Method, device, equipment and storage medium for predicting abnormality of fetal heart rate curve
By analyzing fetal heart rate curves using segmented prediction methods and deep learning models, the problems of missing data and individual differences in fetal heart rate curve interpretation were solved, improving the consistency of interpretation and prediction efficiency, and enhancing interpretability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- PING AN TECH (SHENZHEN) CO LTD
- Filing Date
- 2023-06-30
- Publication Date
- 2026-07-03
Smart Images
Figure CN116831546B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of digital healthcare, and more particularly to a method for predicting the abnormality of a fetal heart rate curve, an apparatus for predicting the abnormality, a computer device, and a computer-readable storage medium. Background Technology
[0002] In clinical practice, medical staff mainly assess fetal oxygenation in the uterus by interpreting the baseline and variability of the fetal heart rate curve. Because fetal heart rate monitoring requires continuous monitoring of the fetal heart rate for a long period of time, situations such as inaccurate probe positioning or the fetus being in a sleep cycle may occur, leading to problems such as missing data in the fetal heart rate curve and large individual differences in the curve. This makes manual interpretation difficult, resulting in poor consistency of interpretation results and a high false positive rate.
[0003] Current methods based on clinical guidelines are not ideal for accurately interpreting fetal heart rate baselines when fetal activity is high or baseline variability is observed. Furthermore, supervised learning-based diagnostic models, such as classification algorithms based on deep learning and decision trees, lack consistent annotation information on the fetal heart rate curve, resulting in slower prediction efficiency and less interpretable results. Summary of the Invention
[0004] This application provides a method, an apparatus for predicting the abnormality of a fetal heart rate curve, a computer device, and a computer-readable storage medium, which aims to determine the abnormality of the fetal heart rate curve through a segmented prediction method, thereby improving both prediction efficiency and interpretability during the prediction process.
[0005] To achieve the above objectives, this application provides a method for predicting the abnormality of a fetal heart rate curve, the method comprising:
[0006] Obtain the target fetal heart rate curve and determine the missing data of the target fetal heart rate curve to obtain the corresponding initial mask;
[0007] The target fetal heart rate curve is divided into several target fetal heart rate curve segments of a first preset duration, and the target mask for each target fetal heart rate curve segment is determined based on the initial mask.
[0008] The target baseline fetal heart rate is obtained by analyzing each target fetal heart rate curve segment and the corresponding target mask using a target deep learning model.
[0009] Based on the target baseline fetal heart rate, the target variation value and the target acceleration number of each target fetal heart rate curve segment are obtained, and the target abnormality of the target fetal heart rate curve is obtained based on the target variation value and the target acceleration number.
[0010] To achieve the above objectives, this application also provides an anomaly prediction device, comprising:
[0011] The acquisition module is used to acquire the target fetal heart rate curve, determine the missing data of the target fetal heart rate curve, and obtain the corresponding initial mask.
[0012] The determining module is used to divide the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and to determine the target mask for each target fetal heart rate curve segment based on the initial mask.
[0013] The analysis module is used to analyze each target fetal heart rate curve segment and the corresponding target mask through a target deep learning model to obtain the target baseline fetal heart rate.
[0014] The prediction module is used to obtain the target variation value and the target acceleration number for each segment of the target fetal heart rate curve based on the target baseline fetal heart rate, and to obtain the target abnormality of the target fetal heart rate curve based on the target variation value and the target acceleration number.
[0015] In addition, to achieve the above objectives, this application also provides a computer device, the computer device including a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and, when executing the computer program, implement the steps of the method for predicting the abnormality of the fetal heart rate curve provided in any of the embodiments of this application.
[0016] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the steps of the method for predicting the abnormality of a fetal heart rate curve as described in any of the embodiments of this application.
[0017] The method, apparatus, computer device, and computer-readable storage medium disclosed in this application for predicting the anomaly degree of a fetal heart rate curve can acquire a target fetal heart rate curve, determine missing data in the fetal heart rate curve, and obtain a corresponding initial mask. Further, the target fetal heart rate curve can be divided into several target fetal heart rate curve segments of a first preset duration, and a target mask for each target fetal heart rate curve segment can be determined based on the initial mask. After analyzing each target fetal heart rate curve segment and its corresponding target mask using a target deep learning model to obtain the target baseline fetal heart rate, the target variation value and target acceleration number of each target fetal heart rate curve segment can be obtained based on the target baseline fetal heart rate, thereby obtaining the target anomaly degree of the target fetal heart rate curve based on the target variation value and target acceleration number. This application aims to predict the target anomaly degree of the target fetal heart rate curve by using a segmented prediction method and a target deep learning model to analyze each target fetal heart rate curve segment and its corresponding target mask to obtain the target baseline fetal heart rate. The method proposed in this application not only improves the prediction efficiency of the abnormality of the target fetal heart rate curve, but also improves the interpretability of the prediction process. Attached Figure Description
[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a schematic diagram of a method for predicting the abnormality of a fetal heart rate curve provided in an embodiment of this application.
[0020] Figure 2 This is a flowchart illustrating a method for predicting the abnormality of a fetal heart rate curve according to an embodiment of this application.
[0021] Figure 3 This is a schematic diagram of a process for obtaining an initial mask provided in an embodiment of this application;
[0022] Figure 4 This is a flowchart illustrating another method for predicting the abnormality of a fetal heart rate curve provided in an embodiment of this application.
[0023] Figure 5 This is a schematic block diagram of an anomaly prediction device provided in an embodiment of this application;
[0024] Figure 6 This is a schematic block diagram of a computer device provided in an embodiment of this application. Detailed Implementation
[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0026] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily need to be performed in the described order. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation. Furthermore, although functional modules are divided in the device diagram, in some cases, a different module division may be used.
[0027] The term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items, as well as all possible combinations, and includes such combinations.
[0028] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.
[0029] like Figure 1 As shown, the method for predicting the abnormality of a fetal heart rate curve provided in this application embodiment can be applied to, for example... Figure 1The application environment shown includes a terminal device 110 and a server 120, wherein the terminal device 110 can communicate with the server 120 via a network. Specifically, the server 120 can acquire the target fetal heart rate curve, determine the missing data of the target fetal heart rate curve, and obtain the corresponding initial mask; then, it divides the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and determines the target mask for each target fetal heart rate curve segment based on the initial mask; it analyzes each target fetal heart rate curve segment and the corresponding target mask through a target deep learning model to obtain the target baseline fetal heart rate; finally, based on the target baseline fetal heart rate, it obtains the target variation value and the target acceleration number of each target fetal heart rate curve segment, and obtains the target anomaly degree of the target fetal heart rate curve based on the target variation value and the target acceleration number, and sends the target anomaly degree to the terminal device 110. The server 120 can be a standalone server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The terminal device 110 can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and server can be directly or indirectly connected via wired or wireless communication; this application does not impose any restrictions on this connection.
[0030] Please see Figure 2 , Figure 2 This is a flowchart illustrating a method for predicting the abnormality of a fetal heart rate curve according to an embodiment of this application. This method for predicting the abnormality of a fetal heart rate curve can be applied in a computer device, thereby achieving a more accurate answer based on the target problem.
[0031] like Figure 2 As shown, the method for predicting the abnormality of the fetal heart rate curve includes steps S11 to S14.
[0032] Step S11: Obtain the target fetal heart rate curve and determine the missing data of the target fetal heart rate curve to obtain the corresponding initial mask.
[0033] The target fetal heart rate curve is the fetal heart rate curve for which the degree of abnormality is to be determined.
[0034] It's important to note that the fetal heart rate curve (FHR curve) is a graphical representation of the fetal heart rate recorded during labor. The FHR curve is typically monitored using a fetal heart rate monitor, and changes in the fetal heart rate are plotted during monitoring. Therefore, the FHR curve is an important indicator during labor, helping healthcare professionals assess the fetus's health and monitor for potential difficulties or complications. Under normal circumstances, the FHR curve should show certain variations, including different patterns such as baseline heart rate, acceleration, and deceleration. The appearance and changes in these patterns provide information about the fetus, helping doctors make appropriate interventions and decisions.
[0035] Understandably, during fetal heart rate monitoring, situations may arise such as inaccurate probe placement, changes in fetal heart rate position, or the fetus being in a sleep cycle, leading to missing data in the fetal heart rate curve. Consequently, it becomes impossible to accurately interpret the fetal heart rate curve, thus requiring prediction of the degree of abnormality in the fetal heart rate curve.
[0036] For the reasons stated above, embodiments of this application can determine the missing data of the target fetal heart rate curve and identify its corresponding initial mask, thus obtaining the corresponding initial mask. In this way, the actual data and missing data in the target fetal heart rate curve can be distinguished, and the missing data can be selectively processed, such as filtering or filling in the missing data.
[0037] In this embodiment, a target fetal heart rate curve can be obtained, and missing data of the target fetal heart rate curve can be identified to obtain a corresponding initial mask. In this way, the missing data can be processed, thereby achieving a more accurate prediction of the anomaly degree of the target fetal heart rate curve.
[0038] Step S12: Divide the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and determine the target mask for each target fetal heart rate curve segment based on the initial mask.
[0039] In this application, the first preset duration is not limited, and can be represented by Window1 for example.
[0040] Furthermore, the target fetal heart rate curve is divided into several target fetal heart rate curve segments of a first preset duration, which can be represented by the following formula:
[0041]
[0042] Among them, S w The target fetal heart rate curve segment, with each segment lasting Window1, i.e.
[0043] Optionally, the target fetal heart rate curve segment includes a first fetal heart rate curve segment and a second fetal heart rate curve segment, and the target mask includes a first mask corresponding to the first fetal heart rate curve segment and a second mask corresponding to the second fetal heart rate curve segment; the target mask for each target fetal heart rate curve segment is determined based on the initial mask by the following expression:
[0044]
[0045]
[0046] Among them, Mask i (t) is the first mask; The second mask; Mask missing (t) is the initial mask; The target fetal heart rate curve segment.
[0047] Furthermore, the first fetal heart rate curve segment corresponds to the normal data; the second fetal heart rate curve segment corresponds to the missing data.
[0048] In this embodiment of the application, the target fetal heart rate curve can be divided into several target fetal heart rate curve segments of a first preset duration, and the target mask of each target fetal heart rate curve segment can be determined based on the initial mask, so as to process the missing data and thereby achieve a more accurate prediction of the abnormality of the target fetal heart rate curve.
[0049] Step S13: Analyze each target fetal heart rate curve segment and its corresponding target mask using the target deep learning model to obtain the target baseline fetal heart rate.
[0050] Specifically, each target fetal heart rate curve segment and its corresponding target mask can be input into the target deep learning model to output the baseline fetal heart rate.
[0051] The baseline fetal heart rate can be expressed by the following formula:
[0052]
[0053] in, Baseline fetal heart rate; FHR(t) is the target fetal heart rate curve segment; Seq2Seq(t) is the output of the model when the target depth model is Seq2Seq(t); Mask i (t) is the target mask.
[0054] Optionally, in order to obtain a smoother baseline fetal heart rate, the baseline fetal heart rate can be filtered by a 10-minute median filter to obtain a filtered baseline fetal heart rate.
[0055] In the embodiments of this application, each target fetal heart rate curve segment and its corresponding target mask can be analyzed by a target deep learning model to obtain the target baseline fetal heart rate, which can be used to predict the abnormality of the target fetal heart rate curve.
[0056] Step S14: Based on the target baseline fetal heart rate, obtain the target variation value and target acceleration number for each target fetal heart rate curve segment, and obtain the target abnormality of the target fetal heart rate curve based on the target variation value and target acceleration number.
[0057] Specifically, for calculating the target variability: the duration of the target baseline fetal heart rate can be determined, which is typically at least 10 minutes or longer. Within this time period, the fetal heart rate values per minute can be calculated, and these values are subtracted from the baseline fetal heart rate to obtain the variability per minute. The absolute values of these variability values are summed and divided by the total number of variability values to obtain the target variability.
[0058] For calculating the target abnormality: A baseline fetal heart rate duration can be determined, typically at least 10 minutes or longer. Within this timeframe, observe whether the fetal heart rate rises above a specific threshold (usually above 15 bpm) and persists for a certain period (usually above 15 seconds). The number of times the fetal heart rate exceeds the threshold is ultimately counted as the target acceleration count.
[0059] Optionally, the target baseline fetal heart rate includes a first baseline fetal heart rate and a second baseline fetal heart rate, the first baseline fetal heart rate corresponding to a first mask; the second baseline fetal heart rate corresponding to a second mask, and the target variation value and target acceleration number for each target fetal heart rate curve segment are obtained based on the target baseline fetal heart rate, including: obtaining the first variation value and first acceleration number for the corresponding first fetal heart rate curve segment based on the first baseline fetal heart rate; and obtaining the second variation value and second acceleration number for the corresponding second fetal heart rate curve segment based on the second baseline fetal heart rate; obtaining the target variation value based on the first variation value and the second variation value, and obtaining the target acceleration number based on the first acceleration number and the second acceleration number.
[0060] The first baseline fetal heart rate is the baseline fetal heart rate corresponding to the normal data, which corresponds to the first variation value and the first acceleration number; the second baseline fetal heart rate is the baseline fetal heart rate corresponding to the missing data, which corresponds to the second variation value and the second acceleration number.
[0061] Therefore, the above steps can be used to determine the variability and number of accelerations corresponding to different baseline fetal heart rates, and then determine the target variability and target number of accelerations.
[0062] Understandably, the target mutation value includes the first mutation value and the second mutation value; the target acceleration number includes the first acceleration number and the second acceleration number.
[0063] Optionally, based on the above embodiments, the target abnormality of the target fetal heart rate curve is obtained based on the target variation value and the target acceleration number, including: sorting the target variation value and the target acceleration number according to a preset rule to obtain a first order and a second order; and obtaining the target abnormality based on the first order, the second order, the target variation value, and the target acceleration number.
[0064] Specifically, this application does not limit the sorting rules; for example, the sorting can be from smallest to largest.
[0065] Furthermore, the target anomaly degree can be obtained using the following formula:
[0066] Ab i =λ V Ord(V i )+λ A Ord(A i )
[0067] Among them, Ab i Ord(V) represents the target anomaly degree. i ) represents V i exist The order in; V i A represents the target variation value; i The number of accelerations for the target; ord(A i ) represents A i exist The order in; λ V ,λ A These are preset parameters.
[0068] In the embodiments of this application, the target variation value and target acceleration number of each target fetal heart rate curve segment can be obtained based on the target baseline fetal heart rate, and the target abnormality of the target fetal heart rate curve can be obtained based on the target variation value and target acceleration number, thereby realizing the prediction of the target abnormality of the target fetal heart rate curve.
[0069] The method for predicting the anomaly of a fetal heart rate curve disclosed in this application can acquire a target fetal heart rate curve, determine missing data in the fetal heart rate curve, and obtain a corresponding initial mask. Further, the target fetal heart rate curve can be divided into several target fetal heart rate curve segments of a first preset duration, and a target mask for each target fetal heart rate curve segment can be determined based on the initial mask. After analyzing each target fetal heart rate curve segment and its corresponding target mask using a target deep learning model to obtain the target baseline fetal heart rate, the target variation value and target acceleration number of each target fetal heart rate curve segment can be obtained based on the target baseline fetal heart rate, thereby obtaining the target anomaly of the target fetal heart rate curve based on the target variation value and the target acceleration number. This application aims to predict the target anomaly of the target fetal heart rate curve by using a segmented prediction method and a target deep learning model to analyze each target fetal heart rate curve segment and its corresponding target mask to obtain the target baseline fetal heart rate. The method proposed in this application not only improves the prediction efficiency of the anomaly of the target fetal heart rate curve but also improves the interpretability of the prediction process.
[0070] Please see Figure 3 , Figure 3 This is a schematic diagram illustrating a process for obtaining an initial mask provided in an embodiment of this application. For example... Figure 3 As shown, the initial mask can be obtained through steps S141 to S143.
[0071] Step S111: Obtain the target fetal heart rate curve, and divide the target fetal heart rate curve into several initial fetal heart rate curve segments based on the first preset threshold and the first preset conditions.
[0072] Optionally, before dividing the target fetal heart rate curve into several initial fetal heart rate curve segments based on the first preset threshold and the first preset conditions, the method further includes: filtering the target fetal heart rate curve through a median filter of a second preset duration to obtain the filtered target fetal heart rate curve.
[0073] It should be noted that this application does not limit the second preset duration. For example, the second preset duration may be 5 minutes or 10 minutes. This application will use a second preset duration of 5 minutes as an example for explanation.
[0074] Understandably, filters are a common signal processing technique used to denoise and smooth curves or signals. Therefore, embodiments of this application can filter the target fetal heart rate curve based on a median filter to remove noise, preserve edges and details, and thus obtain a smoother filtered target fetal heart rate curve.
[0075] Furthermore, based on the above embodiments, the first preset threshold can be expressed as thres5; the first preset condition can be expressed as:
[0076]
[0077] Wherein, C5(t) is the first preset condition; FHR(t) is the target fetal heart rate curve; MedianFilter5(FHR(t)) is the filtered target fetal heart rate curve; and thres5 is the first preset threshold.
[0078] Furthermore, the division of the target fetal heart rate curve into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition can be represented by the following formula:
[0079]
[0080] S5 represents the initial fetal heart rate curve segment; These are different time periods, which this application does not limit and C5(t) = True.
[0081] In this embodiment of the application, a target fetal heart rate curve can be obtained, and the target fetal heart rate curve can be divided into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition, thereby realizing the processing of the initial fetal heart rate curve.
[0082] Step S112: Based on several initial fetal heart rate curve segments, obtain the linearly interpolated initial fetal heart rate curve through linear interpolation.
[0083] Understandably, linear interpolation is a common interpolation method that can achieve effects such as smoothing curves and supplementing missing values. Therefore, this application can process several initial fetal heart rate curve segments based on the linear interpolation method to obtain the initial fetal heart rate curve after linear interpolation.
[0084] Specifically, the initial fetal heart rate curve obtained by linear interpolation based on several initial fetal heart rate curve segments can be represented by the following formula:
[0085]
[0086] in, The initial fetal heart rate curve is obtained after linear interpolation; FHR(t) is the target fetal heart rate curve; S5 is the segment of the initial fetal heart rate curve. This application does not specify different time periods.
[0087] If S5 includes the left and right endpoints of the original curve, it is defined at the endpoints.
[0088]
[0089]
[0090] In the embodiments of this application, the above method can be used to estimate and supplement missing data. By using the linear relationship between known data points, not only can the curve be completely recovered, that is, the initial fetal heart rate curve after linear interpolation can be obtained, but also the approximate value of the missing data can be inferred, thereby determining the initial mask of the missing data.
[0091] Step S113: Based on the initial fetal heart rate curve after linear interpolation, determine the initial mask through the second preset threshold and the second preset conditions.
[0092] Optionally, before determining the initial mask based on the linearly interpolated initial fetal heart rate curve using the second preset threshold and the second preset conditions, the method further includes: filtering the linearly interpolated initial fetal heart rate curve using a median filter of a third preset duration to obtain a filtered initial fetal heart rate curve.
[0093] It should be noted that this application does not limit the second preset duration. For example, the second preset duration is 1 minute, 2 minutes, etc. This application will use a second preset duration of 1 minute as an example for explanation.
[0094] Understandably, filters are a common signal processing technique used to denoise and smooth curves or signals. Therefore, this embodiment can filter the initial fetal heart rate curve after linear interpolation based on a median filter to remove noise, preserve edges and details, and thus obtain a smoother filtered initial fetal heart rate curve.
[0095] Based on the above embodiments, the second preset threshold can be represented as thres1; the second preset condition can be represented as:
[0096]
[0097] Where C1(t) is the second preset condition; FHR(t) is the target fetal heart rate curve; The filtered initial fetal heart rate curve; thres1 is the second preset threshold.
[0098] Furthermore, the initial mask can be expressed by the following formula:
[0099] Mask missing (t)=C5(t)|C1(t)
[0100] Among them, Mask missingC5(t) is the initial mask; C5(t) is the first preset condition; C1(t) is the second preset condition.
[0101] In this embodiment, a target fetal heart rate curve can be obtained, and the target fetal heart rate curve can be divided into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition. The several initial fetal heart rate curve segments are then linearly interpolated to obtain linearly interpolated initial fetal heart rate curves. Based on the linearly interpolated initial fetal heart rate curves, an initial mask is determined using a second preset threshold and a second preset condition.
[0102] Please continue reading. Figure 4 , Figure 4 This is a flowchart illustrating another method for predicting the abnormality of a fetal heart rate curve provided in an embodiment of this application. Figure 4 As shown, the abnormality of the fetal heart rate curve can be predicted through steps S21 to S23.
[0103] Step S21: Obtain the target uterine contraction curve associated with the target fetal heart rate curve.
[0104] Step S22: Based on the target uterine contraction curve, determine the early deceleration value and the late deceleration value.
[0105] The target uterine contraction curve is the uterine contraction curve of the mother corresponding to the target fetal heart rate curve.
[0106] Furthermore, regarding the determination of early deceleration values: the time points at the onset and end of contractions can be identified. Since early deceleration refers to a decrease in fetal heart rate occurring simultaneously with or shortly after the onset of a contraction, and returning to baseline levels before the contraction ends, the time interval from the onset of a contraction to the occurrence of early deceleration, as well as the duration of early deceleration, can be calculated. This allows the determination of the early deceleration value, i.e., the fetal heart rate decrease event occurring at or shortly after the onset of a contraction.
[0107] Determining the value of late decelerations: In the target contraction curve, the time points at the onset and end of contractions can be identified. Since late decelerations refer to a decrease in fetal heart rate occurring simultaneously with or shortly after the onset of a contraction and continuing for a period after the contraction ends, the time interval from the onset of a contraction to the occurrence of late decelerations, as well as the duration of the late decelerations, can be calculated. This allows for the determination of the late deceleration value, which is the prolonged fetal heart rate decrease event occurring at or shortly after the onset of a contraction.
[0108] Step S23: Determine the target abnormality of the target fetal heart rate curve based on the early deceleration value, late deceleration value, target variation value, and target acceleration number.
[0109] Specifically, the target abnormality of the target fetal heart rate curve can be determined using the following formula:
[0110] Ab i =λ V Ord(V i )+λ A Ord(A i )+λ ED Ord(ED i )+λ LD Ord(LD i )
[0111] Among them, Ab i Ord(V) represents the target anomaly degree. i ) represents V i exist The order in; V i A represents the target variation value; i The number of accelerations for the target; ord(A i ) represents A i exist The order in the middle; ED i This represents the early deceleration value; Ord(ED) i ) indicates ED i exist The order in; LD i This refers to the late-stage deceleration value; Ord(LD) i ) represents LD i exist The order in; λ V ,λ A ,λ ED ,λ LD These are preset parameters.
[0112] In this embodiment of the application, a target uterine contraction curve associated with the target fetal heart rate curve can also be obtained. Then, based on the target uterine contraction curve, the early deceleration value and the late deceleration value can be determined. Finally, based on the early deceleration value, the late deceleration value, the target variation value, and the target acceleration number, the target abnormality of the target fetal heart rate curve can be determined.
[0113] Please see Figure 5 , Figure 5 This is a schematic block diagram of an anomaly prediction device provided in an embodiment of this application. The anomaly prediction device can be configured in a server to perform the aforementioned method for predicting the anomaly of a fetal heart rate curve.
[0114] like Figure 5 As shown, the anomaly prediction device 200 includes: an acquisition module 201, a determination module 202, an analysis module 203, and a prediction module 204.
[0115] The acquisition module 201 is used to acquire the target fetal heart rate curve, determine the missing data of the target fetal heart rate curve, and obtain the corresponding initial mask;
[0116] The determining module 202 is used to divide the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and determine the target mask for each target fetal heart rate curve segment based on the initial mask.
[0117] Analysis module 203 is used to analyze each target fetal heart rate curve segment and the corresponding target mask through a target deep learning model to obtain the target baseline fetal heart rate;
[0118] The prediction module 204 is used to obtain the target variation value and the target acceleration number of each segment of the target fetal heart rate curve based on the target baseline fetal heart rate, and to obtain the target abnormality of the target fetal heart rate curve based on the target variation value and the target acceleration number.
[0119] The acquisition module 201 is further configured to acquire the target fetal heart rate curve, and divide the target fetal heart rate curve into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition; based on the several initial fetal heart rate curve segments, obtain the linearly interpolated initial fetal heart rate curve through linear interpolation; and based on the linearly interpolated initial fetal heart rate curve, determine the initial mask through a second preset threshold and a second preset condition.
[0120] The determining module 202 is further configured to filter the target fetal heart rate curve through a median filter of a second preset duration to obtain the filtered target fetal heart rate curve; and to filter the linearly interpolated initial fetal heart rate curve through a median filter of a third preset duration to obtain the filtered initial fetal heart rate curve.
[0121] The determining module 202 is further configured to determine a target mask for each of the target fetal heart rate curve segments based on the initial mask, and to achieve this through the following expression:
[0122]
[0123]
[0124] Among them, the Mask i (t) is the first mask; the The second mask; the Mask missing (t) is the initial mask; the The target fetal heart rate curve segment.
[0125] The prediction module 204 is further configured to obtain a first variation value and a first acceleration number for the corresponding first fetal heart rate curve segment based on the first baseline fetal heart rate; and to obtain a second variation value and a second acceleration number for the corresponding second fetal heart rate curve segment based on the second baseline fetal heart rate; to obtain the target variation value based on the first variation value and the second variation value; and to obtain the target acceleration number based on the first acceleration number and the second acceleration number.
[0126] The prediction module 204 is further configured to sort the target mutation value and the target acceleration number according to a preset rule to obtain a first order and a second order; and to obtain the target anomaly degree based on the first order, the second order, the target mutation value and the target acceleration number.
[0127] The prediction module 204 is further configured to acquire a target uterine contraction curve associated with the target fetal heart rate curve; determine early deceleration values and late deceleration values based on the target uterine contraction curve; and determine the target abnormality of the target fetal heart rate curve based on the early deceleration values, the late deceleration values, the target variation value, and the target acceleration number.
[0128] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the above-described apparatus and its modules and units can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0129] The methods and apparatus of this application can be used in a wide variety of general-purpose or special-purpose computing system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer terminal devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.
[0130] For example, the above-described method and apparatus can be implemented as a computer program, which can be used in, for example... Figure 6 It runs on the computer device shown.
[0131] Please see Figure 6 , Figure 6 This is a schematic diagram of a computer device provided in an embodiment of this application. The computer device may be a server.
[0132] like Figure 6 As shown, the computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include volatile storage media, non-volatile storage media, and internal memory.
[0133] Non-volatile storage media can store an operating system and computer programs. These computer programs include program instructions that, when executed, cause the processor to perform any method for predicting the abnormality of a fetal heart rate curve.
[0134] The processor provides computing and control capabilities, supporting the operation of the entire computer device.
[0135] Internal memory provides an environment for the execution of computer programs in non-volatile storage media. When executed by a processor, the computer program enables the processor to perform any method for predicting the degree of abnormality of the fetal heart rate curve.
[0136] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that the structure of this computer device is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.
[0137] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.
[0138] In some embodiments, the processor is used to run a computer program stored in a memory to perform the following steps: acquiring a target fetal heart rate curve and determining the missing data of the target fetal heart rate curve to obtain a corresponding initial mask; dividing the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and determining a target mask for each target fetal heart rate curve segment based on the initial mask; analyzing each target fetal heart rate curve segment and the corresponding target mask using a target deep learning model to obtain a target baseline fetal heart rate; obtaining the target variation value and the target acceleration number for each target fetal heart rate curve segment based on the target baseline fetal heart rate, and obtaining the target anomaly degree of the target fetal heart rate curve based on the target variation value and the target acceleration number.
[0139] In some embodiments, the processor is further configured to acquire the target fetal heart rate curve, and divide the target fetal heart rate curve into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition; based on the several initial fetal heart rate curve segments, obtain the linearly interpolated initial fetal heart rate curve by linear interpolation; and based on the linearly interpolated initial fetal heart rate curve, determine the initial mask by a second preset threshold and a second preset condition.
[0140] In some embodiments, the processor is further configured to filter the target fetal heart rate curve through a median filter of a second preset duration to obtain the filtered target fetal heart rate curve; and to filter the linearly interpolated initial fetal heart rate curve through a median filter of a third preset duration to obtain the filtered initial fetal heart rate curve.
[0141] In some implementations, the processor is further configured to determine, based on the initial mask, a target mask for each of the target fetal heart rate curve segments by means of the following expression:
[0142]
[0143]
[0144] Among them, the Mask i (t) is the first mask; the The second mask; the Mask missing (t) is the initial mask; the The target fetal heart rate curve segment.
[0145] In some embodiments, the processor is further configured to obtain a first variation value and a first acceleration number for the corresponding first fetal heart rate curve segment based on the first baseline fetal heart rate; and to obtain a second variation value and a second acceleration number for the corresponding second fetal heart rate curve segment based on the second baseline fetal heart rate; to obtain the target variation value based on the first variation value and the second variation value; and to obtain the target acceleration number based on the first acceleration number and the second acceleration number.
[0146] In some embodiments, the processor is further configured to sort the target mutation value and the target acceleration number according to a preset rule to obtain a first order and a second order; and to obtain the target anomaly degree based on the first order, the second order, the target mutation value, and the target acceleration number.
[0147] In some embodiments, the processor is further configured to acquire a target uterine contraction curve associated with the target fetal heart rate curve; determine early deceleration values and late deceleration values based on the target uterine contraction curve; and determine a target abnormality of the target fetal heart rate curve based on the early deceleration values, the late deceleration values, the target variability value, and the target number of accelerations.
[0148] This application also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed, implement any of the methods provided in this application for predicting the abnormality of a fetal heart rate curve.
[0149] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.
[0150] Furthermore, the computer-readable storage medium may primarily include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc.
[0151] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A method for predicting the abnormality of a fetal heart rate curve, characterized in that, The method includes: Obtain the target fetal heart rate curve and determine the missing data of the target fetal heart rate curve to obtain the corresponding initial mask; The target fetal heart rate curve is divided into several target fetal heart rate curve segments of a first preset duration, and a target mask for each target fetal heart rate curve segment is determined based on the initial mask; wherein, the target mask is used to identify whether the data in the corresponding target fetal heart rate curve segment is normal data or missing data; The target baseline fetal heart rate is obtained by analyzing each target fetal heart rate curve segment and the corresponding target mask using a target deep learning model. Based on the target baseline fetal heart rate, the target variation value and the target acceleration number of each segment of the target fetal heart rate curve are obtained, and the target abnormality of the target fetal heart rate curve is obtained based on the target variation value and the target acceleration number. The target fetal heart rate curve segment includes a first fetal heart rate curve segment and a second fetal heart rate curve segment. The target mask includes a first mask corresponding to the first fetal heart rate curve segment and a second mask corresponding to the second fetal heart rate curve segment. The determination of the target mask for each target fetal heart rate curve segment based on the initial mask is achieved through the following expression: Among them, the The first mask; the The second mask; The initial mask; The target fetal heart rate curve segment.
2. The method according to claim 1, characterized in that, The process of acquiring the target fetal heart rate curve, determining the missing data of the target fetal heart rate curve, and obtaining the corresponding initial mask includes: The target fetal heart rate curve is obtained, and the target fetal heart rate curve is divided into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition. Based on several initial fetal heart rate curve segments, the initial fetal heart rate curve after linear interpolation is obtained through linear interpolation. Based on the initial fetal heart rate curve after linear interpolation, the initial mask is determined by a second preset threshold and a second preset condition.
3. The method according to claim 2, characterized in that, Before dividing the target fetal heart rate curve into several initial fetal heart rate curve segments based on a first preset threshold and a first preset condition, the method further includes: The target fetal heart rate curve is filtered by a median filter of a second preset duration to obtain the filtered target fetal heart rate curve; Before determining the initial mask based on the initial fetal heart rate curve after linear interpolation using a second preset threshold and a second preset condition, the method further includes: The initial fetal heart rate curve after linear interpolation is filtered by a median filter of a third preset duration to obtain the filtered initial fetal heart rate curve.
4. The method according to claim 1, characterized in that, The target baseline fetal heart rate includes a first baseline fetal heart rate and a second baseline fetal heart rate, the first baseline fetal heart rate corresponding to the first mask; the second baseline fetal heart rate corresponding to the second mask, and obtaining the target variation value and target acceleration number for each target fetal heart rate curve segment based on the target baseline fetal heart rate includes: Based on the first baseline fetal heart rate, the first variation value and the first number of accelerations for the corresponding first fetal heart rate curve segment are obtained; and... Based on the second baseline fetal heart rate, the second variation value and the second acceleration number of the corresponding second fetal heart rate curve segment are obtained; The target mutation value is obtained based on the first mutation value and the second mutation value, and the target acceleration number is obtained based on the first acceleration number and the second acceleration number.
5. The method according to claim 4, characterized in that, The process of obtaining the target abnormality of the target fetal heart rate curve based on the target variation value and the target acceleration number includes: The target mutation values and the target acceleration numbers are sorted according to preset rules to obtain the first order and the second order; The target anomaly degree is obtained based on the first sequence, the second sequence, the target mutation value, and the target acceleration number.
6. The method according to claim 1, characterized in that, The determination of the target abnormality of the target fetal heart rate curve based on the target variation value and the target acceleration number further includes: Obtain the target uterine contraction curve associated with the target fetal heart rate curve; Based on the target uterine contraction curve, determine the early deceleration value and the late deceleration value; The target abnormality of the target fetal heart rate curve is determined based on the early deceleration value, the late deceleration value, the target variation value, and the target acceleration number.
7. An anomaly prediction device, characterized in that, A method for predicting the abnormality of a fetal heart rate curve as described in any one of claims 1-6, wherein the device for predicting the abnormality comprises: The acquisition module is used to acquire the target fetal heart rate curve, determine the missing data of the target fetal heart rate curve, and obtain the corresponding initial mask. The determining module is used to divide the target fetal heart rate curve into several target fetal heart rate curve segments of a first preset duration, and to determine the target mask for each target fetal heart rate curve segment based on the initial mask. The analysis module is used to analyze each target fetal heart rate curve segment and the corresponding target mask through a target deep learning model to obtain the target baseline fetal heart rate. The prediction module is used to obtain the target variation value and the target acceleration number for each segment of the target fetal heart rate curve based on the target baseline fetal heart rate, and to obtain the target abnormality of the target fetal heart rate curve based on the target variation value and the target acceleration number.
8. A computer device, characterized in that, include: A memory and a processor; wherein the memory is connected to the processor for storing a program for the processor to implement the steps of the method for predicting the abnormality of a fetal heart rate curve as described in any one of claims 1-6 by running the program stored in the memory.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method for predicting the abnormality of a fetal heart rate curve as described in any one of claims 1-6.