline of the pulse index of the to-be-evaluated person, and the determination includes:

[0033] A second decline value is recorded as the decline value of the pulse index of the to-be-evaluated person in the preset time period and in the adjacent two detections;

[0034] A date corresponding to the later detection of the two detections corresponding to the maximum point of all the second decline values is taken as a decline delay date;

[0035] A normalized result of the number of days from the first day of the preset time period to the decline delay date is taken as a PI decline delay coefficient;

[0036] An average value of all the second decline values from the first day of the preset time period to the decline delay date of the to-be-evaluated person is obtained, and an inverse proportional normalized result of the average value of all the second decline values is taken as an actual impact value;

[0037] The PI decline delay impact degree is obtained in combination with the actual impact value and the PI decline delay coefficient; the actual impact value and the PI decline delay coefficient are positively correlated with the PI decline delay impact degree.

[0038] Preferably, the peristalsis over-frequency performance degree is determined based on the uterine contraction condition of the to-be-evaluated person, and the determination includes:

[0039] A normalized result of a ratio between the number of uterine contractions per unit time and a preset contraction number of the to-be-evaluated person at each detection is taken as a peristalsis abnormality coefficient of each detection;

[0040] A second product of the number of uterine contractions per unit time and the peristalsis abnormality coefficient of the same detection at each detection is calculated;

[0041] An average value of all the second products is taken as the peristalsis over-frequency performance degree.

[0042] Preferably, the comprehensive index of endometrial receptivity of the to-be-evaluated person is obtained by comprehensively considering the embryo transfer day, the PI decline delay impact degree, and the peristalsis over-frequency performance degree of all the persons in the class to which the to-be-evaluated person belongs, and the determination includes:

[0043] The endometrial development coefficient is taken as the weight of the peristalsis over-frequency performance degree, a difference between a constant 1 and the endometrial development coefficient is taken as the weight of the PI decline delay impact degree, and the peristalsis over-frequency performance degree and the PI decline delay impact degree are weighted and summed to obtain the comprehensive index of endometrial receptivity of the to-be-evaluated person;

[0044] The endometrial development coefficient is obtained by taking a normalized result of an average value of the embryo transfer day of all the persons in the class to which the to-be-evaluated person belongs as the endometrial development coefficient.

[0045] Preferably, the endometrial change steady state degree and the endometrial development influence degree divide all reference patients into multiple categories, and determine the category in which the to-be-evaluated person is located, including:

[0046] Based on the endometrial change steady state degree and the endometrial development influence degree, the distance between each two reference persons is calculated in the form of Euclidean distance, and all reference persons are clustered based on the distance by using a DBSCAN clustering algorithm, and each cluster is taken as a category;

[0047] The category in which the to-be-evaluated person is located is determined by using a nearest neighbor method.

[0048] The present application has at least the following beneficial effects:

[0049] The present application breaks through the limitation of traditional single one-time detection result evaluation by dynamically tracking the endometrial thickness, hormone content and pulse index of different persons, constructing an endometrial development speed model, quantitatively analyzing the endometrial thickness change steady state degree and hormone synergistic evolution law, i.e., the decline of estrogen content and the change of luteinizing hormone content; coupling the PI decline delay influence degree and the peristalsis over-frequency performance degree multi-modal parameters, specifically identifying the blood flow lag of slow endometrium and the peristalsis abnormality of rapid endometrium and other latent receptivity defects, dynamically determining the influence of latent defects of the to-be-evaluated person based on category analysis, and finally fusing the traditional indicators and latent defect characteristics to realize comprehensive acquisition of comprehensive indicators of different dimensions of the endometrium of the to-be-evaluated person, providing data support for clinical precise locking of individualized embryo transfer window, providing a reference basis for subsequent evaluation of endometrial receptivity of the to-be-evaluated person by the doctor, and improving the evaluation efficiency of the doctor. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, a brief introduction will be given to the drawings needed in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0051] Figure 1 A flowchart of a method executed by a multi-modal fusion-based endometrial receptivity intelligent evaluation system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object, the multi-modal fusion-based endometrial receptivity intelligent evaluation system according to the present application is described in detail as follows by combining the drawings and preferred embodiments.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0054] The specific scheme of the multi-modal fusion based endometrial receptivity intelligent evaluation system provided by the present application will be specifically described below in combination with the drawings.

[0055] Multi-modal fusion based endometrial receptivity intelligent evaluation system embodiment:

[0056] The multi-modal fusion based endometrial receptivity intelligent evaluation system proposed in this embodiment, as shown in Figure 1 The method performed by the multi-modal fusion based endometrial receptivity intelligent evaluation system of this embodiment includes the following steps:

[0057] Step S1, obtaining the endometrial thickness, estrogen content, progesterone content and pulse index of the reference person and the person to be evaluated.

[0058] Endometrial receptivity evaluation is a key process to evaluate whether the endometrium is suitable for embryo implantation. In order to accurately evaluate this receptivity, especially in natural cycles or assisted reproductive treatment, post-ovulation data is very important.

[0059] After ovulation, the normal endometrium will thicken and show typical secretory phase manifestations, showing uniform echo and appropriate thickness. Continuous data collection based on ultrasound examination helps to determine whether the endometrium has reached a suitable thickness and morphology for embryo implantation, and to avoid missing the optimal implantation window. At the same time, the corpus luteum secretes progesterone to support the sustained development of the endometrium, and the changes in hormone levels, especially progesterone and estrogen, directly affect the receptivity of the endometrium. Continuous monitoring of hormone levels can help evaluate the normality of corpus luteum function and timely detect problems such as corpus luteum insufficiency. The development of the endometrium is a delicate process regulated by hormone levels such as estrogen and progesterone. Abnormal development speed may reflect abnormal hormone levels or other physiological problems, which can lead to endometrial dysfunction and affect implantation. At the same time, considering that the receptivity of the endometrium is not persistent, but a short "implantation window". This window usually occurs within about 5-7 days after ovulation, at which time the endometrium has been regulated by hormones and is ready to accept embryos. If the endometrium develops too slowly or too quickly, it may miss this optimal implantation window, leading to failure of embryo implantation or implantation failure. Therefore, it is necessary to collect post-ovulation monitoring data to evaluate hormone levels, which is very important for determining whether the endometrium is suitable for embryo implantation.

[0060] In the evaluation of the endometrial receptivity of the to-be-evaluated person, different dimensions of data of the to-be-evaluated person need to be comprehensively measured to obtain a comprehensive evaluation result, so as to provide a reference basis for doctors, and the doctors evaluate the endometrial receptivity of the to-be-evaluated person based on the result.

[0061] In this embodiment, the thickness of the endometrium, the PI, the estrogen content and the progesterone content of each reference person and the to-be-evaluated person are detected by the portable ultrasound instrument at each detection time in a preset time period after ovulation, and these data are used as the basis for dynamic and continuous monitoring of the corresponding person. The preset time period is a time period after ovulation, and the length of the preset time period and the number of reference persons are set by the implementer according to the specific situation, which will not be described in detail here. It should be noted that each person is detected multiple times in the preset time period.

[0062] At this point, the thickness of the endometrium, the estrogen content, the progesterone content and the PI of each reference person at each detection time in the preset time period, and the thickness of the endometrium, the estrogen content, the progesterone content and the PI of the to-be-evaluated person at each detection time are collected.

[0063] In step S2, the endometrial change stability of the reference person is determined according to the endometrial thickness growth characteristics of the reference person; the endometrial development influence degree of the reference person is determined according to the fluctuation of the estrogen content, the decrease of the estrogen content at adjacent two detections and the change of the progesterone content; all reference patients are divided into multiple categories in combination with the endometrial change stability and the endometrial development influence degree, and the class in which the to-be-evaluated person is located is determined.

[0064] For the thickness of the endometrium, the change rate of the thickness needs to be fully paid attention to, so the thickness growth at each detection is considered here. If the thickness growth value is too small, the thin endometrium receptivity may be poor; and if the thickness growth value is too large, the influence of excessive hyperplasia on the embryo implantation rate may occur.

[0065] Next, this embodiment will be described taking one reference person as an example. The method provided in this embodiment can be used to process other reference persons.

[0066] Specifically, for any reference person:

[0067] Calculate the growth value of the endometrial thickness of the reference person in the preset time period, and there is a corresponding growth value of the endometrial thickness between every two adjacent detections of the reference person in the preset time period; it should be noted that the growth value is obtained by subtracting the endometrial thickness of the previous detection from the endometrial thickness of the subsequent detection in the adjacent two detections. Calculate the average value of all corresponding growth values of the reference person in the preset time period; the absolute value of the difference between each growth value of the reference person and the average value is taken as the fluctuation value of each growth value.

[0068] Determine whether each growth value of the reference person is within the preset reference range, if yes, the abnormal growth coefficient of the corresponding growth value is 0; if not, calculate the difference value between each growth value and the two boundary values (lower limit value and upper limit value) of the preset reference range, and take the smallest difference value as the abnormal growth coefficient of the corresponding growth value; it should be noted that the preset reference range is set by the implementer according to the specific situation, which will not be described in detail here. The difference value between the growth value and the boundary value is obtained by calculating the absolute value of the difference between the growth value and the boundary value, and taking the absolute value as the difference value between the growth value and the boundary value.

[0069] The sum of the abnormal growth coefficient of each growth value of the reference person and the constant 1 is recorded as the first sum value of the corresponding growth value, and each growth value of the reference person has a first sum value. Calculate the product of the fluctuation value of each growth value of the reference person and the first sum value corresponding to the same growth value, and take the product as the first product. The reciprocal normalization result of the cumulative sum of all first products is determined as the endometrial change stability of the reference person. In this embodiment, the specific process of reciprocal normalization of the cumulative sum of all first products is to take the exponential function value with the natural constant as the base and the negative cumulative sum as the exponent as the reciprocal normalization result of the cumulative sum. As other embodiments, other reciprocal normalization methods can also be used, which will not be described in detail here.

[0070] For female hormones, as the dominant factor for endometrial proliferation, the concentration change directly affects the speed of endometrial thickening, which usually reaches a peak before ovulation and then shows a stable downward trend. Therefore, the smooth decline is a normal performance of hormone regulation, and a sharp decline may affect the uterine endometrial receptivity, and a sharp decline usually indicates insufficient luteal function.

[0071] Based on this, the embodiment analyzes whether the luteinizing hormone is in a state of deficiency in the corresponding period based on the smooth decline of female hormones, that is, whether it shows an increase in hormone level.

[0072] For any reference person:

[0073] The reciprocal normalization result of the standard deviation of the estrogen content of all the detections of the reference person within the preset time period is taken as the estrogen smooth decline degree of the reference person. In this embodiment, the specific process of reciprocal normalization of the standard deviation is that the exponential function value with the natural constant as the base and the negative standard deviation as the index is taken as the reciprocal normalization result of the standard deviation. As other implementations, other reciprocal normalization methods can also be used for processing, which will not be described in detail here.

[0074] The decline value of the estrogen content of the reference person in the adjacent two detections within the preset time period is calculated respectively, and the decline value is recorded as the first decline value. There is a first decline value of the estrogen content between each adjacent two detections within the preset time period. It should be noted that the decline value is the difference between the estrogen content of the first detection and the estrogen content of the second detection in the adjacent two detections. The detection time corresponding to the maximum first decline value is the maximum drop point, so the time period before the detection time corresponding to the second detection in the two detections corresponding to the maximum first decline value is taken as the suspected abnormal stage of the reference person, and then whether the luteinizing hormone of the reference person in the suspected abnormal stage is insufficient is analyzed.

[0075] The ratio between the luteinizing hormone content of the second detection and the luteinizing hormone content of the first detection in the adjacent two detections within the suspected abnormal stage of the reference person is calculated, and the ratio is recorded as the first ratio, that is, a plurality of first ratios are obtained. When the ratio is closer to 1 or smaller than 1, it means that the luteinizing hormone is more insufficient, and therefore the influence on the endometrial receptivity from the hormone level is greater. The reciprocal normalization result of the mean of all the first ratios corresponding to the reference person is determined as the luteinizing hormone deficiency performance degree of the reference person. In this embodiment, the method of reciprocal normalization of the mean is that the exponential function value with the natural constant as the base and the negative mean as the index is taken as the reciprocal normalization result of the mean. As other implementations, other reciprocal normalization methods can also be used, which will not be described in detail here.

[0076] At the same time, for the possible luteinizing hormone deficiency of the patient, there may also be transient deficiency and persistent deficiency. In this embodiment, the ratio between the maximum value and the minimum value of the luteinizing hormone content of the adjacent two detections of the reference person is calculated respectively, and the ratio is recorded as the second ratio, that is, a plurality of second ratios are obtained. The mean of all the second ratios corresponding to the reference person is taken as the luteinizing hormone persistent deficiency influence coefficient of the reference person. The sum of the constant 1 and the luteinizing hormone persistent deficiency influence coefficient is calculated, and the sum is recorded as the second sum. The product of the luteinizing hormone deficiency performance degree, the second sum, and the estrogen smooth decline degree is determined as the endometrial development influence degree of the reference person.

[0077] By the above method, the endometrial development influence degree of each reference person can be obtained.

[0078] The endometrial development speed condition is considered from the intuitive aspect of endometrial changes and the indirect aspect of hormone influence by the above process. Next, based on the endometrial change steady state and the endometrial development influence degree, the distance between each two reference persons is calculated in the form of Euclidean distance, and based on the distance, the DBSCAN clustering algorithm is used to cluster all reference persons to obtain multiple clustering clusters, and each clustering cluster is taken as a class. And the class where the to-be-evaluated person is located is determined by the nearest neighbor method. The DBSCAN clustering algorithm and the nearest neighbor method are prior art, and will not be described in more detail here.

[0079] Step S3, based on the decline of the pulse index of the to-be-evaluated person, the PI decline delay influence degree is determined; and based on the uterine contraction condition of the to-be-evaluated person, the peristalsis over-frequency performance degree is determined.

[0080] From a clinical point of view, the speed of endometrial development will show different influence performances, and the slow development type is often accompanied by PI decline delay, while the rapid type is prone to peristalsis over-frequency. Specifically, the pulse index decline delay is common in patients with slow endometrial development, and the endometrial blood flow is slow, and the vascular adaptability is poor, and it is difficult to meet the oxygen and nutrients required for embryo implantation; while the rapidly developing endometrium is prone to peristalsis over-frequency, which not only affects the stability of the endometrium, but also may affect the microenvironment of the endometrium, leading to reduced receptivity. Therefore, the pulse index decline delay influence is analyzed first. Under normal circumstances, the pulse index of the uterine artery should gradually decrease after ovulation, and the blood flow supply increases, while in the case of slow endometrium, the timing and degree of pulse index decline may be affected, showing a state of decline delay.

[0081] The falling value of the pulse index of the to-be-evaluated personnel in the preset time period is recorded as a second falling value, and there is a corresponding second falling value for each adjacent detection in the preset time period; it should be noted that the falling value is the difference between the pulse index of the previous detection and the pulse index of the subsequent detection; all second falling values are fitted according to the chronological order to obtain a corresponding fitting curve, and the maximum point of the fitting curve is obtained as the maximum point of the second falling value. The date corresponding to the subsequent detection of the two detections corresponding to the maximum point of all second falling values is taken as the falling delay day; the normalization result of the number of days from the first day of the preset time period to the falling delay day is determined as the PI falling delay coefficient. In this embodiment, the normalization method of the number of days is that the ratio obtained by dividing the number of days by the total number of days in the preset time period is taken as the normalization result of the number of days. As other embodiments, other data normalization methods can also be used for processing, which will not be described in detail here.

[0082] The average value of all second falling values of the to-be-evaluated personnel in the preset time period is obtained, and the inverse proportional normalization result of the average value is taken as the actual influence value; in this embodiment, the inverse proportional normalization method of the average value is that the exponential function value with the natural constant as the base number and the negative average value as the index is taken as the inverse proportional normalization result of the average value. As other embodiments, other inverse proportional normalization methods can also be used, which will not be described in detail here.

[0083] Next, the PI falling delay influence degree is obtained by combining the actual influence value and the PI falling delay coefficient, and the actual influence value and the PI falling delay coefficient are positively correlated with the PI falling delay influence degree.

[0084] The positive correlation means that the dependent variable will increase with the increase of the independent variable, and the dependent variable will decrease with the decrease of the independent variable, which can be an additive relationship, a multiplication relationship, etc., and is determined by actual application.

[0085] In this embodiment, the product of the actual influence value and the PI falling delay coefficient is taken as the PI falling delay influence degree.

[0086] Hyperperistalsis refers to the muscle of the uterus or the contraction activity of the uterus being too frequent, which is often caused by excessive excitement of the uterus or excessive stimulation of the endometrium. In the case of fast endometrium, due to the rapid development of the endometrium, it may be accompanied by excessive excitement of the uterus, resulting in too frequent contraction activity of the uterus.

[0087] The normalized result of the ratio between the number of uterine contractions per unit time of the person to be evaluated at each detection and the preset number of contractions is taken as the peristalsis abnormality coefficient of each detection; in this embodiment, the unit time is 1 minute, and the preset number of contractions is 3, which can be set according to specific circumstances in specific applications.

[0088] The product of the number of uterine contractions per unit time at each detection and the peristalsis abnormality coefficient of the same detection is calculated, and the product is recorded as a second product; there is a second product for each detection, and the average of all second products is taken as the peristalsis hyperactivity performance.

[0089] Step S4: The embryo transfer day, the PI decline delay impact degree and the peristalsis hyperactivity performance of all persons in the class to which the person to be evaluated belongs are integrated to obtain the comprehensive index of endometrial receptivity of the person to be evaluated.

[0090] The typical hidden defect of the slow type is blood flow lag, and the typical hidden defect of the fast type is peristalsis hyperactivity. When performing intelligent evaluation of receptivity, comprehensive dynamic evaluation and analysis are performed, the PI weight of the slow type is paid more attention to, and the influence of peristalsis hyperactivity is focused on for the fast type.

[0091] In this embodiment, the normalized result of the average of the embryo transfer days of all persons in the class to which the person to be evaluated belongs is taken as the endometrial development coefficient.

[0092] The endometrial development coefficient is taken as the weight of the peristalsis hyperactivity performance, the difference between the constant 1 and the endometrial development coefficient is taken as the weight of the PI decline delay impact degree, and the peristalsis hyperactivity performance and the PI decline delay impact degree are weighted and summed to obtain the comprehensive index of endometrial receptivity of the person to be evaluated.

[0093] In this embodiment, the specific calculation formula of the comprehensive index is given, and the comprehensive index of endometrial receptivity of the person to be evaluated can be expressed as:

[0094] T = V x R + (1-V) x X

[0095] Wherein, T represents the comprehensive index of endometrial receptivity of the person to be evaluated, V represents the endometrial development coefficient, R represents the peristalsis hyperactivity performance, and X represents the PI decline delay impact degree.

[0096] After obtaining the comprehensive index of endometrial receptivity of the person to be evaluated, the doctor can refer to the comprehensive index and experience to evaluate the endometrial receptivity of the person to be evaluated.

[0097] At this point, by using the method provided in this embodiment, the doctor can be provided with accurate reference basis, and the efficiency of the doctor in evaluating the endometrial receptivity can be improved.

[0098] The embodiment constructs an endometrial development speed model by dynamically tracking endometrial thickness, hormone content and pulse index of different persons, quantitatively analyzes the steady state degree of endometrial thickness change and the synergistic evolution law of hormones, that is, the decline of estrogen content and the change of luteinizing hormone content, breaks through the limitation of traditional single detection result evaluation; the multi-modal parameters of PI decline delay influence degree and peristalsis over-frequency performance degree are coupled, and the hidden receptivity defects such as blood flow lag of slow endometrium and peristalsis abnormality of rapid endometrium are specifically identified, and the influence situation of the hidden defects of the person to be evaluated is dynamically determined based on category analysis, finally the traditional index and the hidden defect characteristics are fused, the comprehensive index of the comprehensive acquisition of the data of the endometrium of the person to be evaluated in different dimensions is obtained, the data support is provided for the clinical accurate locking of the individualized embryo transfer window, the reference basis is provided for the subsequent evaluation of the endometrial receptivity of the person to be evaluated by the doctor, and the evaluation efficiency of the doctor is improved.

[0099] It should be noted that: the above only describes the preferred embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A multi-modal fusion based endometrial receptivity intelligent assessment system, characterized in that, The system comprises a memory and a processor, characterized in that the processor executes a computer program stored in the memory to realize the following steps: Obtaining the endometrial thickness, estrogen content, progesterone content and pulse index of the reference personnel and the personnel to be evaluated; According to the endometrial thickness growth characteristics of the reference personnel, determining the endometrial change stability degree of the reference personnel; according to the fluctuation of the estrogen content of the reference personnel, the decline of the estrogen content in the adjacent two detections and the change of the progesterone content, determining the endometrial development influence degree of the reference personnel; combining the endometrial change stability degree and the endometrial development influence degree, dividing all the reference patients into multiple categories, and determining the category where the personnel to be evaluated is located; Based on the decline of the pulse index of the personnel to be evaluated, determining the PI decline delay influence degree; based on the uterine contraction condition of the personnel to be evaluated, determining the peristalsis over-frequency performance degree; Comprehensively obtaining the embryo transfer days, PI decline delay influence degree and peristalsis over-frequency performance degree of all personnel in the category where the personnel to be evaluated is located, and obtaining the comprehensive index of the endometrial receptivity of the personnel to be evaluated.

2. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The endometrial change stability degree of the reference personnel is determined according to the endometrial thickness growth characteristics of the reference personnel, comprising: For any reference personnel: Calculate the growth value of the endometrial thickness detected in the adjacent two times within a preset time period for the any reference personnel; calculate the average value of all the growth values corresponding to the any reference personnel; The difference between each of the growth values corresponding to the any reference personnel and the average value of all the growth values is taken as the fluctuation value of each growth value; It is judged whether each of the growth values corresponding to the any reference personnel is within a preset reference range, if yes, the abnormal growth coefficient of the corresponding growth value is 0; if not, the difference value between each growth value and the two boundary values of the preset reference range is calculated, and the smallest difference value is taken as the abnormal growth coefficient of the corresponding growth value; The endometrial change stability degree of the any reference personnel is determined by combining the fluctuation value and the abnormal growth coefficient.

3. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 2, wherein, The endometrial change stability degree of the any reference personnel is determined by combining the fluctuation value and the abnormal growth coefficient, comprising: The sum value of the abnormal growth coefficient of each of the growth values corresponding to the any reference personnel and the constant 1 is taken as the first sum value of the corresponding growth value; The first product of the fluctuation value of each of the growth values corresponding to the any reference personnel and the first sum value of the corresponding growth value is calculated, and the inverse proportional normalization result of the cumulative sum of all the first products is determined as the endometrial change stability degree of the any reference personnel.

4. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The endometrial development influence degree of the reference personnel is determined according to the fluctuation of the estrogen content of the reference personnel, the decline of the estrogen content in the adjacent two detections and the change of the progesterone content, comprising: For any reference personnel: The inverse proportional normalization result of the standard deviation of all the detected estrogen contents within a preset time period for the any reference personnel is taken as the estrogen smooth decline degree of the any reference personnel; screening a suspected abnormal stage of the any reference person based on a decrease of the estrogen content of the any reference person within a preset time period; calculating a first ratio between the lutein hormone content of the any reference person in the latter detection and the lutein hormone content of the any reference person in the former detection within the suspected abnormal stage; and determining a lutein hormone deficiency performance degree of the any reference person based on an inverse normalization result of a mean value of all the first ratios corresponding to the any reference person; obtaining an endometrial development influence degree of the any reference person by comprehensively considering a difference between the lutein hormone contents of the any reference person in the adjacent two detections, the lutein hormone deficiency performance degree, and the female hormone stable decrease degree.

5. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 4, wherein, The screening a suspected abnormal stage of the any reference person based on a decrease of the estrogen content of the any reference person within a preset time period, comprises: For any reference person: respectively calculating a first decrease value of the estrogen content of the any reference person in the adjacent two detections within a preset time period, and taking a time period before a detection time of the latter detection corresponding to the maximum first decrease value as a suspected abnormal stage of the any reference person.

6. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 4, wherein, The obtaining an endometrial development influence degree of the any reference person by comprehensively considering a difference between the lutein hormone contents of the any reference person in the adjacent two detections, the lutein hormone deficiency performance degree, and the female hormone stable decrease degree, comprises: calculating a second ratio between a maximum value and a minimum value of the lutein hormone contents of the any reference person in the adjacent two detections; and taking a mean value of all the second ratios corresponding to the any reference person as a lutein hormone persistent deficiency influence coefficient of the any reference person; calculating a second sum value of the lutein hormone persistent deficiency influence coefficient and a constant 1; determining an endometrial development influence degree of the any reference person based on a product of the lutein hormone deficiency performance degree, the second sum value, and the female hormone stable decrease degree.

7. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The determining a PI decrease delay influence degree based on a decrease of a pulse index of the to-be-evaluated person, comprises: taking a decrease value of the pulse index of the to-be-evaluated person in the adjacent two detections within a preset time period as a second decrease value; taking a date of the latter detection corresponding to a maximum value point of all the second decrease values as a decrease delay day; determining a PI decrease delay coefficient based on a normalization result of a number of days from a first day of the preset time period to the decrease delay day; obtaining an average value of all the second decrease values of the to-be-evaluated person from the first day of the preset time period to the decrease delay day, and taking an inverse normalization result of the average value of all the second decrease values as an actual influence value; obtaining a PI decrease delay influence degree by combining the actual influence value and the PI decrease delay coefficient; the actual influence value and the PI decrease delay coefficient are in positive correlation with the PI decrease delay influence degree.

8. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The determining a peristalsis over-frequency performance degree based on a uterine contraction condition of the to-be-evaluated person, comprises: The normalization result of the ratio between the number of uterine contractions per unit time of the person to be evaluated at each detection and the preset number of contractions is taken as a peristalsis abnormality coefficient of each detection; The number of uterine contractions per unit time at each detection is calculated as a second product of the peristalsis abnormality coefficient of the same detection; The mean value of all the second products is taken as a hyperperistalsis performance degree.

9. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The comprehensive index of endometrial receptivity of the person to be evaluated is obtained by synthesizing the embryo transfer day, the PI drop delay impact degree and the hyperperistalsis performance degree of all the persons in the class to which the person to be evaluated belongs, and includes: The endometrial development coefficient is taken as the weight of the hyperperistalsis performance degree, the difference between the constant 1 and the endometrial development coefficient is taken as the weight of the PI drop delay impact degree, and the hyperperistalsis performance degree and the PI drop delay impact degree are weighted and summed to obtain the comprehensive index of endometrial receptivity of the person to be evaluated; The acquisition of the endometrial development coefficient includes: the normalization result of the mean value of the embryo transfer day of all the persons in the class to which the person to be evaluated belongs is taken as the endometrial development coefficient.

10. The multi-modal fusion based endometrial receptivity intelligent assessment system according to claim 1, wherein, The reference patients are divided into multiple classes by combining the endometrial change homeostasis degree and the endometrial development impact degree, and the class to which the person to be evaluated belongs is determined, and includes: Based on the endometrial change homeostasis degree and the endometrial development impact degree, the distance between each two reference persons is calculated in the form of Euclidean distance, all the reference persons are clustered based on the distance by using the DBSCAN clustering algorithm, and each cluster is taken as a class; The class to which the person to be evaluated belongs is determined by the nearest neighbor method.

Images