CN115089151A

Movatterモバイル変換

Info

Publication number: CN115089151A
Application number: CN202210652017.XA
Authority: CN
Inventors: 洪申达; 王晶玉; 马婧颖; 刘国莉; 韩芳; 董霄松; 王建六; 张弛
Original assignee: Peking University
Current assignee: Peking University
Priority date: 2022-06-09
Filing date: 2022-06-09
Publication date: 2022-09-23

Abstract

The invention discloses a data quality identification and correction method and a data quality identification and correction system for a fetal heart rate signal, wherein the data quality identification and correction method for the fetal heart rate signal comprises the following steps: acquiring a fetal heart rate signal, wherein the fetal heart rate signal comprises fetal heart rate data; segmenting the fetal heart rate data by using a sliding window to obtain a plurality of fetal heart rate segments; classifying and marking the plurality of fetal heart rate fragments, and marking the fetal heart rate fragments which do not meet the preset data quality condition as low-quality fragments; acquiring the starting position of the low-quality fragment; and calculating a filling segment of the low-quality segment according to the starting position, and replacing the low-quality segment with the filling segment to obtain corrected fetal heart rate data. The technical scheme of the invention can solve the problems that in the prior art, low-quality data in fetal heart rate data are difficult to identify and correct in time, and signal misinformation is easy to generate, so that inappropriate obstetrical intervention is caused.

Description

Translated fromChinese

一种胎儿心率信号的数据质量识别纠正方法和系统A method and system for identifying and correcting data quality of fetal heart rate signal

技术领域technical field

本发明涉及数据处理技术领域，尤其涉及一种胎儿心率信号的数据质量识别纠正方法和系统。The invention relates to the technical field of data processing, in particular to a method and system for identifying and correcting data quality of fetal heart rate signals.

背景技术Background technique

在数据处理领域，胎儿健康数据是监测胎儿健康状况的重要指标；而评估妊娠和分娩过程中的胎儿健康状况，则是胎儿优生优育的重要途径，也是产妇护理的基本要求。测量胎儿心率信号是胎儿健康状况评估最常用的方法之一；通过监测胎儿心率，能够评估胎儿的运动状况、识别胎儿是否缺氧，减少胎儿酸中毒等事故的发病率。In the field of data processing, fetal health data is an important indicator for monitoring fetal health; evaluating fetal health during pregnancy and delivery is an important approach to fetal prenatal and postnatal care, as well as a basic requirement for maternal care. Measuring fetal heart rate signal is one of the most commonly used methods for fetal health assessment; by monitoring fetal heart rate, fetal movement status can be assessed, fetal hypoxia can be identified, and the incidence of fetal acidosis and other accidents can be reduced.

尽管胎儿心率信号的监测技术已被广泛使用，但它仍存在较大的局限性，主要表现在采集的信号质量方面。通常，胎儿心率信号是通过使用放置在孕妇腹部的皮肤电极或探头记录得到，且记录时间通常为半小时左右。由于孕妇或胎儿的运动会影响皮肤电极或探头的接触稳定性，因此，胎儿心率信号很容易受到干扰和噪声的污染，导致信号的数据质量明显下降，胎儿心率信号容易出现低质量片段。这些低质量片段会降低胎儿心率信号的测量准确性，可能会产生大量假阳性的误报，造成错误的运动健康或缺氧报警，夸大胎儿酸中毒的风险，并错误地导致剖宫产率的急剧增加。Although the monitoring technology of fetal heart rate signal has been widely used, it still has major limitations, mainly in the quality of the collected signal. Usually, fetal heart rate signals are recorded using skin electrodes or probes placed on the abdomen of the pregnant woman, and the recording time is usually about half an hour. Because the movement of the pregnant woman or the fetus will affect the contact stability of the skin electrodes or probes, the fetal heart rate signal is easily polluted by interference and noise, resulting in a significant decrease in the data quality of the signal, and low-quality fragments of the fetal heart rate signal are prone to occur. These low-quality segments reduce the measurement accuracy of fetal heart rate signals, may generate a large number of false positives, create false alarms for exercise fitness or hypoxia, exaggerate the risk of fetal acidosis, and falsely lead to cesarean section rates. Dramatic increase.

因此，如何及时地识别出这些低质量数据并进行纠正，是减少胎儿心率信号误报，降低不适当的产科干预的重要需求，也是本领域技术人员亟待解决的技术问题。Therefore, how to identify and correct these low-quality data in a timely manner is an important requirement for reducing false alarms of fetal heart rate signals and inappropriate obstetric intervention, and is also a technical problem to be solved by those skilled in the art.

发明内容SUMMARY OF THE INVENTION

本发明提供一种胎儿心率信号的数据质量识别纠正方法和系统，旨在解决现有技术中难以及时识别出胎儿心率数据中的低质量数据并纠正，容易产生信号误报，造成不适当的产科干预的问题。The present invention provides a method and system for identifying and correcting data quality of fetal heart rate signals, aiming to solve the problem that in the prior art, it is difficult to identify and correct low-quality data in fetal heart rate data in time, and it is easy to generate signal false alarms and cause inappropriate obstetrics. the problem of intervention.

为实现上述目的，根据本发明的第一方面，本发明提出了一种胎儿心率信号的数据质量识别纠正方法，包括：In order to achieve the above object, according to the first aspect of the present invention, the present invention proposes a method for identifying and correcting data quality of fetal heart rate signals, including:

获取胎儿心率信号，其中，胎儿心率信号包括胎儿心率数据；acquiring a fetal heart rate signal, wherein the fetal heart rate signal includes fetal heart rate data;

使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段；Use a sliding window to segment fetal heart rate data to obtain multiple fetal heart rate segments;

根据预设数据质量条件对多个胎儿心率片段进行二分类识别，将不符合预设数据质量条件的胎儿心率片段识别为低质量片段；Perform binary identification of multiple fetal heart rate segments according to preset data quality conditions, and identify fetal heart rate segments that do not meet preset data quality conditions as low-quality segments;

获取低质量片段的起始位置；Get the starting position of the low-quality fragment;

根据起始位置计算低质量片段的填补片段，使用填补片段替换低质量片段，得到纠正后的胎儿心率数据。Calculate the padding segment of the low-quality segment based on the starting position, and replace the low-quality segment with the padding segment to obtain corrected fetal heart rate data.

优选的，上述数据质量识别纠正方法中，使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段的步骤包括：Preferably, in the above data quality identification and correction method, the steps of using a sliding window to segment the fetal heart rate data to obtain multiple fetal heart rate segments include:

选取预定窗口长度的滑动窗口；Select a sliding window with a predetermined window length;

按照预定步长，使用滑动窗口顺序滑动切分胎儿心率数据；According to the predetermined step size, use the sliding window to sequentially slide and segment the fetal heart rate data;

获取切分后的多个胎儿心率片段。Acquire multiple fetal heart rate segments after segmentation.

优选的，上述数据质量识别纠正方法中，对多个胎儿心率片段进行二分类识别的步骤包括：Preferably, in the above data quality identification and correction method, the step of performing two-class identification on a plurality of fetal heart rate segments includes:

使用预设数据质量条件校验多个胎儿心率片段，分别判断每个胎儿心率片段是否满足预设数据质量条件；Verifying multiple fetal heart rate segments using preset data quality conditions, and separately determining whether each fetal heart rate segment satisfies the preset data quality conditions;

若胎儿心率片段满足预设数据质量条件，则标记胎儿心率片段为高质量片段；或者，If the fetal heart rate segment meets the preset data quality condition, the fetal heart rate segment is marked as a high-quality segment; or,

若胎儿心率片段不满足预设数据质量条件，则标记胎儿心率片段为低质量片段。If the fetal heart rate segment does not meet the preset data quality condition, the fetal heart rate segment is marked as a low-quality segment.

优选的，上述数据质量识别纠正方法中，获取低质量片段的起始位置的步骤包括：Preferably, in the above data quality identification and correction method, the step of obtaining the starting position of the low-quality segment includes:

获取胎儿心率数据中的所有每个采样心率；Get all each sampled heart rate in fetal heart rate data;

针对所述所有采样心率中的任一采样心率，获取包含所述任一采样心率的所有胎儿心率片段；For any sample heart rate among all the sample heart rates, obtain all fetal heart rate segments including the any sample heart rate;

计算胎儿心率片段中低质量片段的占比；Calculate the proportion of low-quality fragments in fetal heart rate fragments;

判断所述胎儿心率片段中低质量片段的占比是否大于或等于预设比例阈值；Determine whether the proportion of low-quality segments in the fetal heart rate segments is greater than or equal to a preset proportion threshold;

若低质量片段的占比大于或等于预设比例阈值，则将所述胎儿心率片段的起始位置作为低质量片段的起始位置。If the proportion of the low-quality segments is greater than or equal to the preset proportion threshold, the starting position of the fetal heart rate segment is used as the starting position of the low-quality segment.

优选的，上述数据质量识别纠正方法中，根据起始位置计算低质量片段的填补片段的步骤包括：Preferably, in the above-mentioned data quality identification and correction method, the step of calculating the padding segment of the low-quality segment according to the starting position includes:

使用线性插值法，计算低质量片段的线性插值片段；Use linear interpolation to calculate linearly interpolated segments of low-quality segments;

构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段；以及，constructing a forward time series forecasting model, and using the forward time series forecasting model to predict forward time series forecasting segments of low-quality segments; and,

构建后向时间序列预测模型，使用后向时间序列预测模型预测得到低质量片段的后向时间序列预测片段；Build a backward time series prediction model, and use the backward time series prediction model to predict the backward time series prediction segments of low-quality segments;

对线性插值片段、前向时间序列预测片段和后向时间序列预测片段求平均计算，得到填补片段。The interpolation segment is obtained by averaging the linear interpolation segment, the forward time series prediction segment, and the backward time series prediction segment.

优选的，上述数据质量识别纠正方法中，构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段的步骤，包括：Preferably, in the above data quality identification and correction method, the steps of constructing a forward time series prediction model, and using the forward time series prediction model to predict and obtain the forward time series prediction segments of the low-quality segments, include:

构建前向时间序列预测模型，使用低质量片段的前向预定数量的采样心率训练前向时间序列预测模型；Build a forward time series prediction model, and train the forward time series prediction model using a predetermined number of sample heart rates of low-quality segments in the forward direction;

当前向时间序列预测模型训练完毕时，输入低质量片段的前一采样心率；When the training of the forward time series prediction model is completed, input the previous sampling heart rate of the low-quality segment;

使用前向时间序列预测模型根据前一采样心率迭代预测前向时间序列预测片段，得到前向时间序列预测片段的所有心率数据。Use the forward time series prediction model to iteratively predict the forward time series prediction segment according to the previous sample heart rate, and obtain all the heart rate data of the forward time series prediction segment.

根据本发明的第二方面，本发明还提供了一种胎儿心率信号的数据质量识别纠正系统，包括：According to the second aspect of the present invention, the present invention also provides a data quality identification and correction system for fetal heart rate signals, including:

信号获取模块，用于获取胎儿心率信号，其中，胎儿心率信号包括胎儿心率数据；a signal acquisition module for acquiring a fetal heart rate signal, wherein the fetal heart rate signal includes fetal heart rate data;

数据切分模块，用于使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段；The data segmentation module is used to segment fetal heart rate data using a sliding window to obtain multiple fetal heart rate segments;

分类标记模块，用于根据预设数据质量条件对多个胎儿心率片段进行二分类，将不符合预设数据质量条件的胎儿心率片段标记为低质量片段；The classification and marking module is used to classify the multiple fetal heart rate segments according to the preset data quality conditions, and mark the fetal heart rate segments that do not meet the preset data quality conditions as low-quality segments;

位置获取模块，用于获取低质量片段的起始位置；The position acquisition module is used to obtain the starting position of the low-quality fragment;

片段替换模块，用于根据起始位置计算低质量片段的填补片段，使用填补片段替换低质量片段，得到纠正后的胎儿心率数据。The segment replacement module is used to calculate the padding segment of the low-quality segment according to the starting position, and use the padding segment to replace the low-quality segment to obtain corrected fetal heart rate data.

优选的，上述数据质量识别纠正系统中，数据切分模块包括：Preferably, in the above data quality identification and correction system, the data segmentation module includes:

窗口选取子模块，用于选取预定窗口长度的滑动窗口；The window selection submodule is used to select the sliding window of the predetermined window length;

滑动切分子模块，用于按照预定步长，使用滑动窗口顺序滑动切分胎儿心率数据，获取切分后的多个胎儿心率片段。The sliding molecular segmentation module is used for sequentially sliding and segmenting fetal heart rate data by using a sliding window according to a predetermined step size, and obtaining multiple fetal heart rate segments after segmentation.

优选的，上述数据质量识别纠正系统中，分类标记模块包括：Preferably, in the above data quality identification and correction system, the classification and marking module includes:

条件判断子模块，用于使用预设数据质量条件校验多个胎儿心率片段，分别判断每个胎儿心率片段是否满足预设数据质量条件；a condition judgment sub-module, used to verify a plurality of fetal heart rate segments using preset data quality conditions, and respectively determine whether each fetal heart rate segment satisfies the preset data quality conditions;

第一片段标记子模块，用于若胎儿心率片段满足预设数据质量条件时，标记胎儿心率片段为高质量片段；或者，The first segment marking submodule is used to mark the fetal heart rate segment as a high-quality segment if the fetal heart rate segment meets the preset data quality condition; or,

第二片段标记子模块，用于若胎儿心率片段不满足预设数据质量条件时，标记胎儿心率片段为低质量片段。The second segment marking submodule is configured to mark the fetal heart rate segment as a low-quality segment if the fetal heart rate segment does not meet the preset data quality condition.

优选的，上述数据质量识别纠正系统中，位置获取模块包括：Preferably, in the above data quality identification and correction system, the location acquisition module includes:

心率获取子模块，用于获取胎儿心率数据中所有采样心率；Heart rate acquisition sub-module, used to acquire all sampled heart rates in fetal heart rate data;

占比计算子模块，针对所述所有采样心率中任一采样心率，获取包含该任一采样心率的所有胎儿心率片段，计算胎儿心率片段中低质量片段的占比；The proportion calculation submodule, for any sampling heart rate among all the sampling heart rates, obtains all fetal heart rate segments including the sampling heart rate, and calculates the proportion of low-quality segments in the fetal heart rate segments;

阈值判断子模块，用于判断所述胎儿心率片段中低质量片段的占比是否大于或等于预设比例阈值；a threshold judgment submodule, used for judging whether the proportion of low-quality segments in the fetal heart rate segments is greater than or equal to a preset proportion threshold;

起始位置获取子模块，用于若低质量片段的占比大于或等于预设比例阈值时，将胎儿心率片段的起始位置作为低质量片段的起始位置。The starting position obtaining submodule is used for taking the starting position of the fetal heart rate segment as the starting position of the low-quality segment if the proportion of the low-quality segments is greater than or equal to the preset proportion threshold.

优选的，上述数据质量识别纠正系统中，片段替换模块包括：Preferably, in the above-mentioned data quality identification and correction system, the segment replacement module includes:

插值片段计算子模块，用于使用线性插值法，计算低质量片段的线性插值片段；Interpolated segment calculation sub-module, used to calculate linear interpolation segments of low-quality segments using linear interpolation;

预测模型构建子模块，用于构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段；以及构建后向时间序列预测模型，使用后向时间序列预测模型预测得到低质量片段的后向时间序列预测片段；The forecasting model building sub-module is used to build a forward time series forecasting model, and use the forward time series forecasting model to predict the forward time series forecasting segments of low-quality segments; and build a backward time series forecasting model, using backward time series forecasting The prediction model predicts the backward time series prediction segment of the low-quality segment;

片段平均计算子模块，用于对线性插值片段、前向时间序列预测片段和后向时间序列预测片段求平均计算，得到填补片段。The segment average calculation sub-module is used to average the linear interpolation segment, the forward time series prediction segment and the backward time series prediction segment to obtain the fill segment.

综上，上述胎儿心率信号的数据质量识别纠正方案，通过获取胎儿心率数据并使用滑动窗口进行切分，得到多个胎儿心率片段，每个胎儿心率片段都包含一定时段内的胎儿心率，然后根据预设数据质量条件对上述胎儿心率片段进行二分类，这样符合上述预设数据质量条件的胎儿心率片段即高质量片段，而不符合上述预设数据质量条件的胎儿心率片段则被标记为低质量片段，此时获取低质量片段的起始位置，从该起始位置开始计算低质量片段的采样心率的填补替代心率，就得到包含所有填补替代心率的填补片段，使用该填补片段替换上述低质量片段，就能够纠正低质量的胎儿心率数据，得到纠正后的高质量的胎儿心率数据。综上，通过上述方式能够解决现有技术难以及时识别出胎儿心率数据中低质量数据并予以纠正，低质量的胎儿心率数据容易产生信号误报，进而造成不适当的产科干预的问题。To sum up, the above-mentioned data quality identification and correction scheme of fetal heart rate signal obtains multiple fetal heart rate segments by acquiring fetal heart rate data and using a sliding window for segmentation, each fetal heart rate segment contains fetal heart rate within a certain period, and then according to The preset data quality conditions are used to classify the above-mentioned fetal heart rate fragments into two categories, so that the fetal heart rate fragments that meet the above-mentioned preset data quality conditions are high-quality fragments, and the fetal heart rate fragments that do not meet the above-mentioned preset data quality conditions are marked as low-quality segment, at this time, the starting position of the low-quality segment is obtained, and the fill-in replacement heart rate of the sampling heart rate of the low-quality segment is calculated from the starting position, and the fill-in segment containing all the fill-in replacement heart rates is obtained, and the fill-in segment is used to replace the above low-quality segment. Fragments, low-quality fetal heart rate data can be corrected to obtain corrected high-quality fetal heart rate data. In conclusion, the above methods can solve the problem that the prior art is difficult to identify and correct low-quality fetal heart rate data in a timely manner, and low-quality fetal heart rate data is prone to signal false alarms, thereby causing inappropriate obstetric intervention.

附图说明Description of drawings

为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图示出的结构获得其他的附图。In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can also be obtained according to the structures shown in these drawings without creative efforts.

图1是本发明实施例提供的一种胎儿心率信号的数据质量识别纠正方法方法的流程示意图；1 is a schematic flowchart of a method for identifying and correcting data quality of a fetal heart rate signal according to an embodiment of the present invention;

图2是图1所示实施例提供的一种胎儿心率数据的切分方法的流程示意图；2 is a schematic flowchart of a method for segmenting fetal heart rate data provided by the embodiment shown in FIG. 1;

图3是图1所示实施例提供的一种胎儿心率片段的分类标记方法的流程示意图；3 is a schematic flowchart of a method for classifying and labeling fetal heart rate segments provided by the embodiment shown in FIG. 1;

图4是图3所示实施例提供的一种胎儿心率片段判断方法的流程示意图；4 is a schematic flowchart of a method for determining a fetal heart rate segment provided by the embodiment shown in FIG. 3;

图5是图1所示实施例提供的一种低质量片段的起始位置的获取方法的流程示意图；FIG. 5 is a schematic flowchart of a method for acquiring a starting position of a low-quality segment provided by the embodiment shown in FIG. 1;

图6是图1所示实施例提供的一种填补片段的计算方法的流程示意图；FIG. 6 is a schematic flowchart of a calculation method for filling segments provided by the embodiment shown in FIG. 1;

图7是图6所示实施例提供的一种前向时间序列预测片段的预测方法的流程示意图；7 is a schematic flowchart of a method for predicting a forward time series prediction segment provided by the embodiment shown in FIG. 6;

图8是本发明实施例提供的一种胎儿心率信号的数据质量识别纠正系统的结构示意图；8 is a schematic structural diagram of a system for identifying and correcting data quality of a fetal heart rate signal according to an embodiment of the present invention;

图9是图8所示实施例提供的一种数据切分模块的结构示意图；9 is a schematic structural diagram of a data segmentation module provided by the embodiment shown in FIG. 8;

图10是图8所示实施例提供的一种分类标记模块的结构示意图；10 is a schematic structural diagram of a classification and marking module provided by the embodiment shown in FIG. 8;

图11是图8所示实施例提供的一种位置获取模块的结构示意图；11 is a schematic structural diagram of a location acquisition module provided by the embodiment shown in FIG. 8;

图12是图8所示实施例提供的一种片段填补模块的结构示意图；12 is a schematic structural diagram of a segment filling module provided by the embodiment shown in FIG. 8;

图13-a是本发明实施例提供的一种原始的胎儿心率数据的示意图；Fig. 13-a is a schematic diagram of raw fetal heart rate data provided by an embodiment of the present invention;

图13-b是本发明实施例提供的一种低质量片段的示意图；13-b is a schematic diagram of a low-quality fragment provided by an embodiment of the present invention;

图13-c是本发明实施例提供的一种纠正后的胎儿心率数据的示意图。FIG. 13-c is a schematic diagram of corrected fetal heart rate data provided by an embodiment of the present invention.

本发明目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。The realization, functional characteristics and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments.

具体实施方式Detailed ways

应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

本发明实施例的主要解决的技术问题是：The main technical problem solved by the embodiment of the present invention is:

胎儿心率信号很容易遭受周边信号干扰和噪声的污染，导致信号的数据质量明显下降，胎儿心率信号容易出现低质量片段。这些低质量片段会降低胎儿心率信号的测量准确性，可能会产生大量假阳性的误报，造成错误的运动健康或缺氧报警，夸大胎儿酸中毒的风险，并错误地导致剖宫产率的急剧增加。The fetal heart rate signal is easily polluted by surrounding signal interference and noise, which leads to a significant decrease in the data quality of the signal, and the fetal heart rate signal is prone to low-quality fragments. These low-quality segments reduce the measurement accuracy of fetal heart rate signals, may generate a large number of false positives, create false alarms for exercise fitness or hypoxia, exaggerate the risk of fetal acidosis, and falsely lead to cesarean section rates. Dramatic increase.

为了解决上述问题，及时识别这些低质量片段，减少胎儿心率信号误报，降低不适当的产科干预。本发明下述实施例提供了胎儿心率信号的数据质量识别纠正方案，通过对胎儿心率数据进行切分，得到多个胎儿心率片段，根据预设数据质量条件对胎儿心率片段进行二分类，识别并标记低质量片段，这样就能够及时得到识别到胎儿心率信号包含的所有低质量片段，然后通过获取低质量片段的起始位置，根据该起始位置计算得到低质量片段的填补片段，使用该填补片段替换上述低质量片段，从而得到纠正后的胎儿心率数据，提高胎儿心率数据的质量，减少胎儿心率信号的误报，进而减少不适当的产科干预。In order to solve the above problems, identify these low-quality fragments in time, reduce the false alarm of fetal heart rate signal, and reduce inappropriate obstetric intervention. The following embodiments of the present invention provide a data quality identification and correction scheme for fetal heart rate signals. By segmenting the fetal heart rate data, a plurality of fetal heart rate segments are obtained, and the fetal heart rate segments are classified into two categories according to preset data quality conditions to identify and identify the fetal heart rate segments. Mark the low-quality segments, so that all the low-quality segments included in the fetal heart rate signal can be identified in time, and then obtain the starting position of the low-quality segment, calculate the filling segment of the low-quality segment according to the starting location, and use the padding segment. The segment replaces the above-mentioned low-quality segment, thereby obtaining corrected fetal heart rate data, improving the quality of the fetal heart rate data, reducing false alarms of the fetal heart rate signal, and thereby reducing inappropriate obstetric intervention.

本发明能够使用神经网络模型执行下述胎儿心率信号的数据质量识别纠正方案，神经网络模型的输入是原始的胎儿心率数据，输出是低质量片段的起始位置和纠正后的胎儿心率数据。上述数据质量识别纠正方案主要包含两个关键步骤：(1)识别得到低质量片段、(2)对低质量片段进行纠正。The present invention can use a neural network model to perform the following data quality identification and correction scheme for fetal heart rate signals. The input of the neural network model is the original fetal heart rate data, and the output is the starting position of the low-quality segment and the corrected fetal heart rate data. The above data quality identification and correction scheme mainly includes two key steps: (1) identifying low-quality segments, and (2) correcting low-quality segments.

为实现上述目的，请参见图1，图1是本发明实施例提供的第一种胎儿心率信号的数据质量识别纠正方法的流程示意图，如图1所示，该胎儿心率信号的数据质量识别纠正方法包括以下步骤：In order to achieve the above purpose, please refer to FIG. 1. FIG. 1 is a schematic flowchart of a first method for identifying and correcting data quality of a fetal heart rate signal provided by an embodiment of the present invention. As shown in FIG. 1, the data quality of the fetal heart rate signal is identified and corrected. The method includes the following steps:

S110：获取胎儿心率信号，其中，胎儿心率信号包括胎儿心率数据。本申请实施例可以使用神经网络的相关模型，输入上述胎儿心率数据，神经网络模型输出的即低质量片段的起始位置和纠正后的胎儿心率数据。本申请实施例获取的胎儿心率信号是为一段时间内的胎儿心率数据，具体在输入神经网络模型时能够序列化该胎儿心率数据，形成胎儿心率序列。具体地，记输入的原始胎儿心率数据为X∈Rⁿ，Rⁿ表示长度为n的实数序列。S110: Acquire a fetal heart rate signal, where the fetal heart rate signal includes fetal heart rate data. In this embodiment of the present application, a correlation model of a neural network can be used to input the above-mentioned fetal heart rate data, and the output of the neural network model is the starting position of the low-quality segment and the corrected fetal heart rate data. The fetal heart rate signal obtained in the embodiment of the present application is fetal heart rate data within a period of time, and specifically, the fetal heart rate data can be serialized when inputting the neural network model to form a fetal heart rate sequence. Specifically, the input original fetal heart rate data is denoted as X∈Rⁿ , where Rⁿ represents a real number sequence of length n.

S120：使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段；神经网络模型包括滑动窗口，使用该滑动窗口切分序列化的胎儿心率数据，从而得到多个胎儿心率片段。S120: Use a sliding window to segment the fetal heart rate data to obtain multiple fetal heart rate segments; the neural network model includes a sliding window, and use the sliding window to segment the serialized fetal heart rate data to obtain multiple fetal heart rate segments.

作为一种优选的实施例，如图2所示，上述数据质量识别纠正方法中，步骤S120：使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段的步骤包括：As a preferred embodiment, as shown in FIG. 2, in the above data quality identification and correction method, step S120: using a sliding window to segment the fetal heart rate data, and the steps of obtaining multiple fetal heart rate segments include:

S121：选取预定窗口长度的滑动窗口。S121: Select a sliding window with a predetermined window length.

S122：按照预定步长，使用滑动窗口顺序滑动切分胎儿心率数据。S122: According to a predetermined step size, use a sliding window to sequentially slide and segment the fetal heart rate data.

S123：获取切分后的多个胎儿心率片段。S123: Acquire a plurality of segmented fetal heart rate segments.

具体地，使用滑动窗口对原始的胎儿心率数据进行切分，选取窗口长度为d，步长为s，得到k个长度为d的胎儿心率片段，其中上述k个长度为d的胎儿心率片段中，第i个片段记为X_i∈R^d。Specifically, a sliding window is used to segment the original fetal heart rate data, the window length is d, and the step size is s, to obtain k fetal heart rate segments of length d, wherein among the above k fetal heart rate segments of length d , the i-th segment is denoted as X_i ∈ R^d .

在切分胎儿心率数据，得到多个胎儿心率片段后，图1所示的数据质量识别纠正方法还包括：After segmenting the fetal heart rate data and obtaining multiple fetal heart rate segments, the data quality identification and correction method shown in Figure 1 further includes:

S130：根据预设数据质量条件对多个胎儿心率片段进行二分类识别，将不符合预设数据质量条件的胎儿心率片段识别并标记为低质量片段，从而完成低质量片段的评估。其中，原始的胎儿心率数据如图13-a所示；识别得到的低质量片段如图13-b所示。S130: Perform two-class identification on a plurality of fetal heart rate segments according to preset data quality conditions, identify and mark fetal heart rate segments that do not meet the preset data quality conditions as low-quality segments, so as to complete the evaluation of the low-quality segments. Among them, the original fetal heart rate data is shown in Figure 13-a; the identified low-quality segments are shown in Figure 13-b.

本申请实施例中，能够使用神经网络模型对上述多个胎儿心率片段分别进行卷积操作，然后将卷积后的胎儿心率片段输入至神经网络模型的全连接层MLP，根据上述预设数据质量条件对上述胎儿心率片段进而二分类标记，具体地符合预设数据质量条件的胎儿心率片段为高质量片段，不符合上述预设数据质量条件的胎儿心率片段为低质量片段。In the embodiment of the present application, a neural network model can be used to perform convolution operations on the above-mentioned multiple fetal heart rate segments respectively, and then the convoluted fetal heart rate segments can be input into the fully connected layer MLP of the neural network model. According to the above preset data quality The conditions further classify and label the above-mentioned fetal heart rate segments. Specifically, fetal heart rate segments that meet the preset data quality conditions are high-quality segments, and fetal heart rate segments that do not meet the preset data quality conditions are low-quality segments.

具体作为一种优选的实施例，如图3所示，该根据预设数据质量条件对多个胎儿心率片段进行二分类识别，将不符合预设数据质量条件的胎儿心率片段识别并标记为低质量片段的步骤包括：Specifically, as a preferred embodiment, as shown in FIG. 3 , the multiple fetal heart rate segments are classified into two categories according to the preset data quality conditions, and the fetal heart rate segments that do not meet the preset data quality conditions are identified and marked as low The steps to quality clips include:

S131：使用预设数据质量条件校验多个胎儿心率片段，分别判断每个胎儿心率片段是否满足预设数据质量条件；S131: Verify multiple fetal heart rate segments using preset data quality conditions, and determine whether each fetal heart rate segment satisfies the preset data quality conditions;

S132：若胎儿心率片段满足预设数据质量条件，则标记胎儿心率片段为高质量片段；或者，S132: If the fetal heart rate segment meets the preset data quality condition, mark the fetal heart rate segment as a high-quality segment; or,

S133：若胎儿心率片段不满足预设数据质量条件，则标记胎儿心率片段为低质量片段。S133: If the fetal heart rate segment does not meet the preset data quality condition, mark the fetal heart rate segment as a low-quality segment.

其中，预设数据质量条件能够设置为多个，本申请实施例中对每个长度为d的胎儿心率片段X_i进行信号质量二分类。若X_i同时满足以下3个预设数据质量条件，则记X_i为高质量片段(标记为1)，否则，任一条不满足则记X_i为低质量片段(标记为0)。Wherein, a plurality of preset data quality conditions can be set, and in the embodiment of the present application, the signal quality of each fetal heart rate segment X_i of length d is classified into two. If Xi_meets the following three preset data quality conditions at the same time, mark_Xi as a high-quality segment (marked as 1); otherwise, if any one of them does not satisfy, mark_Xi as a low-quality segment (marked as 0).

其中，本申请实施例中设置预设数据质量条件如下：Wherein, the preset data quality conditions are set as follows in the embodiment of the present application:

(1)平均心率是否在80次/分钟到200次/分钟之间，即：80≤mean(X_i)≤200；(1) Whether the average heart rate is between 80 beats/min and 200 beats/min, that is: 80≤mean(X_i)≤200;

(2)最低心率是否大于60次/分钟，即：min(X_i)>60；(2) Whether the lowest heart rate is greater than 60 beats/min, namely: min(X_i)>60;

(3)最高心率/最低心率是否小于1.5，即：max(X_i)/min(X_i)<1.5。(3) Whether the maximum heart rate/minimum heart rate is less than 1.5, namely: max(X_i)/min(X_i)<1.5.

整体胎儿群体中，通常，(1)心率的可能范围是80到200；(2)瞬时心率最低不应该低于60次/分钟；(3)窗口长度d内心率的变化量不应该太剧烈，因此，这里取的是最大心率比最小心率小于1.5。In the overall fetal population, generally, (1) the possible range of heart rate is 80 to 200; (2) the minimum instantaneous heart rate should not be lower than 60 beats/min; (3) the variation of heart rate within the window length d should not be too drastic, Therefore, the maximum heart rate is taken here to be less than 1.5 than the minimum heart rate.

具体的每个胎儿心率片段是否满足预设数据质量条件的判断流程如图4所示：The specific judgment process of whether each fetal heart rate segment meets the preset data quality conditions is shown in Figure 4:

S1331：输入：长度为d的胎儿心率片段。S1331: Input: fetal heart rate segment of length d.

S1332：判断平均心率是否在80次/分钟到200次/分钟之间；若是，则执行步骤S1333；若否，则输出：低质量。S1332: Determine whether the average heart rate is between 80 beats/min and 200 beats/min; if so, execute step S1333; if not, output: low quality.

S1333：判断最低心率是否大于60次/分钟；若是，则执行步骤S1334；若否，则输出：低质量。S1333: Determine whether the minimum heart rate is greater than 60 beats/min; if so, execute step S1334; if not, output: low quality.

S1334：判断最高心率/最低心率是否小于1.5；若是，则输出：高质量；若否，则输出：低质量。S1334: Determine whether the maximum heart rate/minimum heart rate is less than 1.5; if so, output: high quality; if not, output: low quality.

如图13-b所示，本申请实施例识别得到的图13-b中阴影区域的胎儿心率数据即低质量片段。As shown in Fig. 13-b, the fetal heart rate data in the shaded area in Fig. 13-b identified by the embodiment of the present application is a low-quality segment.

在对多个胎儿心率片段进行二分类后能够得到K个分类结果，K属于实数，本申请实施例需要整合K个分类结果，得到低质量片段的起始位置；因此图1所示的数据质量识别纠正方法包括：K classification results can be obtained after performing binary classification on multiple fetal heart rate segments, and K is a real number. In this embodiment of the present application, K classification results need to be integrated to obtain the starting positions of low-quality segments; therefore, the data quality shown in FIG. 1 Identifying corrections includes:

S140：获取低质量片段的起始位置。其中，对于原始胎儿心率数据X∈Rⁿ中的每个采样点X[i]，计算包含X[i]的所有胎儿心率片段中，被标记为低质量片段的比例P[i]，得到P∈[0,1]ⁿ。给定预设比例阈值thresh，寻找低质量片段的比例P大于阈值thresh的胎儿心率片段，则这些胎儿心率片段对应的起始位置即为低质量片段的起始位置。在本发明实施例中阈值thresh设置为2/3。S140: Obtain the starting position of the low-quality segment. Among them, for each sampling point X[ⁱ ] in the original fetal heart rate data X∈Rn, calculate the proportion P[i] of all fetal heart rate segments that contain X[i] that are marked as low-quality segments, and obtain P ∈[0,1]ⁿ . Given a preset ratio threshold thresh, find fetal heart rate fragments whose ratio P of low-quality fragments is greater than the threshold thresh, and the corresponding starting positions of these fetal heart rate fragments are the starting positions of the low-quality fragments. In this embodiment of the present invention, the threshold thresh is set to 2/3.

作为一种优选的实施例，如图5所示，上述获取低质量片段的起始位置的步骤包括：As a preferred embodiment, as shown in FIG. 5 , the above step of obtaining the starting position of the low-quality segment includes:

S141：获取胎儿心率数据中的所有采样心率。每个采样心率即上述采样点X[i]，原始的胎儿心率数据即X∈Rⁿ。S141: Acquire all sample heart rates in the fetal heart rate data. Each sampling heart rate is the above-mentioned sampling point X[i], and the original fetal heart rate data is^X∈Rn .

S142：针对上述所有采样心率中的任一采样心率，获取包含该任一采样心率的所有胎儿心率片段，计算胎儿心率片段中低质量片段的占比；该低质量片段的占比或者说比例为上述P[i]，P∈[0,1]ⁿ。S142: For any sampling heart rate among all the sampling heart rates above, obtain all fetal heart rate segments including the sampling heart rate, and calculate the proportion of low-quality segments in the fetal heart rate segments; the proportion or ratio of the low-quality segments is: The above P[i], P∈[0,1]ⁿ .

S143：判断胎儿心率片段中低质量片段的占比是否大于或等于预设比例阈值；S143: Determine whether the proportion of low-quality segments in the fetal heart rate segments is greater than or equal to a preset proportion threshold;

S144：若低质量片段的占比大于或等于预设比例阈值，则将胎儿心率片段的起始位置作为低质量片段的起始位置。S144: If the proportion of the low-quality segments is greater than or equal to the preset proportion threshold, use the starting position of the fetal heart rate segment as the starting location of the low-quality segment.

给定预设比例阈值thresh，寻找低质量片段的比例P大于阈值thresh的胎儿心率片段，则这些胎儿心率片段对应的起始位置即为低质量片段的起始位置；其中，预设比例阈值thresh能够设置为2/3。通过上述方法能够准确全面地查找处低质量片段，并准确地找到低质量片段的起始位置，便于从该起始位置对低质量片段进行纠正。Given a preset ratio threshold thresh, look for fetal heart rate fragments whose proportion P of low-quality fragments is greater than the threshold thresh, then the starting positions corresponding to these fetal heart rate fragments are the starting positions of the low-quality fragments; wherein, the preset ratio threshold thresh Can be set to 2/3. Through the above method, the low-quality segment can be accurately and comprehensively searched, and the starting position of the low-quality segment can be accurately found, so that the low-quality segment can be corrected from the starting position.

其中，本申请实施例输出的低质量片段的起始位置为T＝{(s_1,t_1),(s_2,t_2),...}，其中，二元组(s_i,t_i)表示第i个低质量片段的开始位置为s_i，结束位置为t_i，T是包含二元组(s_i,t_i)的集合，二元组的个数不固定。Wherein, the starting position of the low-quality segment output in the embodiment of the present application is T={(s_1, t_1), (s_2, t_2),...}, where the binary group (s_i, t_i) represents the i-th segment The starting position of the low-quality segment is s_i, and the ending position is t_i. T is a set containing two-tuples (s_i, t_i), and the number of two-tuples is not fixed.

在获取低质量片段的起始位置后，图1所示的数据质量识别纠正方法还包括以下步骤：After obtaining the starting position of the low-quality segment, the data quality identification and correction method shown in Figure 1 further includes the following steps:

S150：根据起始位置计算低质量片段的填补片段，使用填补片段替换低质量片段，得到纠正后的胎儿心率数据。该填补片段为采用上述胎儿心率片段中的非低质量片段的采样心率，结合对该低质量片段对应的心率的预测得到，因此该填补片段能够替换上述低质量片段，得到较高质量的胎儿心率数据。其中，纠正后的胎儿心率数据如图13-c所示。S150: Calculate the padding segment of the low-quality segment according to the starting position, and replace the low-quality segment with the padding segment to obtain corrected fetal heart rate data. The filling segment is obtained by using the sampling heart rate of a non-low-quality segment in the above-mentioned fetal heart rate segment, and combining with the prediction of the corresponding heart rate of the low-quality segment. Therefore, the filling segment can replace the above-mentioned low-quality segment to obtain a higher-quality fetal heart rate. data. Among them, the corrected fetal heart rate data is shown in Figure 13-c.

本申请实施例中使用低质量片段的线性插值片段、前向时间序列预测片段和后向时间序列预测片段求平均计算得到低质量片段的填补片段。其中，设低质量片段为X_j，低质量片段X_j的线性插值片段X_j⁰、前向时间序列预测片段X_j¹、后向时间序列预测片段X_j²，注意每个片段的长度均与X_j相同。In the embodiment of the present application, the low-quality segments are obtained by averaging the linear interpolation segments, the forward time series prediction segments, and the backward time series prediction segments to obtain the padding segments of the low-quality segments. Among them, let the low-quality segment be X_j , the linear interpolation segment X_j⁰ of the low-quality segment X_j , the forward time series prediction segment X_j¹ , and the backward time series prediction segment X_j² . Note that the length of each segment is equal to Same as_Xj .

作为一种优选的实施例，如图6所示，上述数据质量识别纠正方法中，步骤S150：根据起始位置计算低质量片段的填补片段的步骤包括：As a preferred embodiment, as shown in FIG. 6 , in the above data quality identification and correction method, step S150: the step of calculating the padding segment of the low-quality segment according to the starting position includes:

S151：使用线性插值法，计算低质量片段的线性插值片段。对于线性插值片段X_j⁰，采用左端点值X[s-1]，右端点值x[t+1]，长度t-s的线性插值法进行计算。具体地，X_j⁰的第i个数值X_j⁰[i]为x[t+1]+i/(t-s)(x[t+1]-X[s-1])。S151: Using a linear interpolation method, calculate the linear interpolation segment of the low-quality segment. For the linear interpolation segment X_j⁰ , the calculation is performed using the linear interpolation method of the left endpoint value X[s-1], the right endpoint value x[t+1], and the length ts. Specifically, the ith value X_j⁰ [i] of X_j⁰ is x[t+1]+i/(ts)(x[t+1]-X[s-1]).

S152：构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段。对于前向时间序列预测片段X_j¹，通过构建前向时间序列预测模型F对低质量片段的前向时间预测片段进行计算，模型F输入i-1时刻的数值，预测i时刻的数值。F可以是任何一种时间序列预测模型，如线性回归模型、AR(自回归模型)、MA(移动平均模型)、ARIMA(差分自回归移动平均模型)、深度神经网络等，这里不做具体限制。S152: Construct a forward time series prediction model, and use the forward time series prediction model to predict and obtain forward time series prediction segments of low-quality segments. For the forward time series prediction segment X_j¹ , the forward time sequence prediction segment of the low-quality segment is calculated by constructing a forward time series prediction model F. The model F inputs the value at time i-1 and predicts the value at time i. F can be any time series prediction model, such as linear regression model, AR (autoregressive model), MA (moving average model), ARIMA (differential autoregressive moving average model), deep neural network, etc., no specific restrictions are made here .

以及，as well as,

S153：构建后向时间序列预测模型，使用后向时间序列预测模型预测得到低质量片段的后向时间序列预测片段。S153: Construct a backward time series prediction model, and use the backward time series prediction model to predict and obtain backward time series prediction segments of low-quality segments.

与前向时间序列预测方式同理，对于后向时间序列预测片段X_j²，通过构建后向时间序列预测模型G进行计算，G输入i时刻的数值，预测i-1时刻的数值。G可以是任何一种时间序列预测模型，如线性回归模型、AR(自回归模型)、MA(移动平均模型)、ARIMA(差分自回归移动平均模型)、深度神经网络等，这里不做具体限制。Similar to the forward time series prediction method, for the backward time series prediction segment X_j² , the calculation is performed by constructing a backward time series prediction model G, where G inputs the value at time i, and predicts the value at time i-1. G can be any time series prediction model, such as linear regression model, AR (autoregressive model), MA (moving average model), ARIMA (differential autoregressive moving average model), deep neural network, etc., no specific restrictions are made here .

S154：对线性插值片段、前向时间序列预测片段和后向时间序列预测片段求平均计算，得到填补片段。具体地，将线性插值片段、前向时间序列预测片段和后向时间序列预测片段三部分进行平均，得到填补片段X_j^*＝(X_j⁰+X_j¹+X_j²)/3。S154: Calculate the average of the linear interpolation segment, the forward time series prediction segment, and the backward time series prediction segment to obtain a fill segment. Specifically, the linear interpolation segment, the forward time series prediction segment and the backward time series prediction segment are averaged to obtain the padding segment X_j^* =(X_j⁰ +X_j¹ +X_j² )/3.

使用该填补片段X_j^*替换上述低质量片段X_j。Replace the above-mentioned low-quality segment_Xj with this padding segment_Xj^* .

综上，本申请实施例提供的技术方案，对每个低质量片段不失一般性，这里记第j个低质量片段为X_j＝{X[s],X[s+1],...,X[t]}，s<t；分别计算低质量片段X_j的线性插值片段X_j⁰、前向时间序列预测片段X_j¹、后向时间序列预测片段X_j²，注意每个片段的长度均与X_j相同；To sum up, the technical solutions provided by the embodiments of the present application do not lose generality to each low-quality segment, and here the jth low-quality segment is denoted as X_j ={X[s],X[s+1],... .,X[t]}, s<t; separately calculate the linear interpolation segment X_j⁰ of the low-quality segment X_j , the forward time series prediction segment X_j¹ , and the backward time series prediction segment X_j² , pay attention to each The lengths of the fragments are all the same as X_j ;

对于线性插值片段X_j⁰，采用左端点值X[s-1]，右端点值x[t+1]，长度t-s的线性插值法进行计算。具体地，X_j⁰的第i个数值X_j⁰[i]为x[t+1]+i/(t-s)(x[t+1]-X[s-1])。For the linear interpolation segment X_j⁰ , the calculation is performed using the linear interpolation method of the left endpoint value X[s-1], the right endpoint value x[t+1], and the length ts. Specifically, the ith value X_j⁰ [i] of X_j⁰ is x[t+1]+i/(ts)(x[t+1]-X[s-1]).

对于前向时间序列预测片段X_j¹，通过构建前向时间序列预测模型F对低质量片段的前向时间预测片段进行计算，模型F输入i-1时刻的数值，预测i时刻的数值。F可以是任何一种时间序列预测模型，如线性回归模型、AR(自回归模型)、MA(移动平均模型)、ARIMA(差分自回归移动平均模型)、深度神经网络等，这里不做具体限制。For the forward time series prediction segment X_j¹ , the forward time sequence prediction segment of the low-quality segment is calculated by constructing a forward time series prediction model F. The model F inputs the value at time i-1 and predicts the value at time i. F can be any time series prediction model, such as linear regression model, AR (autoregressive model), MA (moving average model), ARIMA (differential autoregressive moving average model), deep neural network, etc., no specific restrictions are made here .

具体地，如图7所示，上述步骤S152：构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段的步骤具体包括：Specifically, as shown in FIG. 7 , the above step S152: constructing a forward time series prediction model, and the step of using the forward time series prediction model to predict and obtain the forward time series prediction segment of the low-quality segment specifically includes:

S1521：构建前向时间序列预测模型，使用低质量片段的前向预定数量的采样心率训练前向时间序列预测模型；S1521: Construct a forward time series prediction model, and train the forward time series prediction model by using the forward predetermined number of sampling heart rates of the low-quality segments;

S1522：当前向时间序列预测模型训练完毕时，输入低质量片段的前一采样心率；S1522: When the training of the forward time series prediction model is completed, input the previous sampling heart rate of the low-quality segment;

S1523：使用前向时间序列预测模型根据前一采样心率迭代预测前向时间序列预测片段，得到前向时间序列预测片段的所有心率数据。S1523: Use the forward time series prediction model to iteratively predict the forward time series prediction segment according to the previous sample heart rate, and obtain all the heart rate data of the forward time series prediction segment.

在训练阶段，使用采样点X[s]往前的K个数据点进行训练，即{X[s-K],x[s-K+1],...,x[s-1]}。K是超参数，可以根据实际情况选择。In the training phase, use the K data points before the sampling point X[s] for training, ie {X[s-K],x[s-K+1],...,x[s-1]}. K is a hyperparameter, which can be selected according to the actual situation.

在预测阶段，首先输入X[s-1]预测得到X¹[s]，再输入X¹[s]预测得到X¹[s+1]，循环该步骤，直到预测到X¹[t]。则前向时间序列预测片段X_j¹＝{X¹[s],X¹[s+1],...,X¹[t]}。In the prediction stage, first input X[s-1] to predict X¹ [s], then input X¹ [s] to predict X¹ [s+1], and loop this step until X¹ [t] is predicted. Then the forward time series prediction segment X_j¹ ={X¹ [s], X¹ [s+1], . . . , X¹ [t]}.

同理，对于使用后向时间序列预测模型预测得到低质量片段的后向时间序列预测片段的方式如上：In the same way, the method of using the backward time series prediction model to predict the backward time series prediction segment of the low quality segment is as follows:

在训练阶段，使用X[t]往后的K个数据点进行训练，即{X[t+1],x[t+2],...,x[t+K]}。K是超参数，可以根据实际情况选择。In the training phase, use the K data points after X[t] for training, ie {X[t+1],x[t+2],...,x[t+K]}. K is a hyperparameter, which can be selected according to the actual situation.

在预测阶段，首先输入X[t]预测得到X²[t-1]，再输入X²[t-1]预测得到X²[t-2]，循环该步骤，直到预测到X²[s]。则X_j²＝{X²[s],X²[s+1],...,X²[t]}。In the prediction stage, first input X[t] to predict X² [t-1], then input X² [t-1] to predict X² [t-2], and loop this step until X² [s is predicted ]. Then X_j² ={X² [s], X² [s+1], . . . , X² [t]}.

本申请实施例将线性插值法、前向时间序列预测模型和后向时间序列预测模型三部分的预测结果进行平均，得到填补片段：X_j^*＝(X_j⁰+X_j¹+X_j²)/3。。In the embodiment of the present application, the prediction results of the three parts of the linear interpolation method, the forward time series prediction model and the backward time series prediction model are averaged to obtain the filling segment: X_j^* =(X_j⁰ +X_j¹ +X_j² )/3. .

使用该填补片段X_j^*替换低质量片段X_j。Replace the low-quality segment_Xj with the padding segment_Xj^* .

综上，本申请上述实施例提供的胎儿心率信号的数据质量识别纠正方法，通过获取胎儿心率数据并使用滑动窗口进行切分，得到多个胎儿心率片段，每个胎儿心率片段都包含一定时段内的胎儿心率，然后根据预设数据质量条件对上述胎儿心率片段进行二分类，这样符合上述预设数据质量条件的胎儿心率片段即高质量片段，而不符合上述预设数据质量条件的胎儿心率片段则被标记为低质量片段，此时获取低质量片段的起始位置，从该起始位置开始计算低质量片段的采样心率的填补替代心率，就得到包含所有用于填补替代低质量心率的填补片段，使用该填补片段替换上述低质量片段，就能够纠正低质量的胎儿心率数据，得到纠正后的高质量的胎儿心率数据。综上，通过上述方式，能够解决现有技术难以及时识别出胎儿心率数据中低质量数据并予以纠正，低质量的胎儿心率数据容易产生信号误报，进而造成不适当的产科干预的问题。To sum up, the data quality identification and correction method of fetal heart rate signal provided by the above-mentioned embodiments of the present application obtains a plurality of fetal heart rate segments by acquiring fetal heart rate data and segmenting using a sliding window, and each fetal heart rate segment includes a certain period of time. The fetal heart rate segments are classified according to the preset data quality conditions, so that the fetal heart rate segments that meet the preset data quality conditions are high-quality segments, and the fetal heart rate segments that do not meet the preset data quality conditions. Then it is marked as a low-quality segment. At this time, the starting position of the low-quality segment is obtained, and the fill-in replacement heart rate of the sampling heart rate of the low-quality segment is calculated from the starting position, and the fill-in replacement heart rate containing all the fill-in and replacement low-quality heart rates is obtained. The low-quality fetal heart rate data can be corrected by using the filled-in segment to replace the above-mentioned low-quality segment, and the corrected high-quality fetal heart rate data can be obtained. To sum up, the above method can solve the problem that it is difficult for the prior art to identify and correct low-quality fetal heart rate data in a timely manner, and low-quality fetal heart rate data is prone to signal false alarms, resulting in inappropriate obstetric intervention.

基于上述方法实施例的同一构思，本发明实施例还提出了胎儿心率信号的数据质量识别纠正系统，用于实现本发明的上述方法，由于该系统实施例解决问题的原理与方法相似，因此至少具有上述实施例的技术方案所带来的所有有益效果，在此不再一一赘述。Based on the same concept of the above method embodiments, the embodiments of the present invention also propose a data quality identification and correction system for fetal heart rate signals, which is used to implement the above method of the present invention. All the beneficial effects brought by the technical solutions of the above embodiments are not repeated here.

具体如图8所示，图8为本申请实施例提供的第一种胎儿心率信号的数据质量识别纠正系统的结构示意图。如图8所示，该胎儿心率信号的数据质量识别纠正系统，包括：Specifically, as shown in FIG. 8 , FIG. 8 is a schematic structural diagram of a first system for identifying and correcting data quality of a fetal heart rate signal according to an embodiment of the present application. As shown in Figure 8, the data quality identification and correction system of the fetal heart rate signal includes:

信号获取模块110，用于获取胎儿心率信号，其中，胎儿心率信号包括胎儿心率数据；asignal acquisition module 110, configured to acquire a fetal heart rate signal, wherein the fetal heart rate signal includes fetal heart rate data;

数据切分模块120，用于使用滑动窗口切分胎儿心率数据，得到多个胎儿心率片段；adata segmentation module 120, configured to segment fetal heart rate data using a sliding window to obtain multiple fetal heart rate segments;

分类标记模块130，用于根据预设数据质量条件对多个胎儿心率片段进行二分类识别，将不符合预设数据质量条件的胎儿心率片段识别并标记为低质量片段；The classification andmarking module 130 is configured to perform two-class identification on a plurality of fetal heart rate segments according to preset data quality conditions, and identify and mark fetal heart rate segments that do not meet the preset data quality conditions as low-quality segments;

位置获取模块140，用于获取低质量片段的起始位置；aposition obtaining module 140, configured to obtain the starting position of the low-quality segment;

片段替换模块150，用于根据起始位置计算低质量片段的填补片段，使用填补片段替换低质量片段，得到纠正后的胎儿心率数据。Thesegment replacement module 150 is configured to calculate the padding segment of the low-quality segment according to the starting position, and use the padding segment to replace the low-quality segment to obtain corrected fetal heart rate data.

综上，本申请上述实施例提供的胎儿心率信号的数据质量识别纠正系统，通过信号获取模块110获取胎儿心率数据并通过数据切分模块120使用滑动窗口进行切分，得到多个胎儿心率片段，每个胎儿心率片段都包含一定时段内的胎儿心率，然后分类标记模块130根据预设数据质量条件对上述胎儿心率片段进行二分类，这样符合上述预设数据质量条件的胎儿心率片段即高质量片段，而不符合上述预设数据质量条件的胎儿心率片段则被标记为低质量片段，此时通过位置获取模块140获取低质量片段的起始位置，片段替换模块150从该起始位置开始计算低质量片段的采样心率的填补替代心率，就得到包含所有上述填补替代心率的填补片段，使用该填补片段替换上述低质量片段，就能够纠正低质量的胎儿心率数据，得到纠正后的高质量的胎儿心率数据。综上，通过上述方式，能够解决现有技术难以及时识别出胎儿心率数据中低质量数据并予以纠正，低质量的胎儿心率数据容易产生信号误报，进而造成不适当的产科干预的问题。To sum up, the data quality identification and correction system for fetal heart rate signals provided by the above-mentioned embodiments of the present application acquires fetal heart rate data through thesignal acquisition module 110 and performs segmentation through thedata segmentation module 120 using a sliding window to obtain a plurality of fetal heart rate segments, Each fetal heart rate segment includes fetal heart rate within a certain period of time, and then the classification andlabeling module 130 classifies the fetal heart rate segment according to the preset data quality condition, so that the fetal heart rate segment that meets the preset data quality condition is a high-quality segment , the fetal heart rate segments that do not meet the above-mentioned preset data quality conditions are marked as low-quality segments. At this time, the starting position of the low-quality segment is obtained by theposition acquisition module 140, and thesegment replacement module 150 starts from the starting position to calculate the low-quality segment. The sampling heart rate of the high-quality segment can be filled in to replace the heart rate, and a filled-in segment containing all the above-mentioned filled-in replacement heart rates can be obtained. Using the filled-in segment to replace the above-mentioned low-quality segment, the low-quality fetal heart rate data can be corrected, and the corrected high-quality fetal heart rate can be obtained. Heart rate data. To sum up, the above method can solve the problem that it is difficult for the prior art to identify and correct low-quality fetal heart rate data in a timely manner, and low-quality fetal heart rate data is prone to signal false alarms, resulting in inappropriate obstetric intervention.

作为一种优选的实施例，如图9所示，上述数据质量识别纠正系统中数据切分模块120包括：As a preferred embodiment, as shown in FIG. 9 , thedata segmentation module 120 in the above-mentioned data quality identification and correction system includes:

窗口选取子模块121，用于选取预定窗口长度的滑动窗口；Thewindow selection submodule 121 is used to select a sliding window of a predetermined window length;

滑动切分子模块122，用于按照预定步长，使用滑动窗口顺序滑动切分胎儿心率数据，获取切分后的多个胎儿心率片段。The slidingmolecular segmentation module 122 is used for sequentially slidingly segmenting the fetal heart rate data by using a sliding window according to a predetermined step size, and obtaining a plurality of segmented fetal heart rate segments.

作为一种优选的实施例，如图10所示，上述数据质量识别纠正系统中，分类标记模块130包括：As a preferred embodiment, as shown in FIG. 10 , in the above data quality identification and correction system, the classification andmarking module 130 includes:

条件判断子模块131，用于使用预设数据质量条件校验多个胎儿心率片段，分别判断每个胎儿心率片段是否满足预设数据质量条件；Thecondition judgment sub-module 131 is used to verify a plurality of fetal heart rate segments using preset data quality conditions, and respectively determine whether each fetal heart rate segment satisfies the preset data quality conditions;

第一片段标记子模块132，用于若胎儿心率片段满足预设数据质量条件时，标记胎儿心率片段为高质量片段；或者，The firstsegment marking submodule 132 is configured to mark the fetal heart rate segment as a high-quality segment if the fetal heart rate segment meets the preset data quality condition; or,

第二片段标记子模块133，用于若胎儿心率片段不满足预设数据质量条件时，标记胎儿心率片段为低质量片段。The second segment marking sub-module 133 is configured to mark the fetal heart rate segment as a low-quality segment if the fetal heart rate segment does not meet the preset data quality condition.

作为一种优选的实施例，如图11所示，上述数据质量识别纠正系统中，位置获取模块140包括：As a preferred embodiment, as shown in FIG. 11 , in the above data quality identification and correction system, thelocation acquisition module 140 includes:

心率获取子模块141，用于获取胎儿心率数据中的所有每个采样心率；a heartrate acquisition sub-module 141, configured to acquire all sample heart rates in the fetal heart rate data;

占比计算子模块142，针对上述所有采样心率中的任一采样心率，获取包含该任一采样心率的所有胎儿心率片段，计算胎儿心率片段中低质量片段的占比；Theproportion calculation sub-module 142, for any sampling heart rate among all the above-mentioned sampling heart rates, obtains all fetal heart rate segments including the any sampling heart rate, and calculates the proportion of low-quality segments in the fetal heart rate segments;

阈值判断子模块143，用于判断胎儿心率片段中低质量片段的占比是否大于或等于预设比例阈值；Threshold judging sub-module 143, for judging whether the proportion of low-quality segments in the fetal heart rate segment is greater than or equal to a preset proportion threshold;

起始位置获取子模块144，用于若低质量片段的占比大于或等于预设比例阈值时，将胎儿心率片段的起始位置作为低质量片段的起始位置。The starting position obtaining sub-module 144 is configured to use the starting position of the fetal heart rate segment as the starting position of the low-quality segment if the proportion of the low-quality segments is greater than or equal to a preset ratio threshold.

作为一种优选的实施例，如图12所示，上述数据质量识别纠正系统中，片段替换模块150包括：As a preferred embodiment, as shown in FIG. 12 , in the above data quality identification and correction system, thesegment replacement module 150 includes:

插值片段计算子模块151，用于使用线性插值法，计算低质量片段的线性插值片段；an interpolationsegment calculation sub-module 151, configured to use a linear interpolation method to calculate the linear interpolation segment of the low-quality segment;

预测模型构建子模块152，用于构建前向时间序列预测模型，使用前向时间序列预测模型预测得到低质量片段的前向时间序列预测片段；以及构建后向时间序列预测模型，使用后向时间序列预测模型预测得到低质量片段的后向时间序列预测片段；The predictionmodel construction submodule 152 is used for constructing a forward time series prediction model, using the forward time series prediction model to predict the forward time series prediction segments of low-quality segments; and constructing a backward time series prediction model, using the backward time series prediction model The sequence prediction model predicts the backward time series prediction segment of the low-quality segment;

片段平均计算子模块153，用于对线性插值片段、前向时间序列预测片段和后向时间序列预测片段求平均计算，得到填补片段。The segmentaverage calculation sub-module 153 is configured to perform an average calculation on the linear interpolation segment, the forward time series prediction segment and the backward time series prediction segment to obtain the fill segment.

本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

应当注意的是，在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的部件或步骤。位于部件之前的单词“一”或“一个”不排除存在多个这样的部件。本发明可以借助于包括有若干不同部件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。It should be noted that, in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not preclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. do not denote any order. These words can be interpreted as names.

尽管已描述了本发明的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本发明范围的所有变更和修改。Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be construed to include the preferred embodiment and all changes and modifications that fall within the scope of the present invention.

显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本发明的精神和范围。这样，倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit and scope of the invention. Thus, provided that these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.