CN118098599A - Pregnancy monitoring management system and method for high-risk pregnant and lying-in women - Google Patents
Pregnancy monitoring management system and method for high-risk pregnant and lying-in women Download PDFInfo
- Publication number
- CN118098599A CN118098599A CN202410481137.7A CN202410481137A CN118098599A CN 118098599 A CN118098599 A CN 118098599A CN 202410481137 A CN202410481137 A CN 202410481137A CN 118098599 A CN118098599 A CN 118098599A
- Authority
- CN
- China
- Prior art keywords
- physiological parameter
- time sequence
- feature map
- training
- parameter time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 61
- 230000035935 pregnancy Effects 0.000 title claims abstract description 56
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000012549 training Methods 0.000 claims description 95
- 239000011159 matrix material Substances 0.000 claims description 89
- 238000010586 diagram Methods 0.000 claims description 48
- 238000007726 management method Methods 0.000 claims description 48
- 230000006870 function Effects 0.000 claims description 28
- 238000003062 neural network model Methods 0.000 claims description 25
- 230000036541 health Effects 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 21
- 230000035790 physiological processes and functions Effects 0.000 claims description 20
- 230000002159 abnormal effect Effects 0.000 claims description 19
- 230000005856 abnormality Effects 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 18
- 230000008774 maternal effect Effects 0.000 claims description 16
- 238000001514 detection method Methods 0.000 claims description 14
- 230000004927 fusion Effects 0.000 claims description 14
- 230000007246 mechanism Effects 0.000 claims description 14
- 230000009466 transformation Effects 0.000 claims description 14
- 208000036029 Uterine contractions during pregnancy Diseases 0.000 claims description 13
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 13
- 239000008280 blood Substances 0.000 claims description 13
- 210000004369 blood Anatomy 0.000 claims description 13
- 230000036772 blood pressure Effects 0.000 claims description 13
- 210000002458 fetal heart Anatomy 0.000 claims description 13
- 229910052760 oxygen Inorganic materials 0.000 claims description 13
- 239000001301 oxygen Substances 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 238000012512 characterization method Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 2
- 230000001105 regulatory effect Effects 0.000 claims description 2
- 230000004962 physiological condition Effects 0.000 claims 1
- 238000009826 distribution Methods 0.000 description 10
- 238000011176 pooling Methods 0.000 description 7
- 230000004913 activation Effects 0.000 description 6
- 210000002569 neuron Anatomy 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 210000003754 fetus Anatomy 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000037396 body weight Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000019771 cognition Effects 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000005094 computer simulation Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/0205—Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/43—Detecting, measuring or recording for evaluating the reproductive systems
- A61B5/4306—Detecting, measuring or recording for evaluating the reproductive systems for evaluating the female reproductive systems, e.g. gynaecological evaluations
- A61B5/4343—Pregnancy and labour monitoring, e.g. for labour onset detection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2123/00—Data types
- G06F2123/02—Data types in the time domain, e.g. time-series data
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Heart & Thoracic Surgery (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Physiology (AREA)
- Mathematical Physics (AREA)
- Cardiology (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Pulmonology (AREA)
- Reproductive Health (AREA)
- Probability & Statistics with Applications (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
Abstract
The application discloses a pregnancy monitoring management system and method for high-risk pregnant and lying-in women, and relates to the field of intelligent monitoring.
Description
Technical Field
The application relates to the field of intelligent monitoring, in particular to a pregnancy monitoring management system and method for a high-risk pregnant woman.
Background
High-risk pregnant and lying-in women refer to pregnant women who have a higher health risk during pregnancy for various reasons (e.g., age, history of illness, pre-pregnancy condition, etc.). The pregnant and lying-in women with high risk are monitored and managed very important, and the pregnant and lying-in women can be ensured to be safe and healthy. However, conventional pregnancy monitoring management systems for high-risk pregnant and lying-in women generally rely on regular labor and hospital monitoring, which has some drawbacks. In particular, the conventional pregnancy monitoring management system requires pregnant women to go to hospitals periodically for birth inspection, which may bring inconvenience and stress to them. Moreover, because the monitoring is intermittent, the traditional method may miss some emergencies or early anomalies, and once an anomaly occurs, a doctor needs time to respond in time, which may delay the treatment opportunity. In addition, the monitoring result of the traditional pregnancy monitoring management system is often influenced by subjective judgment of doctors, and different doctors may have different diagnosis standards, so that the pregnancy monitoring management result is inconsistent.
Therefore, an optimized pregnancy monitoring management system and method for high-risk pregnant women is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems.
According to one aspect of the present application, there is provided a pregnancy monitoring management system for a high-risk pregnant woman, comprising:
The physiological parameter data acquisition module is used for acquiring a time sequence of physiological parameter data of the monitored high-risk pregnant and lying-in women, acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in women, wherein the physiological parameter data comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
the physiological parameter data time sequence normalization module is used for arranging the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
the physiological parameter entity content focusing module is used for focusing the physiological parameter time sequence matrix based on entity content to obtain a focused physiological parameter time sequence matrix;
The physiological parameter time sequence correlation feature extraction module is used for extracting features of the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode correlation feature extractor based on a deep neural network model so as to obtain a physiological parameter time sequence correlation feature map;
the multi-scale physiological state characterization module is used for enabling the physiological parameter time sequence correlation characteristic diagram to be used as a physiological parameter time sequence multi-scale correlation characteristic through a physiological parameter time sequence multi-scale characteristic enhancement expressive machine based on the Res2Net module;
the physiological state abnormality detection module is used for determining whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not based on the physiological parameter time sequence multi-scale correlation characteristics;
Wherein, the physical content focusing module of physiological parameter is used for: processing the physiological parameter time sequence matrix through a physiological parameter focusing module based on an entity content attention mechanism according to the following focusing formula to obtain the focusing physiological parameter time sequence matrix;
Wherein, the focusing formula is:
；
；
Wherein, For the/>, in the physiological parameter timing matrixParameter value of location,/>Is the/>, in the weight vectorNumerical value of individual position,/>For the weight vector,/>For the physiological parameter timing matrix,/>A time sequence matrix for the focused physiological parameters.
In the pregnancy monitoring and management system for the high-risk pregnant and lying-in women, the deep neural network model is a convolutional neural network model.
In the pregnancy monitoring management system for high-risk pregnant and lying-in women, the multi-scale physiological state characterization module comprises:
The channel transformation unit is used for enabling the physiological parameter time sequence correlation characteristic diagram to pass through a first convolution layer based on a 1 multiplied by 1 convolution kernel to obtain a channel transformation physiological parameter time sequence correlation characteristic diagram;
the characteristic diagram splitting unit is used for splitting the channel transformation physiological parameter time sequence associated characteristic diagram along the channel dimension to obtain a first branch characteristic diagram, a second branch characteristic diagram, a third branch characteristic diagram and a fourth branch characteristic diagram;
the first branch characteristic extraction unit is used for enabling the first branch characteristic diagram to pass through a convolutional neural network model to obtain a first branch output characteristic diagram;
a second branch feature extraction unit, configured to process the second branch feature map through a second convolution layer based on a3×3 convolution kernel to obtain a second branch output feature map;
A third branch feature extraction unit, configured to fuse the second branch output feature map and the third branch feature map, and then process the fused second branch output feature map and the third branch feature map through a third convolution layer based on a3×3 convolution kernel to obtain a third branch output feature map;
A fourth branch feature extraction unit, configured to fuse the third branch output feature map and the fourth branch feature map, and then process the fused third branch output feature map and the fourth branch feature map through a fourth convolution layer based on a3×3 convolution kernel to obtain a fourth branch output feature map;
The multi-branch feature fusion unit is used for fusing the first branch output feature map, the second branch output feature map, the third branch output feature map and the fourth branch output feature map to obtain a multi-branch fusion feature map;
the dimension reduction unit is used for processing the multi-branch fusion feature map through a fifth convolution layer based on a1 multiplied by 1 convolution kernel to obtain a channel transformation multi-branch fusion feature map;
The physiological parameter time sequence multi-scale feature expression unit is used for fusing the channel transformation multi-branch fusion feature map and the physiological parameter time sequence correlation feature map to obtain the physiological parameter time sequence multi-scale correlation feature map.
In the pregnancy monitoring and management system for high-risk pregnant and lying-in women, the physiological state abnormality detection module is used for: and the physiological parameter time sequence multi-scale associated feature map passes through a health abnormality detector based on a classifier to obtain a detection result, wherein the detection result is used for indicating whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not.
The pregnancy monitoring management system for the high-risk pregnant and lying-in women further comprises a physiological parameter focusing module based on an entity content attention mechanism, a physiological parameter time sequence mode correlation feature extractor based on a deep neural network model, a physiological parameter time sequence multi-scale feature enhancement expression based on a Res2Net module and a training module for training the health abnormality detector based on a classifier.
In the pregnancy monitoring management system for high-risk pregnant and lying-in women, the training module comprises:
The system comprises a training data acquisition unit, a data processing unit and a data processing unit, wherein the training data acquisition unit is used for acquiring training data, the training data comprises a time sequence of training physiological parameter data of a monitored high-risk pregnant and lying-in woman, the training physiological parameter data is acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in woman, and the training physiological parameter data comprises heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
The training physiological parameter data time sequence regulating unit is used for arranging the time sequence of the training physiological parameter data of the monitored high-risk pregnant and lying-in women into a training physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
The training physiological parameter entity content focusing unit is used for focusing the training physiological parameter time sequence matrix based on entity content to obtain a training focusing physiological parameter time sequence matrix;
the training physiological parameter time sequence correlation feature extraction unit is used for extracting features of the training focusing physiological parameter time sequence matrix through a physiological parameter time sequence mode correlation feature extractor based on a deep neural network model so as to obtain a training physiological parameter time sequence correlation feature map;
The multi-scale physiological state characterization unit is used for obtaining a training physiological parameter time sequence multi-scale correlation characteristic diagram through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module;
the optimization unit is used for optimizing each feature matrix of the training physiological parameter time sequence multi-scale associated feature map to obtain an optimized training physiological parameter time sequence multi-scale associated feature map;
the classification loss unit is used for enabling the optimized training physiological parameter time sequence multi-scale associated feature map to pass through a health anomaly detector based on a classifier so as to obtain a classification loss function value;
The training unit is used for training the physiological parameter focusing module based on the entity content attention mechanism, the physiological parameter time sequence mode association feature extractor based on the deep neural network model, the physiological parameter time sequence multi-scale feature enhancement expressive machine based on the Res2Net module and the health abnormality detector based on the classifier based on the classification loss function value.
According to another aspect of the present application, there is provided a pregnancy monitoring and management method for a high-risk pregnant woman, comprising:
Acquiring a time sequence of physiological parameter data of a monitored high-risk pregnant and lying-in woman acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in woman, wherein the physiological parameter data comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
arranging the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
focusing the physiological parameter time sequence matrix based on the entity content to obtain a focused physiological parameter time sequence matrix;
Extracting features of the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode associated feature extractor based on a deep neural network model to obtain a physiological parameter time sequence associated feature map;
The physiological parameter time sequence correlation characteristic map is passed through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module to obtain a physiological parameter time sequence multi-scale correlation characteristic map as a physiological parameter time sequence multi-scale correlation characteristic;
determining whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not based on the physiological parameter time sequence multi-scale correlation characteristics;
Focusing the physiological parameter time sequence matrix based on the entity content to obtain a focused physiological parameter time sequence matrix, wherein the focused physiological parameter time sequence matrix is used for: processing the physiological parameter time sequence matrix through a physiological parameter focusing module based on an entity content attention mechanism according to the following focusing formula to obtain the focusing physiological parameter time sequence matrix;
Wherein, the focusing formula is:
；
；
Wherein, For the/>, in the physiological parameter timing matrixParameter value of location,/>Is the/>, in the weight vectorNumerical value of individual position,/>For the weight vector,/>For the physiological parameter timing matrix,/>A time sequence matrix for the focused physiological parameters.
Compared with the prior art, the pregnancy monitoring management system and method for the high-risk pregnant and lying-in women provided by the application have the advantages that physiological parameter data of the pregnant and lying-in women, including heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate, are monitored and collected in real time through the intelligent wearing equipment worn by the high-risk pregnant and lying-in women, and the time sequence collaborative analysis of the physiological parameter data is carried out by introducing a data processing and analyzing algorithm at the rear end, so that the pregnancy health condition of the high-risk pregnant and lying-in women is monitored more accurately based on the collaborative effect and correlation relation among different physiological parameter time sequence characteristics, which is beneficial to helping doctors identify abnormal conditions, adopting corresponding intervention measures and improving the safety of the pregnant and lying-in women and fetuses.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
Fig. 1 is a block diagram of a pregnancy monitoring management system for a high risk maternal according to an embodiment of the present application;
Fig. 2 is a system architecture diagram of a pregnancy monitoring management system for a high risk maternal according to an embodiment of the present application;
FIG. 3 is a block diagram of a training phase of a pregnancy monitoring management system for high risk pregnant women according to an embodiment of the present application;
Fig. 4 is a flowchart of a pregnancy monitoring management method for a high risk maternal according to an embodiment of the present application.
Detailed Description
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be appreciated that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
Traditional pregnancy monitoring management systems require pregnant women to go to hospitals periodically for a labor test, which can be inconvenient and stressful for them. Moreover, because the monitoring is intermittent, the traditional method may miss some emergencies or early anomalies, and once an anomaly occurs, a doctor needs time to respond in time, which may delay the treatment opportunity. In addition, the monitoring result of the traditional pregnancy monitoring management system is often influenced by subjective judgment of doctors, and different doctors may have different diagnosis standards, so that the pregnancy monitoring management result is inconsistent. Therefore, an optimized pregnancy monitoring management system for high-risk pregnant women is desired.
In the technical scheme of the application, a pregnancy monitoring and management system for high-risk pregnant and lying-in women is provided. Fig. 1 is a block diagram of a pregnancy monitoring management system for a high risk maternal according to an embodiment of the present application. Fig. 2 is a system architecture diagram of a pregnancy monitoring management system for a high-risk pregnant woman according to an embodiment of the present application. As shown in fig. 1 and 2, a pregnancy monitoring management system 300 for a high-risk maternal according to an embodiment of the present application includes: a physiological parameter data acquisition module 310, configured to acquire a time sequence of physiological parameter data of a monitored high-risk pregnant and lying-in woman acquired by an intelligent wearable device worn by the monitored high-risk pregnant and lying-in woman, where the physiological parameter data includes a heart rate, a blood pressure, a body weight, a blood oxygen saturation, a uterine contraction frequency, and a fetal heart rate; a physiological parameter data time sequence regulation module 320, configured to arrange the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to a time dimension and a physiological parameter sample dimension; a physical content focusing module 330 for focusing the physical content on the physiological parameter timing matrix to obtain a focused physiological parameter timing matrix; the physiological parameter time sequence correlation feature extraction module 340 is configured to perform feature extraction on the focused physiological parameter time sequence matrix by using a physiological parameter time sequence pattern correlation feature extractor based on a deep neural network model to obtain a physiological parameter time sequence correlation feature map; the multi-scale physiological state characterization module 350 is configured to obtain the physiological parameter time sequence multi-scale correlation feature map as a physiological parameter time sequence multi-scale correlation feature by using the physiological parameter time sequence multi-scale feature enhancement expressive machine based on the Res2Net module; the physiological state abnormality detection module 360 is configured to determine whether an abnormality exists in the monitored high-risk maternal based on the physiological parameter time sequence multi-scale correlation characteristic.
In particular, the physiological parameter data acquisition module 310 is configured to acquire a time sequence of physiological parameter data of the monitored high-risk pregnant and lying-in woman acquired by a smart wearable device worn by the monitored high-risk pregnant and lying-in woman, where the physiological parameter data includes heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate. The physiological state abnormality detection module 360 is configured to determine whether an abnormality exists in the monitored high-risk maternal based on the physiological parameter time sequence multi-scale correlation characteristic.
In particular, the physiological parameter data timing module 320 is configured to arrange the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter timing matrix according to a time dimension and a physiological parameter sample dimension. Considering that the physiological parameter data of the monitored high-risk pregnant and lying-in women comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate, the physiological parameter data not only have a time sequence dynamic change rule in a time dimension, but also have time sequence cooperative association and correlation characteristics between different physiological parameter data. Therefore, in order to capture the time sequence collaborative correlation characteristics of each physiological parameter data of the monitored high-risk pregnant and lying-in women more conveniently and effectively, the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women needs to be arranged into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension.
In particular, the physical content focusing module 330 is configured to focus the physical content on the physiological parameter timing matrix to obtain a focused physiological parameter timing matrix. It should be understood that, in the process of actually monitoring the physiological state of a high-risk pregnant and lying-in woman, different physiological parameters have different contribution degrees to the judgment of the physiological state and the detection of the health condition, and meanwhile, the data of different parameter items in the actual physiological parameter data have different importance in the time dimension. Therefore, in order to help the system pay more attention to important physiological parameter time sequence information, so as to improve the efficiency of data processing and the accuracy of judging the physiological state, in the technical scheme of the application, the physiological parameter time sequence matrix is required to be focused by a physiological parameter focusing module based on an entity content attention mechanism so as to obtain a physiological parameter focusing matrix. It should be appreciated that the entity content attention mechanism may help the system more efficiently screen and extract critical information when processing large amounts of physiological parameter data, avoiding information overload and redundancy, thereby improving the efficiency and speed of data processing. In this way, physiological parameter time sequence data which is vital for judging the health condition of pregnant women in the process of monitoring high-risk pregnant women can be identified and focused, and the relevance and importance among different physiological parameters can be better understood and utilized, so that the prediction and identification capacity of the model is improved. Specifically, the physiological parameter time sequence matrix is processed through a physiological parameter focusing module based on an entity content attention mechanism according to the following focusing formula to obtain the focused physiological parameter time sequence matrix; wherein, the focusing formula is:
；
；
Wherein, For the/>, in the physiological parameter timing matrixParameter value of location,/>Is the/>, in the weight vectorNumerical value of individual position,/>For the weight vector,/>For the physiological parameter timing matrix,/>A time sequence matrix for the focused physiological parameters.
In particular, the physiological parameter time series correlation feature extraction module 340 is configured to perform feature extraction on the focused physiological parameter time series matrix by using a physiological parameter time series pattern correlation feature extractor based on a deep neural network model to obtain a physiological parameter time series correlation feature map. In particular, in one specific example of the present application, the deep neural network model is a convolutional neural network model. Namely, feature mining is carried out on the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode association feature extractor based on a deep neural network model so as to extract time sequence collaborative association features and change rules among entity contents of each physiological parameter data, thereby obtaining a physiological parameter time sequence association feature map. This helps the system to better understand the dynamic characteristics and correlation relationships of physiological parameter data, thereby improving the ability to identify and predict abnormal conditions. Specifically, each layer of the physiological parameter time sequence mode associated feature extractor based on the deep neural network model is used for respectively carrying out input data in forward transfer of the layer: carrying out convolution processing on input data to obtain a convolution characteristic diagram; pooling the convolution feature images based on the local feature matrix to obtain pooled feature images; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the physiological parameter time sequence pattern correlation feature extractor based on the deep neural network model is the physiological parameter time sequence correlation feature map, and the input of the first layer of the physiological parameter time sequence pattern correlation feature extractor based on the deep neural network model is the focusing physiological parameter time sequence matrix.
Convolutional neural networks (Convolutional Neural Network, CNN) are a type of deep learning model that is particularly useful for processing data having a grid structure, such as images and video. The following is a general structure and a step-wise expansion of the convolutional neural network model: input layer: accepting input data, typically images, audio, text, or the like; convolution layer: the convolutional layer is one of the core components of the CNN. It extracts local features in the input data by applying a series of filters (also called convolution kernels). The convolution operation multiplies and sums the filter and the input data element by element to generate a feature map; activation function: after the convolutional layer, a nonlinear activation function, such as ReLU, is typically applied to introduce nonlinear characteristics; pooling layer: the pooling layer is used to reduce the spatial size of the feature map and preserve the most important features. Common pooling operations have maximum pooling and average pooling; full tie layer: the fully connected layer connects the outputs of the pooling layer to one or more fully connected layers for mapping features to final output categories or regression values. Each neuron in the fully connected layer is connected with all neurons of the previous layer; output layer: the output layer selects proper activation functions according to different tasks, such as softmax functions for multi-classification tasks and linear activation functions for regression tasks; loss function: selecting proper loss functions according to different tasks, such as cross entropy loss functions for classifying tasks and mean square error loss functions for regression tasks; back propagation and optimization: the gradient of the model parameters is calculated from the loss function by a back propagation algorithm and a gradient descent optimization algorithm, and the parameters are updated to minimize the loss function.
In particular, the multi-scale physiological state characterization module 350 is configured to obtain the physiological parameter time-series multi-scale correlation feature map as a physiological parameter time-series multi-scale correlation feature by using the physiological parameter time-series multi-scale feature enhancement expressive device based on the Res2Net module. It should be understood that in the actual process of monitoring the physiological state and detecting the abnormal situation of the high-risk pregnant and lying-in women, different physiological state parameters have different time sequence scales and different degrees of association relations, and meanwhile, different physiological parameter data of the pregnant women also have different correlation characteristics. Therefore, in order to enable the model to better adapt to physiological parameter data of different pregnant women and improve generalization capability and stability of the model, in the technical scheme of the application, the physiological parameter time sequence correlation characteristic diagram is further processed through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module to obtain the physiological parameter time sequence multi-scale correlation characteristic diagram. It should be noted that the physiological parameter time sequence multi-scale feature enhancement expression device based on the Res2Net module is an effective multi-scale feature enhancement module, which can help the system extract richer and more accurate correlation features between various physiological parameter data on different scales. That is, by applying the Res2Net module, the multi-scale information in the physiological parameter time sequence correlation feature map can be enhanced, so that the system can better capture the fine change and important features in the time sequence cooperative correlation feature of each physiological parameter data, and thus the time sequence multi-scale correlation feature between each physiological parameter data of the pregnant woman can be more comprehensively understood. This helps to improve the system's understanding and analysis of the health of the pregnant woman. Specifically, the physiological parameter time sequence correlation characteristic diagram passes through a first convolution layer based on a 1 multiplied by 1 convolution kernel to obtain a channel transformation physiological parameter time sequence correlation characteristic diagram; splitting the channel transformation physiological parameter time sequence association feature map along the channel dimension to obtain a first branch feature map, a second branch feature map, a third branch feature map and a fourth branch feature map; the first branch characteristic diagram is passed through a convolutional neural network model to obtain a first branch output characteristic diagram; processing the second branch characteristic diagram through a second convolution layer based on a 3×3 convolution kernel to obtain a second branch output characteristic diagram; after the second branch output characteristic diagram and the third branch characteristic diagram are fused, the third branch output characteristic diagram is processed through a third convolution layer based on a 3 multiplied by 3 convolution kernel, so that a third branch output characteristic diagram is obtained; after fusing the third branch output characteristic diagram and the fourth branch characteristic diagram, processing the third branch output characteristic diagram and the fourth branch output characteristic diagram through a fourth convolution layer based on a 3 multiplied by 3 convolution kernel to obtain a fourth branch output characteristic diagram; fusing the first branch output feature map, the second branch output feature map, the third branch output feature map and the fourth branch output feature map to obtain a multi-branch fusion feature map; processing the multi-branch fusion feature map through a fifth convolution layer based on a 1 multiplied by 1 convolution kernel to obtain a channel transformation multi-branch fusion feature map; and fusing the channel transformation multi-branch fusion feature map and the physiological parameter time sequence correlation feature map to obtain the physiological parameter time sequence multi-scale correlation feature map.
In particular, the physiological state abnormality detection module 360 is configured to determine whether an abnormality exists in the monitored high-risk maternal based on the physiological parameter time sequence multi-scale correlation characteristic. In particular, in one specific example of the application, the physiological parameter time sequence multi-scale correlation characteristic map is passed through a classifier-based health abnormality detector to obtain a detection result, wherein the detection result is used for indicating whether the monitored high-risk pregnant and lying-in women have abnormal conditions. That is, the time sequence of a plurality of physiological parameter data of the high-risk pregnant and lying-in women is utilized to carry out classification processing by utilizing the time sequence cooperation multi-scale association characteristics, so that the pregnancy health condition of the high-risk pregnant and lying-in women is monitored based on the cooperation and correlation relationship between the time sequence characteristics of different physiological parameters, which is beneficial to helping doctors identify abnormal conditions, corresponding intervention measures are adopted, and the safety of the pregnant and lying-in women and the fetus is improved. Specifically, the physiological parameter time sequence multi-scale associated feature map is unfolded to be a classification feature vector based on a row vector or a column vector; performing full-connection coding on the classification feature vectors by using a plurality of full-connection layers of the classifier to obtain coded classification feature vectors; and passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
That is, in the technical solution of the present application, the label of the classifier includes a monitored high-risk maternal abnormal condition (first label) and a monitored high-risk maternal abnormal condition (second label), wherein the classifier determines, through a soft maximum function, to which classification label the physiological parameter time sequence multi-scale correlation feature map belongs. It should be noted that the first tag p1 and the second tag p2 do not include a manually set concept, and in fact, during the training process, the computer model does not have a concept of "whether there is an abnormal situation in the monitored high-risk pregnant woman", which is simply that there are two kinds of classification tags, and the probability that the output characteristics are the two classification tags sign is one, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the monitored high-risk pregnant and lying-in women have abnormal conditions is actually converted into the classification probability distribution conforming to the natural rule through the classification label, and the physical meaning of the natural probability distribution of the label is essentially used instead of the language text meaning of whether the monitored high-risk pregnant and lying-in women have abnormal conditions.
A classifier refers to a machine learning model or algorithm that is used to classify input data into different categories or labels. The classifier is part of supervised learning, which performs classification tasks by learning mappings from input data to output categories.
Fully connected layers are one type of layer commonly found in neural networks. In the fully connected layer, each neuron is connected to all neurons of the upper layer, and each connection has a weight. This means that each neuron in the fully connected layer receives inputs from all neurons in the upper layer, and weights these inputs together, and then passes the result to the next layer.
The Softmax classification function is a commonly used activation function for multi-classification problems. It converts each element of the input vector into a probability value between 0 and 1, and the sum of these probability values equals 1. The Softmax function is commonly used at the output layer of a neural network, and is particularly suited for multi-classification problems, because it can map the network output into probability distributions for individual classes. During the training process, the output of the Softmax function may be used to calculate the loss function and update the network parameters through a back propagation algorithm. Notably, the output of the Softmax function does not change the relative magnitude relationship between elements, but rather normalizes them. Thus, the Softmax function does not change the characteristics of the input vector, but simply converts it into a probability distribution form.
It should be appreciated that the physical content attention mechanism-based physiological parameter focusing module, the deep neural network model-based physiological parameter timing pattern correlation feature extractor, the Res2Net module-based physiological parameter timing multi-scale feature enhancement expressive device, and the classifier-based health anomaly detector need to be trained prior to the inference using the neural network model described above. That is, the pregnancy monitoring management system 300 for high-risk pregnant women according to the present application further comprises a training stage 400 for training the physical content attention mechanism-based physiological parameter focusing module, the deep neural network model-based physiological parameter timing pattern correlation feature extractor, the Res2Net module-based physiological parameter timing multi-scale feature enhancement expressior, and the classifier-based health abnormality detector.
Fig. 3 is a block diagram of a training phase of a pregnancy monitoring management system for high risk pregnant women according to an embodiment of the present application. As shown in fig. 3, a pregnancy monitoring management system 300 for a high-risk pregnant woman according to an embodiment of the present application includes: training phase 400, comprising: a training data acquisition unit 410, configured to acquire training data, where the training data includes a time sequence of training physiological parameter data of a monitored high-risk pregnant and lying-in woman acquired by an intelligent wearable device worn by the monitored high-risk pregnant and lying-in woman, and the training physiological parameter data includes a heart rate, a blood pressure, a body weight, a blood oxygen saturation level, a uterine contraction frequency, and a fetal heart rate; a training physiological parameter data time sequence normalization unit 420, configured to arrange a time sequence of training physiological parameter data of the monitored high-risk pregnant and lying-in women into a training physiological parameter time sequence matrix according to a time dimension and a physiological parameter sample dimension; a training physiological parameter entity content focusing unit 430, configured to focus the training physiological parameter timing matrix based on entity content to obtain a training focused physiological parameter timing matrix; a training physiological parameter time sequence correlation feature extraction unit 440, configured to perform feature extraction on the training focused physiological parameter time sequence matrix by using a physiological parameter time sequence pattern correlation feature extractor based on a deep neural network model to obtain a training physiological parameter time sequence correlation feature map; the multi-scale physiological state characterization unit 450 is configured to obtain a training physiological parameter time sequence multi-scale correlation feature map by using the training physiological parameter time sequence multi-scale feature enhancement expressive machine based on the Res2Net module; an optimizing unit 460, configured to optimize each feature matrix of the training physiological parameter time sequence multi-scale associated feature map to obtain an optimized training physiological parameter time sequence multi-scale associated feature map; a classification loss unit 470, configured to pass the optimized training physiological parameter time sequence multi-scale correlation feature map through a health anomaly detector based on a classifier to obtain a classification loss function value; a training unit 480, configured to train the physiological parameter focusing module based on the entity content attention mechanism, the physiological parameter timing pattern correlation feature extractor based on the deep neural network model, the physiological parameter timing multi-scale feature enhancement expressive device based on the Res2Net module, and the health abnormality detector based on the classifier based on the classification loss function value.
Particularly, in the technical scheme of the application, the training physiological parameter time sequence association feature map expresses the association feature of local focusing of sample-time sequence cross dimension of the physiological parameter data, so that after the training physiological parameter time sequence association feature map passes through a physiological parameter time sequence multi-scale feature enhancement expressive device based on a Res2Net module, the association feature enhancement is carried out on different feature dimensions based on spatial distribution of feature matrixes by considering convolution branches of each scale, and the obtained training physiological parameter time sequence multi-scale association feature map has feature distribution differences among the feature matrixes, thereby influencing the quasi-probability regression effect of the feature matrix of the training physiological parameter time sequence multi-scale association feature map based on the integral numerical distribution of feature values, and influencing the training speed of classification training by a classifier and the accuracy of the obtained classification result. Therefore, when the training physiological parameter time sequence multi-scale associated feature map carries out training iteration of classification regression through a classifier, the applicant optimizes each feature matrix of the training physiological parameter time sequence multi-scale associated feature map, and the feature matrix is expressed as:
；
Wherein the method comprises the steps of Is the/>, of each feature matrix of the training physiological parameter time sequence multi-scale associated feature mapCharacteristic value of location,/>Probability function representing eigenvalues, i.e. eigenvalues/>Mapping to/>Probability function of interval,/>Is the scale of each feature matrix of the training physiological parameter time sequence multi-scale associated feature map, namely the width multiplied by the height,/>Is a class probability value obtained by the training physiological parameter time sequence multi-scale associated feature map through a classifier, and/>Is a weight super parameter. That is, for the feature scene corresponding to each feature matrix of the training physiological parameter time sequence multi-scale associated feature map, the probability distribution foreground constraint and the relative probability mapping response assumption are used for accepting the class probability reasoning logic association of scene saturation, so that the feature set of each feature matrix of the training physiological parameter time sequence multi-scale associated feature map is endowed with scene concept ontology cognition, that is, the overall distribution is internally aligned with the class probability logic reasoning based on the scene in the classification process, so that the understanding capability of the feature matrix scene distribution of the training physiological parameter time sequence multi-scale associated feature map on the class cognition is improved. Thus, the coefficient/>The weighted optimization is carried out on the corresponding feature matrix of the training physiological parameter time sequence multi-scale associated feature map, so that the quasi-probability regression effect of the optimized physiological parameter time sequence multi-scale associated feature map can be improved, namely, the training speed of classification training through a classifier and the accuracy of the obtained classification result are improved. Therefore, the pregnancy health condition of the high-risk pregnant and lying-in women can be more accurately monitored based on the synergetic effect and the correlation relationship between the time sequence characteristics of different physiological parameters, which is beneficial to helping doctors identify abnormal conditions and taking corresponding intervention measures to improve the safety of the pregnant and lying-in women and the fetus.
As described above, the pregnancy monitoring management system 300 for a high-risk pregnant and lying-in woman according to the embodiment of the present application may be implemented in various wireless terminals, such as a server or the like having a pregnancy monitoring management algorithm for a high-risk pregnant and lying-in woman. In one possible implementation, the pregnancy monitoring management system 300 for high-risk pregnant women according to embodiments of the present application may be integrated into the wireless terminal as one software module and/or hardware module. For example, the pregnancy monitoring management system 300 for high-risk pregnant women may be a software module in the operating system of the wireless terminal, or may be an application developed for the wireless terminal; of course, the pregnancy monitoring management system 300 for high-risk pregnant women can also be one of the hardware modules of the wireless terminal.
Alternatively, in another example, the pregnancy monitoring management system 300 for a high-risk maternal may be a separate device from the wireless terminal, and the pregnancy monitoring management system 300 for a high-risk maternal may be connected to the wireless terminal through a wired and/or wireless network and transmit interactive information according to a contracted data format.
Furthermore, the invention also provides a pregnancy monitoring and management method for the high-risk pregnant and lying-in women.
Fig. 4 is a flowchart of a pregnancy monitoring management method for a high risk maternal according to an embodiment of the present application. As shown in fig. 4, the pregnancy monitoring and management method for a high-risk pregnant woman according to the embodiment of the application includes the steps of: s1, acquiring a time sequence of physiological parameter data of a monitored high-risk pregnant and lying-in woman, wherein the physiological parameter data comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate, and the physiological parameter data are acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in woman; s2, arranging the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension; s3, focusing the physiological parameter time sequence matrix based on the entity content to obtain a focused physiological parameter time sequence matrix; s4, extracting features of the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode associated feature extractor based on a deep neural network model to obtain a physiological parameter time sequence associated feature map; s5, the physiological parameter time sequence correlation characteristic map is passed through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module to obtain a physiological parameter time sequence multi-scale correlation characteristic map as a physiological parameter time sequence multi-scale correlation characteristic; s6, determining whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not based on the physiological parameter time sequence multi-scale correlation characteristics.
In summary, the pregnancy monitoring management method for the high-risk pregnant and lying-in women according to the embodiment of the application is clarified, physiological parameter data of the pregnant and lying-in women, including heart rate, blood pressure, body weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate, are monitored and collected in real time through the intelligent wearing equipment worn by the high-risk pregnant and lying-in women, and a data processing and analyzing algorithm is introduced at the rear end to carry out time sequence collaborative analysis of the physiological parameter data, so that pregnancy health condition monitoring of the high-risk pregnant and lying-in women is more accurately carried out based on collaborative effects and correlation relations among time sequence characteristics of different physiological parameters, which is beneficial to helping doctors identify abnormal conditions, corresponding intervention measures are adopted, and safety of the pregnant and lying-in women and fetuses is improved.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (7)
1. A pregnancy monitoring management system for high-risk pregnant and lying-in women, comprising:
The physiological parameter data acquisition module is used for acquiring a time sequence of physiological parameter data of the monitored high-risk pregnant and lying-in women, acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in women, wherein the physiological parameter data comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
the physiological parameter data time sequence normalization module is used for arranging the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
the physiological parameter entity content focusing module is used for focusing the physiological parameter time sequence matrix based on entity content to obtain a focused physiological parameter time sequence matrix;
The physiological parameter time sequence correlation feature extraction module is used for extracting features of the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode correlation feature extractor based on a deep neural network model so as to obtain a physiological parameter time sequence correlation feature map;
the multi-scale physiological state characterization module is used for enabling the physiological parameter time sequence correlation characteristic diagram to be used as a physiological parameter time sequence multi-scale correlation characteristic through a physiological parameter time sequence multi-scale characteristic enhancement expressive machine based on the Res2Net module;
the physiological state abnormality detection module is used for determining whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not based on the physiological parameter time sequence multi-scale correlation characteristics;
Wherein, the physical content focusing module of physiological parameter is used for: processing the physiological parameter time sequence matrix through a physiological parameter focusing module based on an entity content attention mechanism according to the following focusing formula to obtain the focusing physiological parameter time sequence matrix;
Wherein, the focusing formula is:
；
；
Wherein, For the/>, in the physiological parameter timing matrixParameter value of location,/>Is the/>, in the weight vectorNumerical value of individual position,/>For the weight vector,/>For the physiological parameter timing matrix,/>A time sequence matrix for the focused physiological parameters.
2. The pregnancy monitoring management system for high-risk maternal according to claim 1, wherein the deep neural network model is a convolutional neural network model.
3. The pregnancy monitoring management system for high-risk maternal according to claim 2, wherein the multi-scale physiological state characterization module comprises:
The channel transformation unit is used for enabling the physiological parameter time sequence correlation characteristic diagram to pass through a first convolution layer based on a 1 multiplied by 1 convolution kernel to obtain a channel transformation physiological parameter time sequence correlation characteristic diagram;
the characteristic diagram splitting unit is used for splitting the channel transformation physiological parameter time sequence associated characteristic diagram along the channel dimension to obtain a first branch characteristic diagram, a second branch characteristic diagram, a third branch characteristic diagram and a fourth branch characteristic diagram;
the first branch characteristic extraction unit is used for enabling the first branch characteristic diagram to pass through a convolutional neural network model to obtain a first branch output characteristic diagram;
a second branch feature extraction unit, configured to process the second branch feature map through a second convolution layer based on a3×3 convolution kernel to obtain a second branch output feature map;
A third branch feature extraction unit, configured to fuse the second branch output feature map and the third branch feature map, and then process the fused second branch output feature map and the third branch feature map through a third convolution layer based on a3×3 convolution kernel to obtain a third branch output feature map;
A fourth branch feature extraction unit, configured to fuse the third branch output feature map and the fourth branch feature map, and then process the fused third branch output feature map and the fourth branch feature map through a fourth convolution layer based on a3×3 convolution kernel to obtain a fourth branch output feature map;
The multi-branch feature fusion unit is used for fusing the first branch output feature map, the second branch output feature map, the third branch output feature map and the fourth branch output feature map to obtain a multi-branch fusion feature map;
the dimension reduction unit is used for processing the multi-branch fusion feature map through a fifth convolution layer based on a1 multiplied by 1 convolution kernel to obtain a channel transformation multi-branch fusion feature map;
The physiological parameter time sequence multi-scale feature expression unit is used for fusing the channel transformation multi-branch fusion feature map and the physiological parameter time sequence correlation feature map to obtain the physiological parameter time sequence multi-scale correlation feature map.
4. A pregnancy monitoring management system for high-risk pregnant women according to claim 3, wherein the physiological condition abnormality detection module is configured to: and the physiological parameter time sequence multi-scale associated feature map passes through a health abnormality detector based on a classifier to obtain a detection result, wherein the detection result is used for indicating whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not.
5. The pregnancy monitoring management system for high-risk maternal according to claim 4, further comprising a training module for training the physical content attention mechanism based physiological parameter focusing module, the deep neural network model based physiological parameter timing pattern correlation feature extractor, the Res2Net module based physiological parameter timing multiscale feature enhancement expressior, and the classifier based health anomaly detector.
6. The pregnancy monitoring management system for high-risk pregnant women according to claim 5, wherein the training module comprises:
The system comprises a training data acquisition unit, a data processing unit and a data processing unit, wherein the training data acquisition unit is used for acquiring training data, the training data comprises a time sequence of training physiological parameter data of a monitored high-risk pregnant and lying-in woman, the training physiological parameter data is acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in woman, and the training physiological parameter data comprises heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
The training physiological parameter data time sequence regulating unit is used for arranging the time sequence of the training physiological parameter data of the monitored high-risk pregnant and lying-in women into a training physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
The training physiological parameter entity content focusing unit is used for focusing the training physiological parameter time sequence matrix based on entity content to obtain a training focusing physiological parameter time sequence matrix;
the training physiological parameter time sequence correlation feature extraction unit is used for extracting features of the training focusing physiological parameter time sequence matrix through a physiological parameter time sequence mode correlation feature extractor based on a deep neural network model so as to obtain a training physiological parameter time sequence correlation feature map;
The multi-scale physiological state characterization unit is used for obtaining a training physiological parameter time sequence multi-scale correlation characteristic diagram through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module;
the optimization unit is used for optimizing each feature matrix of the training physiological parameter time sequence multi-scale associated feature map to obtain an optimized training physiological parameter time sequence multi-scale associated feature map;
the classification loss unit is used for enabling the optimized training physiological parameter time sequence multi-scale associated feature map to pass through a health anomaly detector based on a classifier so as to obtain a classification loss function value;
The training unit is used for training the physiological parameter focusing module based on the entity content attention mechanism, the physiological parameter time sequence mode association feature extractor based on the deep neural network model, the physiological parameter time sequence multi-scale feature enhancement expressive machine based on the Res2Net module and the health abnormality detector based on the classifier based on the classification loss function value.
7. A pregnancy monitoring and management method for a high-risk pregnant and lying-in woman, comprising:
Acquiring a time sequence of physiological parameter data of a monitored high-risk pregnant and lying-in woman acquired by intelligent wearing equipment worn by the monitored high-risk pregnant and lying-in woman, wherein the physiological parameter data comprise heart rate, blood pressure, weight, blood oxygen saturation, uterine contraction frequency and fetal heart rate;
arranging the time sequence of the physiological parameter data of the monitored high-risk pregnant and lying-in women into a physiological parameter time sequence matrix according to the time dimension and the physiological parameter sample dimension;
focusing the physiological parameter time sequence matrix based on the entity content to obtain a focused physiological parameter time sequence matrix;
Extracting features of the focused physiological parameter time sequence matrix through a physiological parameter time sequence mode associated feature extractor based on a deep neural network model to obtain a physiological parameter time sequence associated feature map;
The physiological parameter time sequence correlation characteristic map is passed through a physiological parameter time sequence multi-scale characteristic enhancement expression device based on a Res2Net module to obtain a physiological parameter time sequence multi-scale correlation characteristic map as a physiological parameter time sequence multi-scale correlation characteristic;
determining whether the monitored high-risk pregnant and lying-in women have abnormal conditions or not based on the physiological parameter time sequence multi-scale correlation characteristics;
Focusing the physiological parameter time sequence matrix based on the entity content to obtain a focused physiological parameter time sequence matrix, wherein the focused physiological parameter time sequence matrix is used for: processing the physiological parameter time sequence matrix through a physiological parameter focusing module based on an entity content attention mechanism according to the following focusing formula to obtain the focusing physiological parameter time sequence matrix;
Wherein, the focusing formula is:
；
；
Wherein, For the/>, in the physiological parameter timing matrixParameter value of location,/>Is the/>, in the weight vectorNumerical value of individual position,/>For the weight vector,/>For the physiological parameter timing matrix,/>A time sequence matrix for the focused physiological parameters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410481137.7A CN118098599B (en) | 2024-04-22 | 2024-04-22 | Pregnancy monitoring management system and method for high-risk pregnant and lying-in women |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410481137.7A CN118098599B (en) | 2024-04-22 | 2024-04-22 | Pregnancy monitoring management system and method for high-risk pregnant and lying-in women |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118098599A true CN118098599A (en) | 2024-05-28 |
| CN118098599B CN118098599B (en) | 2024-07-05 |
Family
ID=91142365
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410481137.7A Active CN118098599B (en) | 2024-04-22 | 2024-04-22 | Pregnancy monitoring management system and method for high-risk pregnant and lying-in women |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118098599B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103514577A (en) * | 2013-09-13 | 2014-01-15 | 王得玲 | Area type high-risk pregnant and lying-in woman computer monitoring and evaluating system |
| CN109620211A (en) * | 2018-11-01 | 2019-04-16 | 吉林大学珠海学院 | A kind of intelligent abnormal electrocardiogram aided diagnosis method based on deep learning |
| CN115516103A (en) * | 2020-06-04 | 2022-12-23 | 深圳华大基因股份有限公司 | Method for determining pregnancy status of pregnant woman |
| CN116504382A (en) * | 2023-05-29 | 2023-07-28 | 杭州医锐联科技有限公司 | Remote medical monitoring system and method thereof |
| CN117456068A (en) * | 2023-11-07 | 2024-01-26 | 杭州心向无限科技有限公司 | Video generation method and system based on generation countermeasure network |
| CN117542524A (en) * | 2023-12-18 | 2024-02-09 | 吉林大学 | Intelligent nursing information processing system and method |
| CN117612725A (en) * | 2024-01-23 | 2024-02-27 | 南通大学附属医院 | Respirator alarm management method and system for intensive care unit |
| CN117831733A (en) * | 2024-01-19 | 2024-04-05 | 昆山歆轩电子有限公司 | Intelligent health real-time monitoring system based on wearable equipment and implementation method thereof |
| CN117883093A (en) * | 2024-01-19 | 2024-04-16 | 深圳技术大学 | Fetal heart monitoring method and device based on associated data and deep learning |
-
2024
- 2024-04-22 CN CN202410481137.7A patent/CN118098599B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103514577A (en) * | 2013-09-13 | 2014-01-15 | 王得玲 | Area type high-risk pregnant and lying-in woman computer monitoring and evaluating system |
| CN109620211A (en) * | 2018-11-01 | 2019-04-16 | 吉林大学珠海学院 | A kind of intelligent abnormal electrocardiogram aided diagnosis method based on deep learning |
| CN115516103A (en) * | 2020-06-04 | 2022-12-23 | 深圳华大基因股份有限公司 | Method for determining pregnancy status of pregnant woman |
| CN116504382A (en) * | 2023-05-29 | 2023-07-28 | 杭州医锐联科技有限公司 | Remote medical monitoring system and method thereof |
| CN117456068A (en) * | 2023-11-07 | 2024-01-26 | 杭州心向无限科技有限公司 | Video generation method and system based on generation countermeasure network |
| CN117542524A (en) * | 2023-12-18 | 2024-02-09 | 吉林大学 | Intelligent nursing information processing system and method |
| CN117831733A (en) * | 2024-01-19 | 2024-04-05 | 昆山歆轩电子有限公司 | Intelligent health real-time monitoring system based on wearable equipment and implementation method thereof |
| CN117883093A (en) * | 2024-01-19 | 2024-04-16 | 深圳技术大学 | Fetal heart monitoring method and device based on associated data and deep learning |
| CN117612725A (en) * | 2024-01-23 | 2024-02-27 | 南通大学附属医院 | Respirator alarm management method and system for intensive care unit |
Non-Patent Citations (1)
| Title |
|---|
| 吴富菊等: "" 妊娠期糖尿病与妊娠期高血压疾病病理机制的相关性"", 《国际生殖健康/计划生育杂志》, vol. 39, no. 1, 15 January 2020 (2020-01-15), pages 1 - 4 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118098599B (en) | 2024-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Abdar et al. | UncertaintyFuseNet: robust uncertainty-aware hierarchical feature fusion model with ensemble Monte Carlo dropout for COVID-19 detection | |
| CN109544518B (en) | Method and system applied to bone maturity assessment | |
| Yüzkat et al. | Multi-model CNN fusion for sperm morphology analysis | |
| CN117542474A (en) | Remote nursing monitoring system and method based on big data | |
| CN117598674B (en) | Multi-parameter heart function monitoring system and method | |
| CN117643461B (en) | Heart rate intelligent monitoring system and method based on artificial intelligence | |
| CN117952964B (en) | Fundus medical image analysis method based on computer vision technology | |
| CN118098603B (en) | Intensive care clinical information system and method for eicu | |
| CN117542524A (en) | Intelligent nursing information processing system and method | |
| CN117476214A (en) | Data management method and system based on hospital information | |
| Hasan et al. | FPGA design and hardware implementation of heart disease diagnosis system based on NVG-RAM classifier | |
| CN116269212A (en) | Multi-mode sleep stage prediction method based on deep learning | |
| Mehmood et al. | An intelligent information system and application for the diagnosis and analysis of COVID-19 | |
| CN111047590A (en) | Hypertension classification method and device based on fundus images | |
| CN113855008A (en) | Wearable body fluid monitoring device and operation method thereof | |
| Carvalho et al. | COVID-19 diagnosis in CT images using CNN to extract features and multiple classifiers | |
| Saraei et al. | A different traditional approach for automatic comparative machine learning in multimodality COVID-19 severity recognition | |
| CN117831743A (en) | Household care health monitoring management system and method based on Internet of things technology | |
| Biswas et al. | DFU_MultiNet: A deep neural network approach for detecting diabetic foot ulcers through multi-scale feature fusion using the DFU dataset | |
| CN118098599B (en) | Pregnancy monitoring management system and method for high-risk pregnant and lying-in women | |
| Lei et al. | Patent ductus arteriosus (PDA) detection in echocardiograms using deep learning | |
| Han et al. | Novel Transformation Deep Learning Model for Electrocardiogram Classification and Arrhythmia Detection using Edge Computing | |
| CN116631619A (en) | Postoperative leg bending training monitoring system and method thereof | |
| Kaur et al. | Integrating Transfer Learning and Deep Neural Networks for Accurate Medical Disease Diagnosis from Multi-Modal Data | |
| CN117122286B (en) | Intelligent rehabilitation equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |