Virus evolution trend prediction method and system based on time sequence electric signals

Abstract

The embodiment of the invention discloses a time sequence electric signal-based virus evolution trend prediction method and system. The method comprises the following steps of: acquiring bio-electricity signal time sequence data generated by a host individual infected by a virus in real time; performing framing processing and feature deconstruction on the time series data, extracting an electrophysiological disturbance mode representing an infection process, and constructing a virus evolution disturbance feature sequence; establishing an implicit mapping relation between host electrical response and virus adaptive evolution, and converting the disturbance characteristic sequence into a virus mutation pressure indication factor; a time sequence evolution model is constructed based on the indicator factors, the potential evolution trajectory of viruses in a future time window is deduced, and the time sequence evolution model adopts a bidirectional disturbance propagation mechanism and applies trajectory self-consistency constraint; and generating a virus strain differentiation prediction path by taking the electric signal disturbance similarity as a branch basis, and outputting an evolution trend pedigree diagram. According to the method, non-intrusive and indirect prediction of the virus evolution trend is realized, and the detection cost and time are remarkably reduced.

Description

Technical Field

[0001] This application relates to the field of highly pathogenic pathogen monitoring and early warning technology, combining methods from the fields of biomedical engineering and signal processing technology. In particular, it relates to a method and system for predicting viral evolution trends based on time-series electrical signals, which is applicable to non-invasive dynamic tracking and evolution risk assessment of highly infectious and highly variable pathogens. Background Technology

[0002] Currently, predictions of the evolutionary trends of highly pathogenic pathogens (such as Ebola virus and Nipah virus) mainly rely on viral genome sequencing and bioinformatics analysis. For example, host fitness is predicted through the coding of viral gene datasets and machine learning models, or viral-host associations are predicted based on network fusion and graph embedding. While these methods are effective, they have significant limitations: First, they require direct sampling of the viral genome, a process that is invasive, time-consuming, and dependent on laboratory equipment. Second, the predictions are mostly static matches, unable to capture the dynamic evolutionary path of the virus in real time, leading to delayed early warnings and an inability to effectively respond to sudden public health threats caused by highly variable pathogens. Third, they ignore the potential of host physiological responses as proxy indicators, causing predictions to lag behind the actual infection process, failing to achieve early warnings and limiting the sensitivity and timeliness of predictions. In addition, traditional methods lack in-depth exploration of individualized perturbation patterns and multi-scale signal deconstruction, making it difficult to handle the high uncertainty in viral evolution.

[0003] Therefore, in the field of highly pathogenic pathogen surveillance, there is an urgent need to develop a novel, non-invasive prediction method that can reflect the dynamics of viral evolution in real time, so as to achieve early identification and risk assessment of the evolutionary paths of high-risk viruses. Summary of the Invention

[0004] This application provides a method and system for predicting virus evolution trends based on time-series electrical signals, which enables non-invasive, indirect prediction of virus evolution trends and significantly reduces detection costs and time.

[0005] This application provides the following solution: According to the first aspect, a method for predicting viral evolution trends based on time-series electrical signals is provided. The method includes: real-time acquisition of time-series bioelectrical signal data generated by a virus-infected host individual, wherein the signals are selected from at least one of electroencephalogram (EEG), electrocardiogram (ECG), or electromyogram (EMG), and the signals reflect cell membrane potential perturbations or ion channel dynamics; performing frame-by-frame processing and feature deconstruction on the time-series data to extract electrophysiological perturbation patterns characterizing the infection process and constructing a viral evolution perturbation feature sequence; establishing an implicit mapping relationship between the host's electrical response and viral adaptive evolution, and converting the perturbation feature sequence into a viral mutation pressure indicator; constructing a time-series evolution model based on the indicator to infer the potential evolutionary trajectory of the virus within a future time window, wherein the time-series evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints; generating a viral strain differentiation prediction path based on the similarity of electrical signal perturbations and outputting an evolution trend phylogenetic map.

[0006] According to one achievable method in the embodiments of this application, the step of performing frame-based processing and feature deconstruction on the time-series data includes: using a sliding window algorithm to divide the time-series data into frames, with each frame having a length of 5 to 30 minutes; performing an individualized comparison of each frame signal with the corresponding baseline frame to calculate the disturbance offset; and performing multi-scale deconstruction on the disturbance offset, decomposing it into frequency domain components, time domain statistics, and nonlinear dynamic indicators.

[0007] According to one achievable method in the embodiments of this application, the disturbance offset is deconstructed into frequency domain components, time domain statistics, and nonlinear dynamic indicators, including: calculating the power spectral density shift distribution of the disturbance offset in the frequency domain through short-time Fourier transform; extracting the phase lock value change, heart rate variability mutation, and waveform distortion index of the disturbance offset in the time domain; and calculating the approximate entropy increment and Lyapunov exponential shift of the disturbance offset at the nonlinear dynamic level.

[0008] According to one achievable method in the embodiments of this application, the step of establishing an implicit mapping relationship between host electrical response and viral adaptive evolution, and converting the perturbation feature sequence into a viral mutation pressure indicator, includes: converting the perturbation feature sequence into a viral mutation pressure indicator through adaptive weighted calculation of perturbation intensity and infection duration, wherein the weighted calculation dynamically adjusts the weights of perturbation value and time factor based on empirical coefficients.

[0009] According to one achievable method in an embodiment of this application, the temporal evolution model employs a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints, including: the two-way perturbation propagation mechanism includes forward propagation and backward propagation, the forward propagation uses historical perturbation sequences to predict future evolution trajectories, the backward propagation uses hypothetical future sequences to reconstruct historical perturbations, and the self-consistency constraint is achieved through the cross-entropy loss function of the reconstruction errors of the forward propagation and the backward propagation.

[0010] According to one achievable method in the embodiments of this application, the self-consistency constraint achieved by the cross-entropy loss function of the reconstruction error of the forward propagation and the backward propagation includes: calculating the cross-entropy value between the forward predicted trajectory and the backward reconstructed trajectory, and optimizing the propagation parameters by an iterative weight correction mechanism when the cross-entropy value exceeds a preset self-consistency threshold.

[0011] According to one achievable method in the embodiments of this application, the step of generating a virus strain differentiation prediction path based on the similarity of electrical signal perturbation includes: using the cosine similarity of the electrical signal perturbation trajectory as the distance between nodes to generate a tree-like differentiation structure, and automatically marking high-risk branches by perturbation entropy difference.

[0012] According to the second aspect, a virus evolution trend prediction system based on time-series electrical signals is provided. The system includes: a time-series electrical signal acquisition unit configured to acquire in real-time time-series bioelectrical signal data generated by a virus-infected host individual, wherein the signals are selected from at least one of electroencephalogram (EEG), electrocardiogram (ECG), or electromyogram (EMG), and the signals reflect cell membrane potential perturbations or ion channel dynamics; a perturbation feature sequence construction unit configured to perform frame-by-frame processing and feature deconstruction on the time-series data, extract electrophysiological perturbation patterns characterizing the infection process, and construct a virus evolution perturbation feature sequence; a pressure indicator acquisition unit configured to establish an implicit mapping relationship between host electrical response and viral adaptive evolution, and convert the perturbation feature sequence into a viral mutation pressure indicator; a time-series evolution model construction unit configured to construct a time-series evolution model based on the indicator, inferring the potential evolutionary trajectory of the virus within a future time window, wherein the time-series evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints; and a trend prediction result generation unit configured to generate a virus strain differentiation prediction path based on the similarity of electrical signal perturbations and output an evolution trend phylogenetic map.

[0013] According to a third aspect, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any one of the first aspects above.

[0014] According to a fourth aspect, an electronic device is provided, comprising: one or more processors; and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method described in any one of the first aspects.

[0015] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application achieves non-invasive, indirect prediction of viral evolution trends by real-time acquisition of host electroencephalograms (EEGs), electrocardiograms (ECGs), or electromyograms (EMGs). This method eliminates the need for direct viral genome sampling, avoiding the invasiveness and laboratory dependence of traditional methods, and significantly reducing detection costs and time. Through frame-segmentation and feature deconstruction, electrophysiological perturbation patterns are extracted, constructing viral evolution perturbation feature sequences, which are then implicitly mapped to mutational pressure indicators, further enhancing the personalization and accuracy of the prediction. A time-series evolution model employing a bidirectional perturbation propagation mechanism and imposing trajectory self-consistency constraints can dynamically infer the potential trajectory of the virus within future time windows, such as immune escape or replication enhancement paths, solving the lag problem of static prediction and improving the model's robustness and adaptability to high-uncertainty evolution. A phylogenetic diagram generated based on the similarity of electrical signal perturbations provides intuitive visualization output, facilitating clinical early warning and vaccine development applications.

[0016] Of course, any product implementing this application does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a system architecture diagram applicable to the embodiments of this application; Figure 2 A flowchart illustrating the virus evolution trend prediction method based on time-series electrical signals provided in this application embodiment; Figure 3 A structural block diagram of a virus evolution trend prediction system based on time-series electrical signals provided in an embodiment of this application; Figure 4 A schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0021] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0022] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0023] To facilitate understanding of this application, the system architecture on which this application is based will be described first. Figure 1 An exemplary system architecture that can be applied to embodiments of this application is shown, such as Figure 1 As shown, the system architecture may include: user equipment and a virus evolution trend prediction system based on time-series electrical signals located on the server side.

[0024] Users can input bioelectrical signal time-series data through user devices, which then send it to a server-side virus evolution trend prediction system based on time-series electrical signals. The virus evolution trend prediction system based on time-series electrical signals can use the method provided in the embodiments of this application to predict the evolution trend of the bioelectrical signal time-series data, obtaining an evolution trend phylogenetic map. The server can then send the evolution trend phylogenetic map to the user terminal for subsequent operations.

[0025] User devices can include, but are not limited to, smart mobile terminals, smart home devices, wearable devices, and PCs (Personal Computers). Smart mobile devices can include mobile phones, tablets, laptops, PDAs (Personal Digital Assistants), and connected cars. Smart home devices can include smart TVs, smart refrigerators, and so on. Wearable devices can include smartwatches, smart glasses, virtual reality devices, augmented reality devices, and mixed reality devices (i.e., devices that support both virtual and augmented reality).

[0026] The virus evolution trend prediction system based on time-series electrical signals can be configured as a standalone server, a server cluster, or a cloud server. A cloud server, also known as a cloud computing server or cloud host, is a hosting product within the cloud computing service system, designed to address the management difficulties and weak service scalability inherent in traditional physical hosts and Virtual Private Servers (VPS) services. Besides... Figure 1 In addition to the architecture shown, the virus evolution trend prediction system based on time-series electrical signals can also be set up on a computer terminal with strong computing power.

[0027] It should be understood that Figure 1 The user equipment and virus evolution trend prediction system based on time-series electrical signals shown are merely illustrative. Depending on implementation needs, any number of user equipment and virus evolution trend prediction systems based on time-series electrical signals can be included.

[0028] Figure 2 This document presents a flowchart of a method for predicting virus evolution trends based on time-series electrical signals, as provided in an embodiment of this application. Figure 2 As shown, the method may include the following steps: Step 201: Real-time acquisition of bioelectrical signal time-series data generated by the virus-infected host individual, wherein the signal is selected from at least one of electroencephalogram, electrocardiogram or electromyogram, and the signal reflects cell membrane potential perturbation or ion channel dynamics.

[0029] Step 202: Perform frame-by-frame processing and feature deconstruction on the time-series data, extract electrophysiological perturbation patterns that characterize the infection process, and construct a virus evolution perturbation feature sequence.

[0030] Step 203: Establish an implicit mapping relationship between host electrical response and viral adaptive evolution, and convert the perturbation feature sequence into a viral mutation pressure indicator.

[0031] Step 204: Construct a time-series evolution model based on the indicator factors to infer the potential evolution trajectory of the virus within future time windows. The time-series evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints.

[0032] Step 205: Based on the similarity of electrical signal perturbation, generate a predicted path for virus strain differentiation and output an evolutionary trend phylogenetic diagram.

[0033] As can be seen from the above process, this application achieves non-invasive, indirect prediction of viral evolution trends by real-time acquisition of host electroencephalograms, electrocardiograms, or electromyograms (EMGs). This method eliminates the need for direct viral genome sampling, avoiding the invasiveness and laboratory dependence of traditional methods, and significantly reducing detection costs and time. Through frame processing and feature deconstruction, electrophysiological perturbation patterns are extracted, constructing viral evolution perturbation feature sequences, which are then implicitly mapped into mutation pressure indicators, further enhancing the personalization and accuracy of the prediction. The time-series evolution model employing a bidirectional perturbation propagation mechanism and imposing trajectory self-consistency constraints can dynamically infer the potential trajectory of the virus within future time windows, such as immune escape or replication enhancement paths, solving the lag problem of static prediction and improving the model's robustness and adaptability to high-uncertainty evolution. The phylogenetic diagram generated based on the similarity of electrical signal perturbations provides intuitive visualization output, facilitating clinical early warning and vaccine development applications.

[0034] The following describes in detail each step of the above process and the effects that can be further produced, with reference to the embodiments. First, with reference to the embodiments, step 201 above, namely, "real-time acquisition of bioelectrical signal time-series data generated by the virus-infected host individual, wherein the signal is selected from at least one of electroencephalogram, electrocardiogram or electromyogram, and the signal reflects cell membrane potential perturbation or ion channel dynamics", will be described in detail.

[0035] Real-time acquisition of bioelectrical signal time-series data generated by a virus-infected host refers to the continuous and instantaneous acquisition of electrical signal sequences produced by the host body during the viral infection process. These signals form a continuous dynamic data stream along a time axis, capable of capturing real-time changes in the infection from its early to its progressive stages. This acquisition method is typically achieved using wearable devices or medical monitoring instruments to ensure uninterrupted data acquisition, thereby providing highly timely basic information for subsequent analysis.

[0036] The signals are selected from at least one of electroencephalography (EEG), electrocardiography (ECG), or electromyography (EMG), meaning that one or more common types of bioelectrical signals can be acquired. EEG records changes in the electrical potential of neurons in the brain and is often used to monitor the central nervous system's response to viral invasion. ECG captures waveforms of cardiac electrical activity, reflecting the electrophysiological state of cardiomyocytes, which may be abnormal due to viral infection. EMG records electrical signals from muscle fibers to observe potential fluctuations in muscle tissue, particularly prominent in viral-induced inflammation or neuromuscular damage. The selection of these signals depends on the specific type of infection and host symptoms, allowing for flexible application.

[0037] Specifically, the signals reflect cell membrane potential perturbations or ion channel dynamics. Cell membrane potential perturbations manifest as resting potential shifts, abnormal action potential morphology, or synchronicity disruptions, such as alpha wave suppression in electroencephalography (EEG) and QRS distortion in electrocardiography (ECG). Ion channel dynamics involve changes in the opening frequency or duration of channels such as sodium, potassium, and calcium. These microscopic changes are indirectly captured through macroscopic electrical signals. The core of this feature lies in transforming the molecular-level interference of viruses on host cells into measurable macroscopic electrophysiological characterization, constructing a bridge from microscopic viral behavior to macroscopic signal responses.

[0038] The following describes in detail step 202, namely, "performing frame-by-frame processing and feature deconstruction on the time-series data, extracting electrophysiological perturbation patterns characterizing the infection process, and constructing a virus evolution perturbation feature sequence," with reference to the embodiments.

[0039] First, framing is a structured segmentation operation of continuous time-series data. This framing design ensures the temporal locality of signal analysis, enabling the model to capture the dynamic changes at different stages of the virus infection process, while avoiding the computational complexity and noise accumulation caused by long sequences. After framing, each frame of the signal is treated as an independent analysis unit, preserving the local features and evolutionary trends of the original waveform.

[0040] Secondly, feature decomposition is a multi-level, systematic process of decomposing each frame of signal. This process is not simply feature extraction, but rather employs a multi-scale analysis method to decompose the signal into three dimensions: frequency domain components, time domain statistics, and nonlinear dynamic indicators. Frequency domain components reveal the distribution of signal energy across different frequency bands through short-time Fourier transform; time domain statistics focus on waveform morphology, amplitude, and periodic perturbations; and nonlinear dynamic indicators uncover the complexity, chaos, and unpredictability of the signal. This multi-scale decomposition can comprehensively characterize the electrophysiological abnormalities induced by viral infection, representing the evolution of the infection process from different perspectives.

[0041] As an implementable approach, the framing and feature deconstruction of the time-series data includes: using a sliding window algorithm to divide the time-series signal into frames, each frame being 5-30 minutes long; performing an individualized comparison of each frame with the corresponding baseline frame to calculate the perturbation offset; and performing multi-scale deconstruction of the perturbation offset, decomposing it into frequency domain components, time domain statistics, and nonlinear dynamic indices.

[0042] Specifically, a sliding window algorithm is used to frame the time-series signal, with each frame ranging from 5 to 30 minutes in length. This operation divides long-term bioelectrical signals into multiple short-term units by sliding a window with a fixed or variable step size along the time axis. The significance of framing lies in enabling localized analysis of the signal, allowing the model to capture transient electrophysiological responses at different stages of viral infection while controlling computational complexity. The choice of frame length range takes into account the time-scale characteristics of physiological events. For example, periodic changes in electrocardiograms can be observed within minutes, while slow wave evolution in electroencephalograms requires a longer window to ensure statistical stability. Alternatively, an initial frame length of 10 minutes can be set, with real-time monitoring of the rate of change in perturbation intensity. If the mutation rate is >20% / frame, the frame length is automatically shortened to 5 minutes; if it is <5%, it is extended to 20 minutes, ensuring a balance between computational efficiency and prediction accuracy, combining versatility and real-time performance.

[0043] Each frame of signal is individually compared with its corresponding baseline frame to calculate the perturbation offset. This comparison is not a simple subtraction of absolute values, but rather uses the baseline signal of the same host before infection as a dynamic reference, achieving precise calibration through frame-by-frame alignment. The baseline frame is typically taken from the average electrophysiological data of the same time period in a non-infected state, thus filtering out individual normal fluctuations, circadian rhythms, or environmental interference. The perturbation offset, as a quantitative indicator, characterizes the net electrophysiological abnormalities caused by viral infection, such as power spectrum energy deviation, waveform morphology distortion, or abrupt changes in complexity, providing a clean "perturbation signal" for subsequent analysis.

[0044] The perturbation offset is deconstructed at multiple scales, into frequency domain components, time domain statistics, and nonlinear dynamic indices. This deconstruction process employs a multi-level analytical framework: in the frequency domain, short-time Fourier transform reveals the energy offset distribution in specific frequency bands; in the time domain, statistical features such as amplitude, interval, and correlation are extracted to capture the dynamic evolution of the waveform; and at the nonlinear dynamic level, entropy indices or chaotic parameters are calculated to explore the response of signal complexity to the microscopic behavior of the virus. The synergistic representation of these three components overcomes the limitations of a single dimension, constructing a multi-dimensional perturbation image of the virus infection process.

[0045] Preferably, this application performs multi-scale decomposition of the disturbance offset into frequency domain components, time domain statistics, and nonlinear dynamic indices, including: calculating the power spectral density shift distribution of the disturbance offset in the frequency domain through short-time Fourier transform; extracting the phase lock value change, heart rate variability mutation, and waveform distortion index of the disturbance offset in the time domain; and calculating the approximate entropy increment and Lyapunov exponential shift of the disturbance offset at the nonlinear dynamic level.

[0046] Specifically, at the frequency domain level, the power spectral density shift distribution of the perturbation offset is calculated using a short-time Fourier transform. This operation first applies a short-time window function to the perturbation offset of each frame to achieve time-frequency localization analysis, and then uses a fast Fourier transform to obtain the energy distribution changes in different frequency bands. The power spectral density shift distribution specifically refers to the spectral energy deviation relative to the baseline after infection, such as delta-band enhancement or theta-band suppression. This frequency domain deconstruction can capture the periodic electrical activity reconstruction caused by viral interference with ion channels or neural network synchronization, and is an important macroscopic fingerprint reflecting cellular-level metabolic disorders.

[0047] At the time domain level, changes in phase lock-in values, heart rate variability mutations, and waveform distortion indices are extracted from the perturbation offset. Phase lock-in value changes quantify the synchronicity decay between neuronal populations or cardiomyocytes, reflecting virus-induced neural or cardiac conduction disorders; heart rate variability mutations characterize autonomic nervous system regulatory imbalances through time-domain statistical indicators such as standard deviation or root mean square difference; and waveform distortion indices assess structural damage to action potentials or QRS complexes through area ratio, peak offset, or duration anomalies. These time-domain features collectively characterize the dynamic perturbation patterns of viral infection on millisecond to second timescales.

[0048] At the nonlinear dynamics level, the approximate entropy increment and Lyapunov exponential shift of the perturbation offset are calculated. The approximate entropy increment measures the increase in signal complexity and unpredictability, reflecting the transformation of cellular electrical activity from order to chaos caused by viral replication; the Lyapunov exponential shift characterizes the increased sensitivity of the system to initial conditions, indicating that the infection process has entered a nonlinear amplification stage. These nonlinear indices break through the limitations of traditional linear analysis and, for the first time, link the dynamic chaotic behavior of the host electrophysiological system with viral adaptive evolution.

[0049] Based on deconstruction, the core innovation of this method lies in extracting electrophysiological perturbation patterns that characterize the infection process. These patterns specifically refer to the unique electrophysiological responses induced by viral invasion of host cells, such as abnormal waveforms caused by cell membrane depolarization, ion channel dysfunction, or disruption of neuronal synchronization. These perturbation patterns are not random noise, but rather dynamic fingerprints with biological significance, indirectly reflecting microscopic processes such as viral replication, immune escape, or enhanced pathogenicity. By quantitatively extracting these patterns, the method achieves a semantic-level mapping from macroscopic host electrical signals to microscopic viral behavior.

[0050] Finally, the extracted perturbation patterns are reconstructed into a virus evolution perturbation feature sequence. This sequence is a high-dimensional vector stream that evolves over time. The vector at each time step is fused from the results of multi-scale deconstruction, including components such as frequency domain energy shift, temporal morphological exponent, and nonlinear complexity increment. The reconstruction process employs a weighted fusion mechanism to ensure the collaborative representational ability of features from different dimensions. This feature sequence serves as a proxy representation of virus evolution, directly inputting into subsequent implicit mappings and temporal evolution models to drive accurate inferences about future mutation trends.

[0051] The following describes in detail step 203, namely, "establishing an implicit mapping relationship between host electrical response and viral adaptive evolution, and converting the perturbation feature sequence into a viral mutation pressure indicator," with reference to the embodiments.

[0052] This step achieves a cross-scale semantic bridge from macroscopic electrophysiological perturbations of the host to microscopic evolutionary pressures of the virus. It can quantify the adaptive mutation tendency without directly sequencing the viral genome, providing key driving variables for subsequent temporal evolution prediction.

[0053] The implicit mapping relationship is not established through explicit biochemical equations, but rather through a data-driven functional form that correlates the multidimensional perturbation patterns of the host's electrical response with the selective pressures of the virus in the host environment. The host's electrical response manifests as a perturbation feature sequence after multi-scale deconstruction, including components such as frequency domain energy shift, temporal domain morphological anomalies, and nonlinear complexity increments. These components collectively characterize the physiological cascade effects induced by viral invasion, such as cell membrane potential disturbances, ion channel dysfunction, or neural network reconstruction. Viral adaptive evolution refers to the direction of mutations at the gene level that occur to evade immunity, increase replication efficiency, or spread across hosts.

[0054] The process of transforming perturbation feature sequences into indicators of viral mutation pressure employs an adaptive weighting mechanism. Specifically, the intensity of the perturbation feature sequences over time and the duration of infection are dynamically weighted to generate a pressure factor that evolves over time. The intensity of the perturbation reflects the degree of electrophysiological abnormalities in the current stage of infection; for example, a sudden increase in the high-frequency power spectrum corresponds to the acute inflammatory phase. The duration of infection introduces the cumulative effect of the game between the virus and the host; for example, prolonged low-intensity perturbation may predict chronic adaptation. The weighting coefficients are adaptively adjusted based on experience or model feedback to ensure the generalization ability of the mapping relationship across different virus types and individual hosts.

[0055] Specifically, viral mutation stress indicators Calculated using the following implicit mapping relationship: in, The input is a perturbation feature sequence after normalization. Its L2 norm represents the current perturbation strength; Duration of infection; This is the first derivative of the disturbance intensity, reflecting the evolving pressure acceleration; 、 、 The initial values ​​for the adaptive weight coefficients are set to 0.6, 0.3, and 0.1, respectively, and are dynamically optimized through gradient descent during model training.

[0056] The following describes in detail step 204, namely, "constructing a time-series evolution model based on indicator factors to infer the potential evolution trajectory of the virus within a future time window, wherein the time-series evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints," with reference to the embodiments.

[0057] The temporal evolution model is constructed using a mutational pressure indicator as its core input. This indicator, the output of the aforementioned implicit mapping step, characterizes the comprehensive selective pressures faced by the virus in the host microenvironment, including factors such as immune clearance, replication restriction, and environmental adaptation. The model treats the indicator sequence as a driving signal of temporal evolution, deducing its changing trends within future time windows through recursive or iterative mechanisms. These future time windows are typically set to several hours to several days, covering key stages of the viral replication cycle to ensure the predictions are timely for clinical intervention.

[0058] The temporal evolution model in this application does not rely on a specific named neural network, but is implemented based on a custom bidirectional perturbation propagation mechanism. Its core components are recursive perturbation dynamics equations and trajectory self-consistency constraints. The bidirectional perturbation propagation mechanism is a unique structural design of the model. It includes two complementary processes: forward propagation and backward propagation. Forward propagation starts with the historical perturbation sequence and infers the future trajectory forward in time, simulating the natural evolution path of the virus under pressure. Backward propagation, on the other hand, takes the hypothetical future trajectory as input and reconstructs the historical perturbation in reverse, verifying the physical consistency of the inference logic. This bidirectional mechanism draws on the temporal symmetry of physical systems, overcoming the shortcomings of unidirectional temporal models that are prone to getting trapped in local optima or overfitting, and significantly improving the robustness and global convergence of predictions.

[0059] Trajectory self-consistency constraints serve as an optimization guarantee for bidirectional propagation. By calculating the reconstruction error between the forward predicted trajectory and the backward reconstructed trajectory, and quantifying the inconsistency using the cross-entropy loss function, the model enforces self-consistency requirements during training or inference. When the error exceeds a preset threshold, weight correction or trajectory adjustment is triggered to ensure that the final output evolution trajectory is logically closed in both forward and backward propagation. This constraint mechanism not only enhances the model's robustness against noise and uncertainty but also gives the prediction results higher interpretability; for example, high self-consistency trajectories correspond to low-risk evolution, while low self-consistency trajectories indicate potential mutation branches.

[0060] As an feasible approach, achieving self-consistency constraints through the cross-entropy loss function of the reconstruction error between the two involves: calculating the cross-entropy value between the forward predicted trajectory and the backward reconstructed trajectory; and optimizing the propagation parameters through an iterative weight correction mechanism when the cross-entropy value exceeds a preset self-consistency threshold.

[0061] Cross-entropy quantifies this distributional inconsistency; the smaller the value, the more closed the trajectory logic, and the more consistent it is with physical and biological evolutionary laws. When the cross-entropy value exceeds a preset self-consistency threshold, it indicates a significant logical conflict in the current trajectory, requiring an iterative weight correction mechanism for optimization. This threshold is preset based on the infection stage, signal quality, or clinical needs; for example, it is set to 0.1 during the acute infection phase to ensure high sensitivity. Iterative correction adjusts the propagation weights within the model using gradient descent or Bayesian updates, such as the contribution coefficients of forward and backward paths or perturbation sensitivity factors, so that the trajectory pairs generated by the next bidirectional propagation gradually converge to consistency. This process is similar to a closed-loop control system, continuously calibrating until the cross-entropy converges below the threshold.

[0062] The following describes in detail step 205, namely, "generating a virus strain differentiation prediction path based on the similarity of electrical signal perturbation and outputting an evolutionary trend phylogenetic chart," with reference to the embodiments.

[0063] Using the similarity of electrical signal perturbations as the basis for branching is the core criterion for generating predicted paths. Perturbation similarity is determined by quantifying the consistency of electrophysiological perturbation patterns between different time points or hypothetical trajectories. For example, cosine similarity or dynamic time-warped distance can be used to measure the vector proximity of perturbation feature sequences. When the similarity is higher than a preset threshold, it is considered a continuation of the same evolutionary path; below the threshold, branching is triggered, representing significant differentiation of viral strains under pressure. The biological significance of this criterion is that similar perturbation patterns correspond to stable adaptation of the virus in the host microenvironment, while mutational perturbations predict the emergence of new strains such as immune escape, enhanced replication, or cross-host transmission.

[0064] The process of generating viral strain differentiation prediction paths employs a tree-structured model. Starting from the current perturbation trajectory as the root node, it expands along multiple potential trajectories output by the temporal evolution model, with each trajectory corresponding to a possible differentiation direction. Path generation can combine Monte Carlo sampling or clustering algorithms to ensure coverage of high-probability evolution scenarios, while also labeling the mutational pressure increment and risk level of each path segment. For example, one path shows a continuous decrease in perturbation similarity, indicating that the virus is evolving towards a direction with low immune recognition; another path shows a sharp increase in perturbation entropy, suggesting a risk of highly pathogenic variants.

[0065] As an feasible approach, generating a virus strain differentiation prediction path based on the similarity of electrical signal perturbation includes: using the cosine similarity of the electrical signal perturbation trajectory as the distance between nodes to generate a tree-like differentiation structure, and automatically marking high-risk branches by perturbation entropy difference.

[0066] Perturbation trajectories are continuous temporal representations of feature sequences deconstructed at multiple scales, with each trajectory corresponding to a possible viral evolutionary path. Cosine similarity is calculated by taking the cosine of the angle between two trajectory vectors, ranging from -1 to 1. Values ​​closer to 1 indicate more consistent perturbation patterns. This distance definition ensures that branching is based on the intrinsic similarity of the host's electrophysiological response; for example, similar trajectories reflect stable viral replication under the same stress, while dissimilar trajectories indicate electrophysiological remodeling triggered by mutation events.

[0067] The process of generating the tree-like differentiation structure employs hierarchical clustering or a minimum spanning tree algorithm. Starting from the current perturbation trajectory as the root node, and using the reciprocal of the cosine similarity as the edge weight, child nodes are gradually expanded to form a tree topology. Each node represents the state of the virus strain within a specific time window, and edge connections represent evolutionary continuity. When the similarity falls below a preset threshold, new branches are automatically generated, resulting in a multi-path differentiation structure. This tree-like form intuitively simulates the phylogenetic evolution of the virus population, covering the complete scenario from low-probability drift to high-probability escape.

[0068] Automatically marking high-risk branches using perturbation entropy difference is an innovative risk assessment feature. Perturbation entropy difference calculates the increment of approximate entropy or Shannon entropy between adjacent nodes; a larger entropy difference indicates a sharp increase in trajectory complexity, corresponding to the virus entering a chaotic adaptation phase. The system sets a threshold, such as 0.15 bits; branches exceeding this threshold are automatically marked as high-risk, for example, highlighted in red with a note indicating potential mutation types. This marking mechanism requires no manual intervention, achieving intelligent early warning of evolutionary paths.

[0069] The output evolutionary trend phylogenetic diagram is a visual representation of this feature. The phylogenetic diagram is presented as a tree structure or a directed acyclic graph, where nodes represent the viral strain state at a specific time point, edge lengths or colors encode perturbation similarity, and branch points are labeled with differentiation probabilities and key mutation types. An automatic perturbation entropy difference labeling mechanism is integrated into the diagram, with high-risk branches highlighted in red, facilitating quick identification of evolutionary paths requiring priority intervention. This phylogenetic diagram not only visually displays the dynamic differentiation process of the virus from the current strain to future variants but also supports interactive exploration, such as zooming to view the details of electrical signal perturbations in specific branches.

[0070] The methods provided in this application can be applied to various scenarios, including but not limited to: for example, in hospital intensive care units, wearable devices can be used to collect patients' electroencephalogram (EEG) and electrocardiogram (ECG) signals in real time to predict the immune escape evolution path of influenza or COVID-19, identify high-risk variants 48 hours in advance, and guide personalized antiviral treatment adjustments. Another scenario is public health epidemic monitoring, where multi-host signal aggregation systems can be deployed in communities or airports to anonymize population electrophysiological perturbation patterns, construct population-level viral drift phylogenetic maps, detect cross-regional transmission risks early, and support precise lockdown and vaccine iteration. A third scenario is vaccine research laboratories, where time-series electromyography data from animal models can be used to simulate human infection, rapidly screen candidate vaccines for coverage efficacy against potential evolutionary branches, and accelerate the translation cycle from laboratory to clinical practice.

[0071] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0072] According to another embodiment, a virus evolution trend prediction system based on time-series electrical signals is provided. For example... Figure 3 As shown, the system 300 includes: The timing electrical signal acquisition unit 301 is configured to acquire timing data of bioelectrical signals generated by a virus-infected host individual in real time. The signals are selected from at least one of electroencephalogram (EEG), electrocardiogram (ECG), or electromyogram (EMG), and the signals reflect cell membrane potential disturbances or ion channel dynamics.

[0073] The perturbation feature sequence construction unit 302 is configured to perform frame-by-frame processing and feature deconstruction on the time-series data, extract electrophysiological perturbation patterns that characterize the infection process, and construct a virus evolution perturbation feature sequence.

[0074] The stress indicator acquisition unit 303 is configured to establish an implicit mapping relationship between the host electrical response and the adaptive evolution of the virus, and to convert the perturbation feature sequence into a virus mutation stress indicator.

[0075] The temporal evolution model construction unit 304 is configured to construct a temporal evolution model based on the indicator factors, and infer the potential evolution trajectory of the virus within a future time window. The temporal evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints.

[0076] The trend prediction result generation unit 305 is configured to generate a virus strain differentiation prediction path based on the similarity of electrical signal perturbation and output an evolution trend phylogenetic diagram.

[0077] As an implementable approach, the perturbation feature sequence construction unit 302 can be configured to: divide the time-series data into frames using a sliding window algorithm, with each frame being 5 to 30 minutes long; perform individualized comparisons between each frame signal and the corresponding baseline frame to calculate the perturbation offset; and perform multi-scale decomposition of the perturbation offset into frequency domain components, time domain statistics, and nonlinear dynamic indices.

[0078] As an implementable approach, the perturbation feature sequence construction unit 302, when performing multi-scale decomposition of the perturbation offset into frequency domain components, time domain statistics, and nonlinear dynamic indices, can be configured to: calculate the power spectral density shift distribution of the perturbation offset in the frequency domain using short-time Fourier transform; extract the phase lock value change, heart rate variability mutation, and waveform distortion index of the perturbation offset in the time domain; and calculate the approximate entropy increment and Lyapunov exponential shift of the perturbation offset at the nonlinear dynamic level.

[0079] As an implementable approach, the pressure indicator acquisition unit 303, when establishing an implicit mapping relationship between host electrical response and viral adaptive evolution and converting the perturbation feature sequence into a viral mutation pressure indicator, can be configured to: convert the perturbation feature sequence into a viral mutation pressure indicator through adaptive weighted calculation of perturbation intensity and infection duration, wherein the weighted calculation dynamically adjusts the weights of perturbation value and time factor based on empirical coefficients.

[0080] As an implementable approach, the temporal evolution model in the temporal evolution model construction unit 304 adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints. The two-way perturbation propagation mechanism includes forward propagation and backward propagation. The forward propagation uses historical perturbation sequences to predict future evolution trajectories, and the backward propagation uses hypothetical future sequences to reconstruct historical perturbations. Self-consistency constraints are achieved through the cross-entropy loss function of the reconstruction errors of the forward propagation and the backward propagation.

[0081] As an implementable approach, the temporal evolution model building unit 304, when implementing self-consistency constraints through the cross-entropy loss function of the reconstruction errors of the forward propagation and the backward propagation, can be configured to: calculate the cross-entropy value between the forward predicted trajectory and the backward reconstructed trajectory, and optimize the propagation parameters through an iterative weight correction mechanism when the cross-entropy value exceeds a preset self-consistency threshold.

[0082] As an feasible approach, the trend prediction result generation unit 305 can be configured to generate a virus strain differentiation prediction path based on the similarity of electrical signal perturbation as the branching basis, and generate a tree-like differentiation structure by using the cosine similarity of the electrical signal perturbation trajectory as the distance between nodes, and automatically mark high-risk branches by perturbation entropy difference.

[0083] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. Components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0084] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0085] In addition, embodiments of this application also provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method described in any of the foregoing method embodiments.

[0086] And an electronic device comprising: one or more processors; and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method described in any of the foregoing method embodiments.

[0087] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the foregoing method embodiments.

[0088] in, Figure 4 An exemplary architecture of an electronic device is shown, which may include a processor 410, a video display adapter 411, a disk drive 412, an input / output interface 413, a network interface 414, and a memory 420. The processor 410, video display adapter 411, disk drive 412, input / output interface 413, network interface 414, and memory 420 can communicate with each other via a communication bus 430.

[0089] The processor 410 can be implemented using a general-purpose CPU, microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits to execute relevant programs and implement the technical solution provided in this application.

[0090] The memory 420 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 420 can store the operating system 421 for controlling the operation of the electronic device 400, and the basic input / output system (BIOS) 422 for controlling the low-level operations of the electronic device 400. Additionally, it can store a web browser 423, a data storage management system 424, and a virus evolution trend prediction system 425 based on timing electrical signals, etc. The aforementioned virus evolution trend prediction system 425 based on timing electrical signals can be the application program that specifically implements the aforementioned steps in this embodiment. In summary, when implementing the technical solution provided in this application through software or firmware, the relevant program code is stored in the memory 420 and executed by the processor 410.

[0091] Input / output interface 413 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0092] Network interface 414 is used to connect a communication module (not shown in the figure) to enable communication and interaction between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0093] Bus 430 includes a pathway for transmitting information between various components of the device, such as processor 410, video display adapter 411, disk drive 412, input / output interface 413, network interface 414, and memory 420.

[0094] It should be noted that although the above-described device only shows the processor 410, video display adapter 411, disk drive 412, input / output interface 413, network interface 414, memory 420, bus 430, etc., in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the solution of this application, and does not necessarily include all the components shown in the figures.

[0095] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer program product. This computer program product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0096] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for predicting virus evolution trends based on time-series electrical signals, characterized in that, The method includes: Real-time acquisition of bioelectrical signal time-series data generated by virus-infected host individuals, wherein the signals are selected from at least one of electroencephalogram (EEG), electrocardiogram (ECG), or electromyogram (EMG), and the signals reflect cell membrane potential perturbations or ion channel dynamics; The time-series data is processed by frame segmentation and feature deconstruction to extract electrophysiological perturbation patterns that characterize the infection process and construct a virus evolution perturbation feature sequence. An implicit mapping relationship between host electrical response and viral adaptive evolution is established, and the perturbation feature sequence is transformed into an indicator of viral mutation pressure. A time-series evolution model is constructed based on the aforementioned indicator factors to infer the potential evolution trajectory of the virus within future time windows. The time-series evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints. Based on the similarity of electrical signal perturbation, a virus strain differentiation prediction path is generated, and an evolutionary trend phylogenetic diagram is output.

2. The method according to claim 1, characterized in that, The step of performing frame segmentation and feature deconstruction on the time-series data includes: The time-series data is divided into frames using a sliding window algorithm, with each frame ranging from 5 to 30 minutes in length. Each frame of signal is individually compared with the corresponding baseline frame, and the disturbance offset is calculated. The disturbance offset is decomposed into frequency domain components, time domain statistics, and nonlinear dynamic indices through multi-scale decomposition.

3. The method according to claim 2, characterized in that, The multi-scale decomposition of the disturbance offset into frequency domain components, time domain statistics, and nonlinear dynamic indices includes: The power spectral density shift distribution of the disturbance offset is calculated in the frequency domain using a short-time Fourier transform. The phase lock value change, heart rate variability mutation, and waveform distortion index of the disturbance offset are extracted in the time domain. The approximate entropy increment and Lyapunov exponential shift of the disturbance offset are calculated at the nonlinear dynamic level.

4. The method according to claim 1, characterized in that, The process of establishing an implicit mapping relationship between host electrical response and viral adaptive evolution, and converting the perturbation feature sequence into an indicator of viral mutation pressure, includes: The perturbation feature sequence is transformed into a viral mutation pressure indicator by adaptive weighted calculation of perturbation intensity and infection duration. The weighted calculation dynamically adjusts the weights of perturbation value and time factor based on empirical coefficients.

5. The method according to claim 1, characterized in that, The time-series evolution model employs a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints, including: The bidirectional perturbation propagation mechanism includes forward propagation and backward propagation. The forward propagation uses historical perturbation sequences to predict future evolution trajectories, and the backward propagation uses hypothetical future sequences to reconstruct historical perturbations. Self-consistency constraints are achieved through the cross-entropy loss function of the reconstruction errors of the forward and backward propagations.

6. The method according to claim 5, characterized in that, The self-consistency constraint achieved through the cross-entropy loss function of the reconstruction errors from the forward and backward propagation includes: The cross-entropy value between the forward predicted trajectory and the backward reconstructed trajectory is calculated. When the cross-entropy value exceeds a preset self-consistent threshold, the propagation parameters are optimized through an iterative weight correction mechanism.

7. The method according to claim 1, characterized in that, The method of generating a virus strain differentiation prediction path based on the similarity of electrical signal perturbations includes: Using the cosine similarity of the electrical signal perturbation trajectory as the distance between nodes, a tree-like differentiation structure is generated, and high-risk branches are automatically marked by the perturbation entropy difference.

8. A virus evolution trend prediction system based on time-series electrical signals, characterized in that, The system includes: The time-series electrical signal acquisition unit is configured to acquire time-series data of bioelectrical signals generated by a virus-infected host individual in real time. The signals are selected from at least one of electroencephalogram (EEG), electrocardiogram (ECG), or electromyogram (EMG), and the signals reflect cell membrane potential perturbations or ion channel dynamics. The perturbation feature sequence construction unit is configured to perform frame-by-frame processing and feature deconstruction on the time-series data, extract electrophysiological perturbation patterns characterizing the infection process, and construct a virus evolution perturbation feature sequence. The stress indicator acquisition unit is configured to establish an implicit mapping relationship between host electrical response and viral adaptive evolution, and to convert the perturbation feature sequence into a viral mutation stress indicator. The temporal evolution model construction unit is configured to construct a temporal evolution model based on the indicator factors, and infer the potential evolution trajectory of the virus within a future time window. The temporal evolution model adopts a two-way perturbation propagation mechanism and applies trajectory self-consistency constraints. The trend prediction result generation unit is configured to generate a virus strain differentiation prediction path based on the similarity of electrical signal perturbations, and output an evolutionary trend phylogenetic diagram.

9. An electronic device, characterized in that, include: One or more processors; and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 7.

Images