Wireless sensor system for monitoring intrauterine activity of fetus

Abstract

The invention provides a wireless sensor system for monitoring fetus intrauterine activities, and belongs to the technical field of collection and monitoring of biological measurement signals, and the system comprises a signal collection module which is used for synchronously collecting abdomen multi-dimensional biological field signals through a heterogeneous biosensor array, and generating an original data stream; the model construction module is used for extracting fetal movement associated feature subspaces and constructing a space-time cause and effect diagram; the intelligent decision-making module is used for determining a fetal movement credibility index and determining an abnormal level through a preset multi-agent decision-making system; the parameter updating module is used for sending the fetal movement credibility index and the abnormal level to a hospital verification node and obtaining feedback data corrected by a doctor so as to update parameters of the space-time cause and effect graph and a preset fetal movement knowledge graph; and the parameter correction module is used for projecting a holographic fetal movement map on the flexible electronic skin interface and acquiring an electromyographic signal adjusted by a pregnant woman to correct the fetal movement credibility index and the abnormal level. And the credibility and reliability of fetal movement monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of acquisition and monitoring of biometric signals, and particularly to a wireless sensor system for monitoring fetal intrauterine activities. Background Art

[0002] With the development of fetal monitoring technology, traditional fetal movement monitoring methods mostly rely on the mother's active perception or simple fetal movement counting. These methods are more dependent on the pregnant woman's subjective judgment and are easily affected by environmental interference, resulting in insufficient accuracy of fetal movement monitoring.

[0003] Existing technologies monitor fetal movements by using sensors, but these technologies usually fail to effectively integrate multi-dimensional biometric signals, making it difficult to evaluate the credibility of fetal movements in real time and accurately, and lacking an adaptive optimization mechanism to cope with dynamically changing physiological conditions. The existing technical solutions cannot accurately and comprehensively collect multi-dimensional biometric signals, and lack an effective intelligent decision-making system to dynamically adjust the credibility index of fetal movements, resulting in possible errors in the monitoring results. In addition, existing solutions usually do not consider the mechanism of hospital doctor feedback and real-time correction, and it is difficult to adaptively adjust monitoring parameters under different environments and individual differences, thus affecting the accuracy and real-time performance of fetal movement monitoring.

[0004] Therefore, the present invention provides a wireless sensor system for monitoring fetal intrauterine activities. Summary of the Invention

[0005] The present invention provides a wireless sensor system for monitoring fetal intrauterine activities, which is used to collect multi-dimensional biofield signals through a wireless sensor system, and in combination with a spatio-temporal causal graph and an intelligent decision-making system, can evaluate the credibility and abnormal level of fetal movements in real time and accurately. Compared with the existing technologies, through an intelligent optimization and doctor feedback mechanism, the fetal movement monitoring parameters are continuously updated and corrected, improving the monitoring accuracy and real-time performance. The combination of a flexible electronic skin interface and electromyographic signals further enhances the adjustment ability of the system, effectively reducing interference factors and improving the credibility and reliability of fetal movement monitoring.

[0006] The present invention provides a wireless sensor system for monitoring fetal intrauterine activities, comprising:

[0007] A signal acquisition module: synchronously acquires multi-dimensional biofield signals of the abdomen through a heterogeneous biosensing array to generate an original data stream;

[0008] A model construction module: extracts a fetal movement correlation feature subspace based on the original data stream and constructs a spatio-temporal causal graph;

[0009] An intelligent decision module: determines the fetal movement credibility index and determines the abnormal level through a preset multi-agent decision-making system based on the spatio-temporal causal graph and a preset fetal movement knowledge graph;

[0010] Parameter update module: Send the fetal movement credibility index and the abnormal level to the hospital verification node and obtain the feedback data corrected by the doctor, and then update the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph;

[0011] Parameter correction module: Project the holographic fetal movement atlas on the flexible electronic skin interface, and obtain the corrected electromyogram signal adjusted by the pregnant woman to correct the fetal movement credibility index and the abnormal level.

[0012] Preferably, the signal acquisition module includes:

[0013] Signal acquisition unit: Fix the heterogeneous biosensing array on the abdomen of the pregnant woman and collect the multi-dimensional biofield signals of the abdomen;

[0014] Signal processing unit: Preprocess the collected signals to generate the original data stream.

[0015] Preferably, the model construction module includes:

[0016] Data encoding unit: Encode the original data stream into a quantum state through a preset encoding method, and obtain the Hamiltonian based on the quantum state through a preset model construction method;

[0017] Feature extraction unit: Determine the ground state through a preset adiabatic evolution algorithm and Hamiltonian, and obtain the fetal movement correlation feature subspace based on the ground state through a preset expected value calculation method;

[0018] Model construction unit: Determine the spatio-temporal causal graph based on the fetal movement correlation feature subspace.

[0019] Preferably, the intelligent decision-making module includes:

[0020] Sequence reconstruction unit: Map the fetal movement correlation feature subspace to a preset biomechanical simulation engine to reconstruct the fetal limb movement vector sequence;

[0021] Atlas matching unit: Call the preset fetal movement knowledge atlas to match the topological similarity of the spatio-temporal causal graph through a preset graph neural network;

[0022] Sequence encoding unit: Perform dynamic trajectory encoding on the movement vector sequence to generate a four-dimensional spatio-temporal movement fingerprint;

[0023] Decision-making unit: Input the four-dimensional spatio-temporal movement fingerprint and the topological similarity score into a preset multi-agent decision-making system, and then generate the fetal movement credibility index and the abnormal level.

[0024] Preferably, the atlas matching unit includes:

[0025] Node matching subunit: Perform node matching on the spatio-temporal causal graph and the fetal movement knowledge graph, determine the corresponding node pairs, and generate a set of matching node pairs;

[0026] Edge determination subunit: Determine the corresponding edges in the fetal movement knowledge graph for each edge of the spatio-temporal causal graph based on the set of matching node pairs;

[0027] Matrix generation subunit: Determine the semantic distance of the corresponding edges in the fetal movement knowledge graph for each edge of the spatio-temporal causal graph, and generate an edge semantic distance matrix;

[0028] Similarity calculation subunit: Determine the topological similarity based on the edge semantic distance matrix.

[0029] Preferably, the parameter update module includes:

[0030] Data transmission unit: Transmit the fetal movement credibility index and abnormalities, etc. to the hospital verification node through a preset quantum encryption channel;

[0031] Feedback acquisition unit: Obtain the feedback data of the doctor at the hospital verification node;

[0032] Parameter update unit: Update the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph based on the doctor's feedback data.

[0033] Preferably, the parameter update unit includes:

[0034] Data parsing subunit: Parse the four-dimensional spatio-temporal fingerprint based on the doctor's feedback data at the hospital verification node, and mark the false trace segments in the four-dimensional spatio-temporal fingerprint;

[0035] Error determination subunit: Determine the error ratio based on the false trace segments, and then update the parameters of the spatio-temporal causal graph;

[0036] Graph update subunit: Update the weights of the fetal movement knowledge graph based on the doctor's feedback data through a preset differential privacy protection mechanism.

[0037] Preferably, the parameter correction module includes:

[0038] Pressure acquisition unit: Detect the pressure distribution when the pregnant woman's finger presses based on the microfluidic tactile units distributed on the surface of the flexible electronic skin, and determine the pressure distribution matrix;

[0039] Trajectory correction unit: Determine the trajectory correction amount based on the pressure distribution matrix;

[0040] Fingerprint update unit: Update the four-dimensional spatio-temporal motion fingerprint based on the trajectory correction vector;

[0041] Parameter correction unit: Input the updated four-dimensional spatio-temporal motion fingerprint into a preset multi-agent decision system to obtain the updated fetal movement credibility index and abnormal level.

[0042] Compared with the prior art, the beneficial effects of the present application are as follows:

[0043] By collecting multi-dimensional biofield signals through a wireless sensor system and combining a spatio-temporal causal graph with an intelligent decision-making system, it is possible to evaluate the credibility and abnormal level of fetal movement in real time and accurately. Compared with the prior art, through intelligent optimization and doctor feedback mechanisms, the fetal movement monitoring parameters are continuously updated and corrected, improving the monitoring accuracy and real-time performance. The combination of a flexible electronic skin interface and electromyogram signals further enhances the system's adjustment ability, effectively reducing interference factors and improving the credibility and reliability of fetal movement monitoring. Description of the Drawings

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0045] Figure 1 is a schematic structural diagram of a wireless sensor system for monitoring fetal intrauterine activities provided by an embodiment of the present invention. Detailed Embodiments

[0046] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0047] Embodiment 1:

[0048] An embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, as Figure 1 shown, including:

[0049] Signal acquisition module: Synchronously collect multi-dimensional biofield signals of the abdomen through a heterogeneous biosensing array to generate an original data stream;

[0050] Model construction module: Extract a fetal movement correlation feature subspace based on the original data stream and construct a spatio-temporal causal graph;

[0051] Intelligent decision module: Determine the fetal movement credibility index and determine the abnormal level through a preset multi-agent decision-making system based on the spatio-temporal causal graph and a preset fetal movement knowledge graph;

[0052] Parameter Update Module: Send the fetal movement credibility index and the abnormality level to the hospital verification node and obtain the feedback data corrected by the doctor, and then update the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph;

[0053] Parameter Correction Module: Project a holographic fetal movement map on the flexible electronic skin interface, and obtain the myoelectric signals adjusted by the pregnant woman to correct the fetal movement credibility index and the abnormality level.

[0054] In this embodiment, the heterogeneous biosensing array refers to a sensor array composed of multiple different types of sensors, including: quantum dot strain sensors, photoacoustic tomography probes, electromagnetic topological grids), which can synchronously collect multi-dimensional data from different biological signal sources;

[0055] In this embodiment, the abdominal multi-dimensional biofield signal, generating the original data stream refers to a variety of biological signals collected from the pregnant woman's abdomen through different types of sensors, generating an original data stream containing mechanical deformation, blood flow fluctuation and electromagnetic topology;

[0056] In this embodiment, the fetal movement associated feature subspace is a set of features extracted from the original data stream, used to represent the specific patterns or features of fetal movement, and these features help to distinguish normal fetal movement from abnormal fetal movement;

[0057] In this embodiment, the spatio-temporal causal graph is a graphical model that describes the causal relationship and time evolution between signals and events. It associates the changes in signals with the corresponding time and space positions, helping to analyze the causal relationship between fetal movement and external environment or physiological factors. Example: If there is a causal relationship between fetal movement and the activities of the pregnant woman, the spatio-temporal causal graph can show whether the occurrence of fetal movement is related to a certain activity of the pregnant woman (such as walking, staying still, etc.).

[0058] In this embodiment, the preset fetal movement knowledge graph is a graphical representation of fetal movement-related knowledge constructed based on existing medical research and clinical experience. It includes information such as fetal movement patterns, abnormal manifestations, and the status of the pregnant woman, and is used to assist system decision-making. Example: The knowledge graph may include the normal range of fetal movement, the definition of abnormal fetal movement, and the possible causes of different fetal movement patterns.

[0059] In this embodiment, the flexible electronic skin interface is a bendable and skin-fitting electronic device, usually used for real-time monitoring and feedback of biological signals, and can be directly in contact with the human body. Example: The flexible electronic skin interface is a wearable device, like a thin electronic film, which can fit the abdomen of the pregnant woman and display the fetal movement map in real time or provide other feedback information.

[0060] In this embodiment, the myoelectric signals adjusted by the pregnant woman refer to the electrophysiological signals generated by the voluntary muscle activities of the pregnant woman, and these signals can be used to adjust or correct the fetal movement credibility index and its abnormal level. For example, when the pregnant woman generates muscle activities due to certain reasons (such as discomfort, tension or relaxation), the flexible electronic skin interface can capture these myoelectric signals and use them to adjust the analysis results of the system to ensure more accurate monitoring of fetal movement.

[0061] The beneficial effects of the above technical solution: By collecting multi-dimensional biofield signals through a wireless sensor system and combining a spatio-temporal causal graph with an intelligent decision-making system, it is possible to evaluate the credibility and abnormal level of fetal movement in real time and accurately. Compared with the prior art, through intelligent optimization and doctor feedback mechanism, the fetal movement monitoring parameters are continuously updated and corrected, improving the monitoring accuracy and real-time performance. The combination of the flexible electronic skin interface and myoelectric signals further enhances the adjustment ability of the system, effectively reducing interference factors and improving the credibility and reliability of fetal movement monitoring.

[0062] Embodiment 2:

[0063] The embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, and the signal acquisition module includes:

[0064] Signal acquisition unit: Fix a heterogeneous biosensing array on the abdomen of the pregnant woman to collect multi-dimensional biofield signals on the abdomen;

[0065] Signal processing unit: Preprocess the collected signals to generate an original data stream.

[0066] In this embodiment, a heterogeneous biosensing array is fixed on the abdomen of the pregnant woman to collect multi-dimensional biofield signals on the abdomen. For example, it includes: Deploying a heterogeneous biosensing array: Quantum dot strain sensor: Using CdSe / ZnS core-shell structure quantum dots with a diameter of 5±0.2 nm; the width / spacing of the interdigital electrodes is 10 μm / 15 μm, and the substrate is a PDMS-graphene composite material; collecting the abdominal surface strain tensor ∈(t)=[∈xx,∈xy,∈yy] with a sampling rate of 500 Hz; Photoacoustic tomography probe: Emitting a 10 MHz pulsed laser and receiving the frequency shift signal Δf(t) through an ultrasonic transducer array; Solving the blood flow velocity field; Electromagnetic topology grid: Emitting non-ionizing electromagnetic waves in the 0.5-3 THz frequency band and reconstructing the abdominal electromagnetic conductivity distribution σEM(x,y,z,t) through the reflection coefficient phase difference ΔΦ(x,y,t);

[0067] In this embodiment, the collected signals are preprocessed to generate an original data stream, including: performing noise reduction processing on the collected signals: using wavelet transform to separate the fetal movement signals from the maternal breathing / heartbeat interference; eliminating environmental noise through an adaptive filter; outputting a standardized data stream: DataStream = {∈(t), σEM(x, y, z, t), v(x, y, z, t)}, where v(x, y, z, t) is the blood flow velocity.

[0068] Advantages of the above technical solution: By fixing a heterogeneous biosensing array on the abdomen of a pregnant woman, precise synchronous acquisition of multi-dimensional biofield signals is achieved. Compared with the prior art, it can comprehensively monitor the intrauterine activities of the fetus, covering biosignals such as fetal movement and heartbeat. Through the preprocessing function of the signal processing unit, the generated original data stream can be more effectively analyzed for subsequent analysis, providing more accurate fetal movement monitoring, helping to evaluate the fetal health status in real time, and improving the monitoring efficiency and accuracy.

[0069] Embodiment 3:

[0070] An embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, including a model construction module:

[0071] Data encoding unit: Encoding the original data stream into a quantum state through a preset encoding method, and obtaining a Hamiltonian based on the quantum state through a preset model construction method;

[0072] Feature extraction unit: Determining the ground state through a preset adiabatic evolution algorithm and the Hamiltonian, and obtaining a fetal movement correlation feature subspace based on the ground state through a preset expected value calculation method;

[0073] Model construction unit: Determining a spatio-temporal causal graph based on the fetal movement correlation feature subspace.

[0074] In this embodiment, the original data stream is encoded into a quantum state through a preset encoding method, and the preset encoding method is a quantum encoding method, and the quantum state is:

[0075]

[0076] This formula represents the quantum state of a multi-qubit system, where: N represents the number of qubits, and the state of each qubit consists of two parts: |0> and |1>, which are the ground state and the excited state respectively. Through the angle θ k and the phase φ k control the state of each qubit, where, This expression describes the qubit rotation angle θ rz (t k )) related to the strain tensor (∈ k ), which reflects the strain information of the system, ∈min and ∈ max are the minimum and maximum values of the strain respectively, This formula represents the phase information φ of the qubit related to the electromagnetic field k , σ EM (t k ) is a certain measured value of the electromagnetic field, σ min and σ max are the minimum and maximum values of the electromagnetic field. Physical scenario corresponding to each qubit: It is stated in the formula that each qubit is associated with a specific physical scenario. Through different combinations of qubits, encoding of multiple physical quantities can be achieved.

[0077] The Hamiltonian is obtained based on the quantum state through a preset model construction method. It is obtained through the Ising model construction method. The specific steps are as follows: Input: Quantum state |ψ>

[0078]

[0079] Among them, H(t) is the Hamiltonian, which determines the evolution of the system over time. A(t) and B(t) are time-dependent functions that control the relative importance of different terms in the system. A(t) and B(t) are used to control the change of the system over time. A(0) = 1, B(0) = 0 represent the initial conditions, while A(T) = 0, B(T) = 1 represent the final conditions. represents the Pauli Z operator, acting on the i-th qubit, A(t)∑ i is the single-body term, representing the interaction of all qubits in the Z-axis direction. The second term is the interaction term, which consists of two parts: This is an interaction term between qubits, This is an interaction term between a single qubit and the external field, J ij is the coupling strength between qubits i and j, representing the electromagnetic interaction strength between qubits i and j, which depends on their relative position |r i -r j |. Its calculation formula is:

[0080]

[0081] This is a function that decays according to the relative distance. α is a constant, which is usually related to the physical properties of the system (such as the decay of the coupling strength). In this expression, the role of α is to control the decay of J ij as the relative distance |r i -r j | between qubits increases, h iis the intensity of the external field, representing the local external field intensity of each qubit, usually determined by physical quantities such as fluid flow rate, magnetic field, etc.

[0082] Output: The Hamiltonian of the Ising model.

[0083] In this embodiment, based on the Hamiltonian H(t), through the adiabatic evolution method, the ground state |ψ g s> is obtained: Input: Hamiltonian H(t), evolving from the initial Hamiltonian H(0) to the target Hamiltonian H(T) through adiabatic evolution, and the ground state |ψ g s> is obtained:

[0084]

[0085] Among them, the initial Hamiltonian H(0): represents the initial state of the system, which consists of the Z-axis parts of all qubits, and the initial ground state is |0>. The target Hamiltonian H(T): the final Hamiltonian, including the interaction term and the local external field term between qubits. Asymptotic condition:

[0086] Through adiabatic evolution, the system gradually evolves from the initial Hamiltonian H(0) to the target Hamiltonian H(T), and finally reaches the ground state |ψ g s>, and satisfies the following conditions:

[0087] <ψ(T)|H(T)|ψ(T)> ≤ 1

[0088] This ensures that the system does not exceed the range described by the target Hamiltonian during the evolution process. Output: Ground state |ψ g s>: represents the final ground state of the system, which is obtained through the adiabatic evolution method.

[0089] Beneficial effects of the above technical solution: By combining quantum state encoding and the adiabatic evolution algorithm, the processing accuracy and efficiency of fetal movement signals are improved. Compared with the prior art, using the Hamiltonian and the expected value calculation method to extract fetal movement correlation features can more accurately identify fetal movement patterns, construct a high-precision spatio-temporal causal graph, enhance the intelligent decision-making ability of the system, provide more reliable and accurate fetal movement monitoring, and provide stronger guarantee for fetal health.

[0090] Embodiment 4:

[0091] The embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, and an intelligent decision-making module, including:

[0092] Sequence reconstruction unit: mapping the fetal movement correlation feature subspace to a preset biomechanical simulation engine to reconstruct the fetal limb movement vector sequence;

[0093] Spectrum matching unit: Call the pre-set fetal movement knowledge graph and match the topological similarity of the spatio-temporal causal graph through the pre-set graph neural network;

[0094] Sequence encoding unit: Perform dynamic trajectory encoding on the motion vector sequence to generate a four-dimensional spatio-temporal motion fingerprint;

[0095] Decision-making unit: Input the four-dimensional spatio-temporal motion fingerprint and the topological similarity score into the pre-set multi-agent decision-making system, and then generate a fetal movement credibility index and an abnormal level.

[0096] In this embodiment, the fetal limb movement vector sequence refers to detecting the fetal limb movement in the uterus through a sensor and converting these movements into vector form. The vector usually includes information such as the movement direction, speed, and acceleration here, forming a set of vector sequences representing the movement state. Example: The movement data at each moment may contain the following information: Time point t1: The fetal right hand moves 2 cm to the right, with a speed of 0.5 cm / s. Time point t2: The fetal right hand moves 3 cm upward, with a speed of 0.8 cm / s. Time point t3: The fetal right hand is stationary. These data can be represented as a movement vector sequence: V1=(2,0.5), V2=(3,0.8), V3=(0,0)V1=(2,0.5), V2=(3,0.8), V3=(0,0). This sequence is the "fingerprint" of the fetal movement and can be used for subsequent analysis. Based on the fetal movement correlation feature subspace, through the biomechanical simulation engine, the fetal limb movement vector sequence is obtained, including: the fetal movement correlation feature subspace FF; Method: Map F to the finite element biomechanical model and solve the motion equation:

[0097]

[0098] Among them, σ = C: ∈ is the constitutive equation, C is the material stiffness matrix; f is the fetal acting force, derived from the feature subspace F; and then output: the fetal limb movement vector sequence u(x,y,z,t);

[0099] In this embodiment, the dynamic trajectory encoded four-dimensional spatio-temporal fingerprint maps the fetal motion vector sequence into a four-dimensional spatio-temporal coordinate system and encodes it. The four-dimensional spatio-temporal fingerprint usually includes information such as time, spatial position, and motion state, and represents the entire process of motion by combining the spatial and temporal dimensions. This encoding assigns a unique identifier to each motion trajectory. Example: Suppose a fetal limb movement occurs according to the following time sequence: Time t1: The fetal hand is at position (x1, y1, z1) and moves at a speed V1 = (0.5, 0.8). Time t2: The fetal hand is at position (x2, y2, z2) and moves at a speed V2 = (1.0, 1.3). Time t3: The fetal hand is at position (x3, y3, z3) and moves at a speed V3 = (0.2, 0.5). The information at each time point includes information such as time, spatial position, and speed. The four-dimensional spatio-temporal fingerprint can be expressed as: Fingerprint1 = (t1, x1, y1, z1, 0.5, 0.8), Fingerprint2 = (t2, x2, y2, z2, 1.0, 1.3), Fingerprint3 = (t3, x3, y3, z3, 0.2, 0.5). This encoding method encodes the motion trajectory at each moment and combines spatial and temporal factors to form a unique fingerprint in the four-dimensional spatio-temporal space.

[0100] In this embodiment, based on the fetal limb motion vector sequence u(x, y, z, t), through the dynamic trajectory encoding method, the four-dimensional spatio-temporal motion fingerprint Fingerprint is obtained: Input: Fetal limb motion vector sequence u(x, y, z, t); Method: Perform Fourier transform on u(x, y, z, t) to generate the four-dimensional spatio-temporal motion fingerprint:

[0101]

[0102] where: m, n, k, l are frequency indices; w x , w y , w z , w t are spatial and temporal frequencies; and then output: Four-dimensional spatio-temporal motion fingerprint Fingerprint = {Cmnkl}.

[0103] In this embodiment, the four-dimensional spatio-temporal motion fingerprint and the topological similarity score of path C2 are input into the multi-agent decision-making system. Among them, the multi-agent decision-making system includes: Agent A: Calculate the Nash equilibrium solution of the motion trajectory and the normal development pattern based on game theory; Agent B: Optimize the matching weight of the pathological pattern using a bio-inspired algorithm; Integrate the output results of Agent A / B to generate a fetal movement credibility index and an abnormal level assessment.

[0104] Advantages of the above technical solution: Through sequence reconstruction and graph matching technologies, efficient extraction of fetal movement characteristics and intelligent decision-making are achieved. Compared with the prior art, the fetal movement vectors can be reconstructed by a biomechanical simulation engine, and the topological similarity of the spatio-temporal causal graph can be matched by using a graph neural network, so as to accurately identify fetal movement patterns. The combination of dynamic trajectory encoding and multi-agent decision-making system provides higher accuracy and adaptability, and can effectively improve the reliability of fetal movement monitoring and the ability of abnormal detection.

[0105] Embodiment 5:

[0106] The embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities. The graph matching unit includes:

[0107] Node matching sub-unit: Perform node matching on the spatio-temporal causal graph and the fetal movement knowledge graph to determine the corresponding node pairs, and generate a set of matching node pairs;

[0108] Edge determination sub-unit: Determine the corresponding edges of each edge of the spatio-temporal causal graph in the fetal movement knowledge graph based on the set of matching node pairs;

[0109] Matrix generation sub-unit: Determine the semantic distance of the corresponding edges of each edge of the spatio-temporal causal graph in the fetal movement knowledge graph, and generate an edge semantic distance matrix;

[0110] Similarity calculation sub-unit: Determine the topological similarity based on the edge semantic distance matrix.

[0111] In this embodiment, performing node matching on the spatio-temporal causal graph and the fetal movement knowledge graph to determine the corresponding node pairs, and generating a set of matching node pairs includes:

[0112] Input: Spatio-temporal causal graph G = (V, E), fetal movement knowledge graph Gknowledge = (Vknowledge, Eknowledge); Method: Traverse each node v i ∈V in the spatio-temporal causal graph G, and search for a semantically similar node v j ∈Vknowledge in the fetal movement knowledge graph Gknowledge. Node matching condition: Sim(v i , v j ) > τ, where: Sim(v i , v j ) is the semantic similarity between node v i and v j , and the calculation method is:

[0113] Where and are the vectors of node v i and v jthe k-th attribute (fetal movement type, concurrent symptoms, time evolution pattern, intervention effectiveness), is an attribute matching function that returns 0 or 1; ω k is the attribute weight, satisfying τ is the similarity threshold, usually taking the value of 0.7; Output: a set of matching node pairs {(v i , v j )}.

[0114] Beneficial effects of the above technical solution: By accurately matching the spatio-temporal causal graph with the fetal movement knowledge graph through the graph matching unit, the relationship between fetal movement characteristics and the knowledge graph can be effectively identified. Compared with the prior art, the node matching, edge determination, and generation of the semantic distance matrix improve the recognition accuracy of the fetal movement pattern by the system. By calculating the topological similarity, the system can more accurately judge the credibility and abnormal conditions of fetal movement, improve the monitoring effect and decision-making accuracy, and contribute to providing more scientific and personalized fetal movement monitoring.

[0115] Example 6:

[0116] The embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, and a parameter update module, including:

[0117] Data transmission unit: Transmits the fetal movement credibility index and abnormalities, etc. to the hospital verification node through a preset quantum encryption channel;

[0118] Feedback acquisition unit: Obtains the feedback data of the doctor at the hospital verification node;

[0119] Parameter update unit: Updates the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph based on the doctor's feedback data.

[0120] In this embodiment, the hospital verification node refers to a system node set in a hospital environment for receiving and verifying fetal movement monitoring data. This node can receive the fetal movement credibility index and abnormal level data transmitted through the quantum encryption channel, and provide the data feedback by the doctor. The core role of the hospital verification node is to conduct professional review on the monitoring results generated by the system, ensure the accuracy of the data, and adjust the monitoring parameters and knowledge graph based on the doctor's feedback to improve the intelligent decision-making and accuracy of the system. For example, when the wireless sensor system monitors that the fetal movement of a pregnant woman is abnormal and generates the corresponding credibility index, these data are transmitted to the hospital verification node through the encryption channel. In the hospital, the doctor views the data provided by the system and confirms the actual situation of the fetal movement. If the system makes a wrong judgment, the doctor will provide feedback through the hospital verification node, and the system will immediately update the parameters of the spatio-temporal causal graph and the fetal movement knowledge graph according to the feedback to further optimize the monitoring effect.

[0121] Advantages of the above technical solution: The secure transmission of the fetal movement credibility index and abnormal data is achieved through a quantum encryption channel, ensuring data privacy and security. Compared with the prior art, the system can obtain doctor feedback at the hospital verification node and update the parameters of the spatio-temporal causal graph and the fetal movement knowledge graph in a timely manner based on the professional opinions of doctors. This feedback mechanism continuously optimizes the monitoring accuracy and intelligent decision-making ability of the system, improves the accuracy and real-time performance of fetal movement monitoring, and ensures the effective monitoring of fetal health.

[0122] Embodiment 7:

[0123] An embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities, and a parameter update unit, including:

[0124] A data parsing subunit: Parse the four-dimensional spatio-temporal fingerprint based on the feedback data of doctors at the hospital verification node, and mark the false trace segments in the four-dimensional spatio-temporal fingerprint;

[0125] An error determination subunit: Determine the error ratio based on the false trace segments, and then update the parameters of the spatio-temporal causal graph;

[0126] A graph update subunit: Update the weights of the fetal movement knowledge graph based on the feedback data of doctors through a preset differential privacy protection mechanism.

[0127] In this embodiment, parsing the four-dimensional spatio-temporal fingerprint based on the feedback data of doctors at the hospital verification node and marking the false trace segments in the four-dimensional spatio-temporal fingerprint are based on the corrected feedback of doctors at the hospital verification node. Use a topological data analyzer to parse the four-dimensional spatio-temporal fingerprint; compare the current movement trajectory with the doctor's corrected trajectory, and then mark the difference area, and then obtain the false trace segments in the four-dimensional spatio-temporal fingerprint.

[0128] In this embodiment, based on the marked false trajectory segments, determine the error ratio, and then determine the quantum bit phase rotation parameter based on the error ratio;

[0129] In this embodiment, based on the backpropagation algorithm, determine the loss function, and update the parameters of the spatio-temporal causal graph in combination with the quantum bit phase rotation parameter and the gradient descent method.

[0130] Advantages of the above technical solution: Through the data parsing and error determination subunits, false trace segments in the four-dimensional spatio-temporal fingerprint can be effectively identified and marked, thereby reducing the impact of errors on the accuracy of the system. Compared with the prior art, the system optimizes the parameters of the spatio-temporal causal graph through doctor feedback data and uses the differential privacy protection mechanism to securely update the weights of the fetal movement knowledge graph, ensuring privacy protection while improving the monitoring accuracy. This closed-loop feedback mechanism significantly improves the reliability and accuracy of fetal movement monitoring.

[0131] Embodiment 8:

[0132] An embodiment of the present invention provides a wireless sensor system for monitoring fetal intrauterine activities. The parameter correction module includes:

[0133] A pressure acquisition unit: Based on the microfluidic tactile units distributed on the surface of the flexible electronic skin, it detects the pressure distribution when the pregnant woman's finger presses, and determines the pressure distribution matrix;

[0134] A trajectory correction unit: Determines the trajectory correction amount based on the pressure distribution matrix;

[0135] A fingerprint update unit: Updates the four-dimensional spatio-temporal motion fingerprint based on the trajectory correction vector;

[0136] A parameter correction unit: Inputs the updated four-dimensional spatio-temporal motion fingerprint into a preset multi-agent decision system to obtain the updated fetal movement credibility index and abnormal level.

[0137] In this embodiment, the pressure distribution matrix is obtained by the microfluidic tactile units on the flexible electronic skin sensing the pressure data generated when the pregnant woman's finger presses. These pressure data reflect the pressure distribution when the finger contacts the sensor surface. By converting these pressure data into matrix form, the system can capture the pressure intensity and distribution pattern at different positions, and then be used to analyze the subtle changes in fetal movement. Example: Suppose the pregnant woman's finger presses on different positions of the electronic skin, and the sensor will sense the pressure at each position. For example, the pressure at the fingertip may be relatively large, while the pressure at the root of the palm is relatively small. The pressure distribution matrix records the pressure values of each sensor node. For example:

[0138]

[0139] This matrix represents the pressure values of the finger in different regions.

[0140] In this embodiment, the trajectory correction amount refers to the adjustment amount of the fetal movement trajectory according to the pressure distribution matrix. By analyzing the pressure change, the system can correct the trajectory deviation caused by factors such as finger pressing, so as to more accurately reflect the real movement trajectory of the fetus. Example: If at certain pressing points, the pressure is relatively high, it may cause the sensor to misjudge the deviation of the fetal movement trajectory, and the trajectory correction amount will be adjusted according to these pressure values. For example, if the pressure matrix shows that the pressure in a certain area is relatively high, the trajectory correction amount may calculate a correction value through a certain algorithm, indicating the correction to be made to the trajectory, so as to eliminate the error caused by pressing.

[0141] Advantages of the above technical solution: By precisely perceiving the pressure distribution when the pregnant woman's finger presses through the flexible electronic skin and the microfluidic tactile unit, and effectively correcting the fetal movement trajectory using the pressure distribution matrix, the accuracy of monitoring is improved. Compared with the prior art, the calculation of the trajectory correction amount and the update of the four-dimensional spatio-temporal motion fingerprint enable the system to more accurately reflect the fetal movement pattern. The combination of the multi-agent decision-making system further improves the accuracy of the fetal movement credibility index and the abnormal level, ensuring the reliability and real-time nature of fetal activity monitoring.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wireless sensor system for monitoring fetal intrauterine activities, characterized in that, Including: Signal acquisition module: Synchronously acquire multi-dimensional biofield signals in the abdomen through a heterogeneous biosensing array to generate an original data stream; Model construction module: Extract a fetal movement associated feature subspace based on the original data stream and construct a spatio-temporal causal graph; Intelligent decision-making module: Determine the fetal movement credibility index and the abnormal level through a preset multi-agent decision-making system based on the spatio-temporal causal graph and a preset fetal movement knowledge graph; Parameter update module: Send the fetal movement credibility index and the abnormal level to the hospital verification node and obtain the feedback data corrected by the doctor, and then update the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph; Parameter correction module: Project a holographic fetal movement map on the flexible electronic skin interface, and obtain the myoelectric signals adjusted by the pregnant woman to correct the fetal movement credibility index and the abnormal level.

2. The wireless sensor system for monitoring fetal intrauterine activities according to claim 1, characterized in that The signal acquisition module includes: Signal acquisition unit: Fix a heterogeneous biosensing array on the abdomen of the pregnant woman to acquire multi-dimensional biofield signals in the abdomen; Signal processing unit: Preprocess the acquired signals to generate an original data stream.

3. The wireless sensor system for monitoring fetal intrauterine activities according to claim 1, characterized in that, The model construction module includes: Data encoding unit: Encode the original data stream into a quantum state through a preset encoding method, and obtain a Hamiltonian based on the quantum state through a preset model construction method; Feature extraction unit: Determine the ground state through a preset adiabatic evolution algorithm and the Hamiltonian, and obtain a fetal movement associated feature subspace based on the ground state through a preset expected value calculation method; Model construction unit: Determine a spatio-temporal causal graph based on the fetal movement associated feature subspace.

4. A wireless sensor system for monitoring fetal intrauterine activities according to claim 1, characterized in that, The intelligent decision-making module includes: Sequence reconstruction unit: Map the fetal movement associated feature subspace to a preset biomechanical simulation engine to reconstruct the fetal limb movement vector sequence; Map matching unit: Call a preset fetal movement knowledge graph to match the topological similarity of the spatio-temporal causal graph through a preset graph neural network; Sequence encoding unit: Perform dynamic trajectory encoding on the movement vector sequence to generate a four-dimensional spatio-temporal movement fingerprint; Decision-making unit: Input the four-dimensional spatio-temporal movement fingerprint and the topological similarity score into a preset multi-agent decision-making system, and then generate a fetal movement credibility index and an abnormal level.

5. A wireless sensor system for monitoring fetal intrauterine activity according to claim 4, characterized in that, The map matching unit includes: Node matching sub-unit: Perform node matching on the spatio-temporal causal graph and the fetal movement knowledge graph to determine the corresponding node pairs and generate a set of matching node pairs; Edge determination sub-unit: Determine the corresponding edges of each edge of the spatio-temporal causal graph in the fetal movement knowledge graph based on the set of matching node pairs; Matrix generation sub-unit: Determine the semantic distance of the corresponding edges of each edge of the spatio-temporal causal graph in the fetal movement knowledge graph, and generate an edge semantic distance matrix; Similarity calculation sub-unit: Determine the topological similarity based on the edge semantic distance matrix.

6. The wireless sensor system for monitoring fetal intrauterine activities according to claim 1, wherein The parameter update module includes: Data transmission unit: Transmit the fetal movement credibility index and the abnormality through a preset quantum encryption channel to the hospital verification node; Feedback acquisition unit: Obtain the feedback data of the doctor at the hospital verification node; Parameter update unit: Update the parameters of the spatio-temporal causal graph and the preset fetal movement knowledge graph based on the feedback data of the doctor.

7. A wireless sensor system for monitoring fetal intrauterine activities according to claim 4, characterized in that, The parameter update unit includes: Data analysis sub-unit: Analyze the four-dimensional spatio-temporal fingerprint based on the feedback data of the doctor at the hospital verification node, and mark the false trace segments in the four-dimensional spatio-temporal fingerprint; Error determination subunit: Determine the error ratio based on the false trace segment, and then update the parameters of the spatio-temporal causal graph; Atlas update subunit: Update the weights of the fetal movement knowledge atlas based on the feedback data of doctors through a preset differential privacy protection mechanism.

8. A wireless sensor system for monitoring fetal intrauterine activities according to claim 1, characterized in that, Parameter correction module, including: Pressure acquisition unit: Detect the pressure distribution when the pregnant woman's finger presses based on the microfluidic tactile units distributed on the surface of the flexible electronic skin, and determine the pressure distribution matrix; Trajectory correction unit: Determine the trajectory correction amount based on the pressure distribution matrix; Fingerprint update unit: Update the four-dimensional spatio-temporal motion fingerprint based on the trajectory correction vector; Parameter correction unit: Input the updated four-dimensional spatio-temporal motion fingerprint into a preset multi-agent decision system to obtain the updated fetal movement credibility index and abnormal level.

Images