Real-time monitoring method and device for respiratory tract infection state and electronic equipment

Abstract

The invention provides a real-time monitoring method and device for a respiratory tract infection state and electronic equipment. The method comprises the following steps: fusing demographic data and sign data of a user, performing adaptive baseline compensation and calibration through body position jump Mask coding and body position coding, performing Hampel filtering on data after baseline calibration, and processing the data; for the processed data, three types of complementary features are extracted by adopting a body position dependent wavelet multi-scale abnormal energy algorithm, a physiological coupling anomaly detection algorithm and night stable section hidden state drift detection; and finally, judging the respiratory tract infection state through time sequence normalization, confidence evaluation and gating weighted fusion, and fusing a result into a subsequent detection process to realize closed-loop baseline adaptive updating, so that the accuracy of respiratory tract infection state detection is improved.

Description

Technical Field

[0001] This invention relates to the field of medical technology, and in particular to a method, apparatus, and electronic device for real-time monitoring of respiratory infection status. Background Technology

[0002] Currently, the issue of "delayed assessment of respiratory infection status" is particularly prominent in children at night, and also frequently occurs in nighttime fever monitoring of adults and the elderly. In common fever-related respiratory infection risk monitoring, the main drawbacks of traditional testing methods are: 1. The monitoring methods are singular and discontinuous. The first and almost only indicator that parents and supervisors pay continuous attention to is "body temperature". Moreover, the monitoring is mainly sporadic and does not involve real-time monitoring or effective combination with other physiological characteristics for comprehensive judgment.

[0003] 2. Using a single fever threshold without considering individual and postural differences, and failing to perform individual baseline adaptive adjustment based on postural and diurnal cycles, leads to misjudgment and low accuracy.

[0004] 3. The lack of multi-timescale time series analysis makes it difficult to identify slow-progressing respiratory infection patterns, resulting in respiratory infections in individuals who are already experiencing fever being frequently and systematically not diagnosed in a timely manner. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method, device and electronic device for real-time monitoring of respiratory infection status, so as to perform adaptive detection of respiratory infection status and improve the accuracy of respiratory infection status detection.

[0006] In a first aspect, embodiments of the present invention provide a method for real-time monitoring of respiratory infection status, the method comprising: collecting multi-source data from users; wherein the multi-source data includes: demographic data and vital sign data; constructing positional coding and jump mask coding based on the vital sign data; performing adaptive baseline compensation and calibration on the vital sign data based on the positional coding and jump mask coding; performing Hampshire filtering on the vital sign data after adaptive baseline compensation and calibration; extracting respiratory infection status features based on the multi-source data; performing time-series normalization, confidence assessment, and gated weighted fusion on the respiratory infection status features; and determining the respiratory infection status and updating the closed-loop baseline based on the respiratory infection status features.

[0007] In an optional embodiment of this application, the above-mentioned step of collecting multi-source data of users includes: acquiring the user's age data, weight data, and height data as demographic data; and collecting the user's body temperature data, body position data, respiratory apparent diffusion coefficient signal, and heart rate time series data as vital sign data through a wearable device.

[0008] In optional embodiments of this application, the steps of constructing position coding and transition mask coding based on vital sign data include: discretizing the night position sequence in the position data into a finite set of states to form position coding; constructing a position transition mask in the position change interval of the position data; wherein the transition mask is used to identify the region of abrupt change in the statistical characteristics of the signal; and performing transition intensity coding based on the position transition mask to construct transition mask coding.

[0009] In optional embodiments of this application, the steps of adaptive baseline compensation and calibration of vital sign data based on position coding and jump mask coding include: fitting the initial baseline of vital sign data based on position coding and jump mask coding, and calibrating it by means of steady-state mean or variance; resetting the baseline of vital sign data with jump neighborhood baseline under mask constraints; and performing time-series dynamic drift tracking and adaptive baseline update on the baseline of vital sign data.

[0010] In an optional embodiment of this application, the step of performing Hampshire filtering on the vital signs data after adaptive baseline compensation and calibration includes: performing Hampshire filtering on the vital signs data after the baseline of the mask-constrained jump neighborhood is reset.

[0011] In an optional embodiment of this application, the above-mentioned step of extracting respiratory infection status features based on multi-source data includes: performing position-dependent wavelet multi-scale abnormal energy detection, physiological coupling abnormality detection, and nighttime stable segment latent state drift detection based on multi-source data to extract respiratory infection status features.

[0012] In optional embodiments of this application, the steps of performing temporal normalization, confidence assessment, and gated weighted fusion on respiratory infection status features include: normalizing the respiratory infection status features on a sliding window to obtain a normalized score; calculating the confidence weight of each normalized score; and fusing the respiratory infection status features and confidence weights using a gated weighted method.

[0013] In an optional embodiment of this application, the steps of determining respiratory infection status and updating closed-loop baseline based on respiratory infection status characteristics include: determining respiratory infection status based on adaptive thresholds and determining respiratory infection status at multiple scales based on respiratory infection status characteristics; and updating the baseline based on historical data and the results of respiratory infection status determination.

[0014] Secondly, embodiments of the present invention also provide a real-time monitoring device for respiratory infection status. The device includes: a multi-source data acquisition module for acquiring multi-source data from users; wherein the multi-source data includes: demographic data and vital sign data; a position coding and jump mask coding construction module for constructing position coding and jump mask coding based on vital sign data; an adaptive baseline compensation and calibration module for performing adaptive baseline compensation and calibration on vital sign data based on position coding and jump mask coding; a Hampshire filtering module for performing Hampshire filtering on the vital sign data after adaptive baseline compensation and calibration; a respiratory infection status feature extraction module for extracting respiratory infection status features based on multi-source data; a time-series normalization, confidence assessment, and gated weighted fusion module for performing time-series normalization, confidence assessment, and gated weighted fusion on the respiratory infection status features; and a respiratory infection status judgment and closed-loop baseline update module for judging respiratory infection status and updating closed-loop baseline based on respiratory infection status features.

[0015] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the above-described method for real-time monitoring of respiratory infection status.

[0016] The embodiments of the present invention bring the following beneficial effects: This invention provides a method, device, and electronic device for real-time monitoring of respiratory infection status. It integrates user demographic data (age, weight, height) and vital sign data (time-series data of body temperature, body position, ADC (Apparent Diffusion Coefficient), respiration, and heart rate collected by a multi-parameter respiratory infection monitor). Adaptive baseline compensation and calibration are performed through body position jump mask encoding and body position encoding. The baseline-calibrated data is processed using Hampel filtering. The processed data is then used to extract three types of complementary features using a position-dependent wavelet multi-scale anomaly energy algorithm, a physiological coupling anomaly detection algorithm, and nighttime stable segment hidden state drift detection. Finally, the respiratory infection status is determined through time-series normalization, confidence assessment, and gated weighted fusion, and the results are incorporated into subsequent detection processes to achieve closed-loop adaptive baseline updates, thereby improving the accuracy of respiratory infection status detection.

[0017] Other features and advantages of this disclosure will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above.

[0018] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a method for real-time monitoring of respiratory infection status provided in an embodiment of the present invention; Figure 2 A schematic diagram of another method for real-time monitoring of respiratory infection status provided in an embodiment of the present invention; Figure 3 A schematic diagram of the structure of a real-time monitoring device for respiratory infection status provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Currently, in the risk monitoring of common fever-related respiratory infections, traditional detection methods have drawbacks such as being limited to a single and discontinuous monitoring approach, using a single fever threshold, and lacking time-series analysis across multiple time scales.

[0023] Based on this, the present invention provides a method, device and electronic device for real-time monitoring of respiratory infection status, specifically a method for real-time monitoring of respiratory infection status based on multi-source time-series adaptive monitoring, which can improve the accuracy of respiratory infection status detection.

[0024] To facilitate understanding of this embodiment, a method for real-time monitoring of respiratory infection status disclosed in this embodiment of the invention will first be described in detail.

[0025] Example 1: This invention provides a method for real-time monitoring of respiratory infection status, see [link to relevant documentation]. Figure 1 The flowchart shown illustrates a method for real-time monitoring of respiratory infection status, which includes the following steps: Step S102: Collect multi-source data from users; wherein, multi-source data includes: demographic data and vital signs data.

[0026] In this embodiment, oral data (age, weight, height) and vital signs data (body temperature, heart rate, respiratory ADC signal and body position time sequence data) of the user can be collected.

[0027] In some embodiments, user age, weight, and height data can be acquired as demographic data; and user body temperature, body position, respiratory apparent diffusion coefficient signal, and heart rate time series data can be collected through wearable devices as vital signs data.

[0028] In this embodiment, wearable devices (e.g., multi-parameter respiratory infection monitors) can be used to collect the user's body temperature data, body position data, respiratory ADC signals, and heart rate time series data in real time as vital sign data.

[0029] Step S104: Construct position coding and transition mask coding based on vital sign data.

[0030] In this embodiment, body position coding and transition mask coding can be constructed to perform body position-related feature evolution and combine with other physiological data for corresponding data processing.

[0031] In some embodiments, the nighttime body position sequence in the body position data can be discretized into a finite set of states to form a body position code; a body position transition mask can be constructed in the body position change range of the body position data; wherein, the transition mask is used to identify the region of abrupt change in the statistical characteristics of the signal; and transition intensity encoding is performed based on the body position transition mask to construct a transition mask code.

[0032] In this embodiment, the entire night's body position sequence can be discretized into a finite set of states to form a body position code; a body position transition mask is constructed in the body position change interval to identify regions where the signal statistical characteristics change rapidly; transition intensity encoding is performed, and users can statistically measure the instability by analyzing the different body position transition patterns and whether there are continuous transitions.

[0033] Step S106: Adaptive baseline compensation and calibration of vital sign data based on position coding and jump mask coding.

[0034] In this embodiment, adaptive baseline compensation and calibration can be performed. Based on body position and body position change mask encoding, baseline offset of data such as body temperature, ADC respiration and heart rate is calibrated. At the same time, age and body position are combined to establish an adaptive individual baseline.

[0035] In some embodiments, the initial baseline of the vital signs data can be fitted based on position encoding and jump mask encoding, and calibrated by means of steady-state segment mean or variance; the baseline of the vital signs data is reset by jump neighborhood with mask constraint; and the baseline of the vital signs data is subjected to time-series dynamic drift tracking and adaptive baseline update.

[0036] In this embodiment, initial baseline fitting can be performed and calibrated using the mean and variance of the steady-state segment; baseline reset of the jump neighborhood under mask constraints can be performed to eliminate baseline drift of physiological signals collected by the multi-parameter respiratory infection monitor due to jumps in different body positions; and time-series dynamic drift tracking and adaptive baseline update can be performed.

[0037] Step S108: Perform Hampshire filtering on the vital signs data after adaptive baseline compensation and calibration.

[0038] In this embodiment, Hampel filtering can be used to smooth noisy data.

[0039] In some embodiments, the trait data after the baseline of the mask-constrained transition neighborhood can be subjected to Hampshire filtering.

[0040] In this embodiment, for the input time-series vital signs data, the baseline of the transition neighborhood can be reset first by mask constraint, and then the outlier can be processed by Hampel filtering to achieve noise reduction.

[0041] Step S110: Extract respiratory infection status features based on multi-source data.

[0042] In this embodiment, multi-source data can be fused to extract respiratory infection status features from different dimensions and scales.

[0043] In some embodiments, position-dependent wavelet multi-scale anomaly energy detection, physiological coupling anomaly detection, and nighttime stable segment hidden state drift detection can be performed based on multi-source data to extract respiratory infection status features.

[0044] The extraction of respiratory infection status features in this embodiment may include: (1) Position-dependent wavelet multi-scale anomaly energy, using multi-scale wavelet transform to calculate energy changes in different frequency bands, and mapping anomaly point locations through position coding constraints; (2) Detection of physiological coupling anomalies: Construct a coupling model of body temperature, respiration, heart rate and body position, and detect coupling instability events through Pearson correlation, entropy divergence or mutual information.

[0045] (3) Nighttime stable segment hidden state drift detection: The sliding window hidden Markov model (HMM) is used to detect the probability mutation of the stable segment, so as to realize the identification of the slow inflammatory state evolution.

[0046] Step S112 involves performing temporal normalization, confidence assessment, and gated weighted fusion on the respiratory infection status characteristics.

[0047] In this embodiment, the extracted features can be processed through time-series normalization and confidence assessment, and the confidence level can be calculated. Then, gated weighted fusion is performed to generate fusion weights.

[0048] In some embodiments, the respiratory infection status features can be normalized on a sliding window to obtain normalized scores; the confidence weight of each normalized score can be calculated; and the respiratory infection status features and confidence weights can be fused using a gating weighting method.

[0049] In this embodiment, the extracted respiratory infection features can be normalized on a sliding window, and the confidence weight of each score can be calculated. Then, a gating weighting method is used to fuse the respiratory infection status features and the confidence weights.

[0050] Step S114: Determine the respiratory infection status and update the closed-loop baseline based on the respiratory infection status characteristics.

[0051] In this embodiment, respiratory infection status can be determined and closed-loop baseline can be updated, thereby enabling adaptive detection of respiratory infection status and improving the accuracy of respiratory infection status detection.

[0052] In some embodiments, respiratory infection status can be determined based on adaptive thresholds and multi-scale respiratory infection status based on respiratory infection status characteristics; the baseline can be updated based on historical data and the results of respiratory infection status determination.

[0053] The respiratory infection status assessment and closed-loop baseline update in this embodiment may include: (1) Adaptive threshold for respiratory infection status assessment.

[0054] (2) Assessment of respiratory infection status at multiple scales.

[0055] (3) Update the baseline based on historical data and respiratory infection status.

[0056] This invention provides a real-time monitoring method for respiratory infection status. It integrates user demographic data (age, weight, height) and vital sign data (body temperature, body position, and ADC collected by a multi-parameter respiratory infection monitor). Adaptive baseline compensation and calibration are performed using body position jump mask encoding and body position encoding. The baseline-calibrated data is then processed using Hampel filtering. The processed data is then analyzed using a position-dependent wavelet multi-scale anomaly energy algorithm, a physiological coupling anomaly detection algorithm, and nighttime stable segment hidden state drift detection to extract three types of complementary features. Finally, the respiratory infection status is determined through temporal normalization, confidence assessment, and gated weighted fusion, and the results are incorporated into subsequent detection processes to achieve closed-loop adaptive baseline updates, thereby improving the accuracy of respiratory infection status detection.

[0057] Example 2: This embodiment provides another method for real-time monitoring of respiratory infection status, which is implemented based on the above embodiment. The focus is on describing the specific implementation of the real-time monitoring method for respiratory infection status. (See also...) Figure 2 This is a schematic diagram of another method for real-time monitoring of respiratory infection status.

[0058] In this embodiment, the multi-source data collected can be structured data, and the data fields and structure should be clear and effective. For demographic data, only one collection is needed within a complete testing cycle. For vital sign data, time-series data collection should be performed according to a fixed sampling frequency and a fixed cycle. For example: demographic data: age, weight (w), height (h); vital sign time-series data: body temperature. body position ADC respiratory signal Heart rate Where t is time.

[0059] Assuming the sampling frequency of vital sign data is n, and data is collected and uploaded every m seconds, then each collection yields n×m data points. At the end of the collection period, the remaining uncollected data is uploaded, uploading as many data points as possible. This is based on the time-series vital sign data. The data structure is as follows:

[0060]

[0061]

[0062] ].

[0063] In this embodiment, during the construction of position coding and transition mask coding, the collected supine, left lateral, prone, right lateral, and other five positions are one-hot encoded (a coding technique that converts categorical variables into binary vectors). For any consecutive time intervals... and Define the body position transition mask as follows: ; in, This is the result of the body position change mask.

[0064] The intensity of postural transitions is encoded and continuous transitions are identified through postural transitions: = , .

[0065] in, This is an indicator of postural change. This represents the statistical result of the number of positional jumps within the window, i.e., the total count of consecutive jumps. N is the upper limit of the number of samples; This is the body position change threshold, used to distinguish between significant changes and minor perturbations; This indicates that only when the range of body position change... Greater than the threshold The number of jumps is only counted at certain times, thus filtering out jitter or sensor noise.

[0066] In this embodiment, during adaptive baseline compensation and calibration, the data can be automatically divided into multiple stable segments based on body position coding, and local baseline calculation can be performed on the signal within each stable segment.

[0067] in, The result is calculated based on the baseline. The mean is given by , and MAD is given by . For adaptive weights, The vital signs data collected at time i.

[0068] Compensation is performed on the signals of different segments after the jump, and exponential smoothing is used to track the chronic drift trend, combined with individual parameters for normalization.

[0069] in, The signal after baseline calibration. This is a chronic drift trend. The baseline value at time t. The baseline value at time t-1, Forgetting factor, ; , ; in, To remove signals after baseline drift and chronic drift, The signal is normalized by demographic factors. , These are the weighting coefficients.

[0070] For unstable regions that abruptly change, body position can be used... The position will be processed into a missing value, and a polynomial will be used to fill the missing value.

[0071] For individual baselines of demographic data characteristics, the age factor Defined as: , Age factor This represents the amplification factor of age relative to the baseline. This serves as a basic bias term to ensure the stability of baseline compensation in early childhood. When age At the age of 12, , =0.5, age At the age of 12, , =0.5 Individualized baseline compensation for: ,in This is the individualized baseline compensation at time t.

[0072] The risk assessment results will then be updated: ; in, Let be the baseline preservation coefficient at time t, and , This is the initial baseline preservation coefficient. The risk score at time t, The suppression term makes When it is close to 1 The baseline almost stopped, which made It has the function of a logic switch: ; in, Risk freeze threshold and It achieves high-risk freezing and low-risk normal adaptive updating, so that the baseline is not absorbed by abnormalities and avoids pathological values ​​from contaminating the baseline estimate.

[0073] In this embodiment, the Hampel filtering uses the Hampel function to remove local outliers from the baseline-calibrated signal, eliminating extreme noise and interference samples.

[0074] In this embodiment, the extraction of three complementary features during the respiratory infection status feature extraction process is carried out in parallel, specifically as follows: 1) Position-dependent wavelet multi-scale anomaly energy, for the filtered energy according to the sliding window... (Initially, if the data is less than 30 minutes long, use 10-minute or 20-minute intervals; later, use 30-minute intervals. If the data collection time is more than 2 hours, use 60-minute intervals.) Divide the data into multiple windows, and then perform an L-layer (3 layers recommended) DWT in each window to obtain an approximate coefficient. With detail coefficient And calculate the coefficient energy of each layer. : ; Then calculate scale anomalies. : ; Among them, the initial weights The values ​​are 0.5 for low frequency, 0.3 for mid frequency, and 0.2 for high frequency, and will be adjusted automatically later.

[0075] in, The energy baseline for each body position is calculated as follows: ; Among them, the baseline update coefficient α, α=0.02, is used to ensure slow baseline updates.

[0076] 2) Physiological coupling anomaly detection algorithm, using a short time window (The window for the first two data uploads is 5 minutes, and the window for subsequent data uploads is 10 minutes.) Calculate three pairs of Pearson correlation coefficients for each sliding window to construct the short-time Pearson coupling matrix: , for The correlation coefficient, for The correlation coefficient, for The correlation coefficient; And construct a body position baseline for each pair of couplings. And use L2 distance to calculate coupling offset :

[0077] in, This represents the correlation coefficient between the i-th and j-th physiological signals within the current time window. This represents the baseline value of the correlation coefficient between the i-th and j-th physiological signals under body position p, where i,j represent the indexes of the physiological signals involved in the coupling analysis, such as body temperature, heart rate, and respiration.

[0078] Then, the offset is mapped using z-score to obtain the coupling anomaly. : ; Where, σ( ) represents the Sigmoid function. The mean value of the coupling offset is statistically obtained under conditions of stable body position and low risk. The standard deviation of the coupling offset is statistically obtained under the same conditions. Slope adjustment coefficient, =3, used to control the steepness of the curve.

[0079] 3) Nighttime stable phase latent state drift detection: If the body position does not change within the past window, it is marked as a stable phase. Minimum length of the stable window. Set to 10 minutes, respectively for and Calculate the empirical distribution using KDE When using KL calculation With history distance : ; Finally, the offset fraction was calculated. : .

[0080] In this embodiment, during the time series normalization and confidence assessment process, time series normalization mainly involves... , Normalize them all to first follow the fixed window The z-score can be used to standardize the feature data, and then the Sigmoid function can be used to normalize it to [0,1]. The confidence assessment in this embodiment mainly uses the intensity of positional jumps and the positional factors of continuous jumps. and fever intensity factor Conduct confidence assessment, among which The calculations are mainly as follows:

[0081] in, Let w represent the intensity of the positional jump within the window w, where w represents the length of the sliding time window for the positional change. This is an indicator variable for body position changes.

[0082] in Let represent the confidence level related to body position within the current window, indicating the credibility of the i-th type of feature under the condition of stable body position. This represents the instability that continuously jumps within the current window. This represents the sensitivity coefficient for each feature, with values ​​varying for different features. , , However, if it's in the transition zone... .

[0083] Based on the temperature baseline under different body position conditions, the fever intensity factor is... It can be obtained in the following ways: Calculate the position-dependent baseline temperature difference: ; Let t be the temperature difference at time t. Let t be the real-time body temperature measurement value at time t. Indicates time t. Temperature baseline under body position.

[0084] The temperature difference is standardized using the tanh function. , and then used it for calculation Amplification factor. Temperature difference normalization factor. The calculation is as follows:

[0085] in, For the temperature difference scale parameter, in this example .

[0086] final The calculation is as follows: ; in, For different fever intensities, This represents the sensitivity coefficient for each feature, with values ​​varying for different features. , , However, if Then directly .

[0087] In this embodiment, during the gated weighted fusion process, fusion weights are generated by merging the confidence scores of each processed feature into an overall confidence score. The final weights are then calculated using a temperature-controlled softmax function, and exponential smoothing is employed to avoid instantaneous weight fluctuations. Specifically: The confidence levels are combined into an overall confidence level: , To prevent small constants with zero weights, it is used for numerical stability.

[0088] Final weighting with temperature: ; Standardized confidence level, temperature Temperature control parameters are used to control the probability of selecting a specific feature. In this embodiment... .

[0089] Smooth image stabilization: These are the smoothed feature weights. The smoothing coefficient is preferably selected in this embodiment. .

[0090] In this embodiment, respiratory infection status assessment and closed-loop baseline update refer to fusing normalized features with their respective weights, performing multi-scale risk status assessments at short, medium, and long scales, and then feeding back the final risk at each scale to adaptive baseline compensation and calibration to achieve individual adaptive baseline updates. The method for fusing the feature weights is as follows: , For the normalized features; The final multi-scale risk state is obtained: ; in, This represents a multi-scale risk state. Scores can be merged based on the current gradient / instantaneous values. This is the hourly average of the sliding window values. This is the average cumulative over 3 hours. , , To introduce a time forgetting curve for determination, so that more recent anomalies are given more attention by the short-term channel and the effects of more distant accumulations are retained by the long-term channel, the normalized integral or instantaneous weights of the three curves are calculated to obtain dynamic weights.

[0091] The method provided in the embodiments of the present invention has the following main advantages: 1. Multi-parameter + body position coding mechanism: Existing wearable respiratory infection detection methods do not include body position labeling and body position change event analysis. This invention is the first to construct a body position mask decision structure to reduce false alarms caused by body position interference.

[0092] 2. Most existing algorithms use a single static threshold. This invention adopts an evolvable baseline and combines individual differences to achieve adaptive updates.

[0093] 3. A three-algorithm parallel detection approach was proposed, which consists of a position-dependent wavelet multi-scale anomaly energy algorithm, a physiological coupling anomaly detection algorithm, and a nighttime stable segment hidden state drift detection algorithm. This approach enables the detection of short-term anomalies, physiological coupling instability, and slow evolution risks, thereby identifying the evolution process of hidden nighttime respiratory infections.

[0094] 5. Gated weighted fusion and closed-loop learning: By using confidence weighting and risk results to feed back to the baseline, a continuously optimized closed-loop detection is formed.

[0095] 6. Through real-time and continuous monitoring, the status of respiratory infection can be assessed at multiple scales.

[0096] Example 3: Corresponding to the above method embodiments, this invention provides a real-time monitoring device for respiratory infection status, see [link to relevant documentation]. Figure 3 The diagram shows a real-time monitoring device for respiratory infection status, which includes: The multi-source data acquisition module 31 is used to collect multi-source data from users; the multi-source data includes: demographic data and vital signs data; The body position coding and jump mask coding construction module 32 is used to construct body position coding and jump mask coding based on vital sign data; The adaptive baseline compensation and calibration module 33 is used to perform adaptive baseline compensation and calibration on vital sign data based on body position coding and jump mask coding. The Hampshire filter module 34 is used to perform Hampshire filtering on the vital signs data after adaptive baseline compensation and calibration. The respiratory infection status feature extraction module 35 is used to extract respiratory infection status features based on multi-source data. The temporal normalization, confidence assessment, and gated weighted fusion module 36 is used to perform temporal normalization, confidence assessment, and gated weighted fusion on respiratory infection status characteristics. The respiratory infection status assessment and closed-loop baseline update module 37 is used to assess the respiratory infection status and update the closed-loop baseline based on the respiratory infection status characteristics.

[0097] This invention provides a real-time monitoring device for respiratory infection status. It integrates user demographic data (age, weight, height) and vital sign data (body temperature, body position, and ADC data collected by a multi-parameter respiratory infection monitor). Adaptive baseline compensation and calibration are performed through body position jump mask encoding and body position encoding. The baseline-calibrated data is processed using Hampel filtering. The processed data is then used to extract three complementary features using a body position-dependent wavelet multi-scale anomaly energy algorithm, a physiological coupling anomaly detection algorithm, and nighttime stable segment hidden state drift detection. Finally, the respiratory infection status is determined through temporal normalization, confidence assessment, and gated weighted fusion, and the results are incorporated into subsequent detection processes to achieve closed-loop adaptive baseline updates, thereby improving the accuracy of respiratory infection status detection.

[0098] The aforementioned multi-source data acquisition module is used to acquire users' age, weight, and height data as demographic data; and to acquire users' body temperature, body position, respiratory apparent diffusion coefficient signal, and heart rate time series data as vital signs data through wearable devices.

[0099] The aforementioned body position encoding and jump mask encoding construction module is used to discretize the night body position sequence in the body position data into a finite set of states to form a body position encoding; construct a body position jump mask in the body position change interval of the body position data; wherein, the jump mask is used to identify the region of sudden change in the statistical characteristics of the signal; and perform jump intensity encoding based on the body position jump mask to construct the jump mask encoding.

[0100] The aforementioned adaptive baseline compensation and calibration module is used to fit the initial baseline of vital sign data based on position encoding and jump mask encoding, and to calibrate it by means of steady-state mean or variance; to reset the baseline of vital sign data by jump neighborhood with mask constraints; and to perform time-series dynamic drift tracking and adaptive baseline update of the baseline of vital sign data.

[0101] The aforementioned Hampshire filtering module is used to perform Hampshire filtering on the vital signs data after the baseline of the transition neighborhood under mask constraints is reset.

[0102] The aforementioned respiratory infection status feature extraction module is used to extract respiratory infection status features based on multi-source data, including position-dependent wavelet multi-scale abnormal energy detection, physiological coupling abnormality detection, and nighttime stable segment hidden state drift detection.

[0103] The aforementioned time-series normalization, confidence assessment, and gated weighted fusion modules are used to normalize respiratory infection status features on a sliding window to obtain normalized scores; calculate the confidence weight of each normalized score; and fuse respiratory infection status features with confidence weights using a gated weighted approach.

[0104] The aforementioned respiratory infection status assessment and closed-loop baseline update module is used for adaptive threshold assessment of respiratory infection status based on respiratory infection status characteristics, multi-scale assessment of respiratory infection status, and baseline update based on historical data and the results of respiratory infection status assessment.

[0105] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the real-time monitoring device for respiratory infection status described above can be referred to the corresponding process in the embodiments of the aforementioned real-time monitoring method for respiratory infection status, and will not be repeated here.

[0106] Example 4: This invention also provides an electronic device for running the above-described real-time monitoring method for respiratory infection status; see [link to related documentation]. Figure 4 The diagram shows the structure of an electronic device, which includes a memory 100 and a processor 101. The memory 100 is used to store one or more computer instructions, which are executed by the processor 101 to realize the above-mentioned real-time monitoring method for respiratory infection status.

[0107] Furthermore, Figure 4 The electronic device shown also includes a bus 102 and a communication interface 103, with the processor 101, the communication interface 103 and the memory 100 connected via the bus 102.

[0108] The memory 100 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 103 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 102 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0109] Processor 101 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 101 or by instructions in software form. Processor 101 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 100, and processor 101 reads information from memory 100 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0110] This invention also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are called and executed by a processor, they cause the processor to implement the above-mentioned method for real-time monitoring of respiratory infection status. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0111] The computer program product of the real-time monitoring method, device and electronic device for respiratory infection status provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0112] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and / or device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0113] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0114] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0115] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0116] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions 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, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for real-time monitoring of respiratory infection status, characterized in that, The method includes: Collect multi-source data from users; wherein, the multi-source data includes: demographic data and vital signs data; Based on the aforementioned vital sign data, positional coding and transition mask coding are constructed; Adaptive baseline compensation and calibration are performed on the vital sign data based on the body position encoding and the jump mask encoding. The vital signs data after adaptive baseline compensation and calibration are subjected to Hampshire filtering; Respiratory infection status features were extracted based on the multi-source data; The respiratory infection status characteristics were subjected to time-series normalization, confidence assessment, and gated weighted fusion. The respiratory infection status is determined and the closed-loop baseline is updated based on the respiratory infection status characteristics.

2. The method according to claim 1, characterized in that, The steps for collecting multi-source user data include: Acquire users' age, weight, and height data as demographic data; The user's body temperature data, body position data, respiratory apparent diffusion coefficient signal, and heart rate time series data are collected by wearable devices as vital sign data.

3. The method according to claim 2, characterized in that, The steps for constructing positional coding and transition mask coding based on the vital sign data include: The nighttime body position sequence in the body position data is discretized into a finite set of states to form a body position code; A positional transition mask is constructed within the positional change range of the positional data; wherein, the transition mask is used to identify regions of abrupt changes in the statistical characteristics of the signal; Based on the body position transition mask, transition intensity encoding is performed to construct transition mask encoding.

4. The method according to claim 1, characterized in that, The steps of adaptive baseline compensation and calibration of the vital sign data based on the position encoding and the jump mask encoding include: The initial baseline of the vital signs data is fitted based on the body position code and the jump mask code, and then calibrated by the mean or variance of the steady-state segment. The baseline of the aforementioned vital signs data is reset by masking the transition neighborhood baseline; The baseline of the vital signs data is subjected to time-series dynamic drift tracking and adaptive baseline update.

5. The method according to claim 4, characterized in that, The steps of performing Hampshire filtering on the vital signs data after adaptive baseline compensation and calibration include: The vital signs data after the baseline of the transition neighborhood under mask constraints is reset are subjected to Hampshire filtering.

6. The method according to claim 1, characterized in that, The steps for extracting respiratory infection status features based on the multi-source data include: Based on the multi-source data, position-dependent wavelet multi-scale anomaly energy detection, physiological coupling anomaly detection, and nighttime stable segment hidden state drift detection are performed to extract respiratory infection status features.

7. The method according to claim 1, characterized in that, The steps of performing time-series normalization, confidence assessment, and gated weighted fusion on the respiratory infection status characteristics include: The respiratory infection status characteristics are normalized on a sliding window to obtain normalized scores; Calculate the confidence weight for each of the normalized scores; The respiratory infection status features and the confidence weights are fused using a gating weighting method.

8. The method according to claim 1, characterized in that, The steps for determining the respiratory infection status and updating the closed-loop baseline based on the respiratory infection status characteristics include: Based on the respiratory infection status characteristics, adaptive threshold respiratory infection status judgment and multi-scale respiratory infection status judgment are performed. The baseline is updated based on historical data and the results of respiratory infection status assessment.

9. A real-time monitoring device for respiratory infection status, characterized in that, The device includes: A multi-source data acquisition module is used to collect multi-source data from users; wherein, the multi-source data includes: demographic data and vital sign data; A body position encoding and jump mask encoding construction module is used to construct body position encoding and jump mask encoding based on the vital sign data; An adaptive baseline compensation and calibration module is used to perform adaptive baseline compensation and calibration on the vital sign data based on the body position code and the jump mask code; The Hampshire filter module is used to perform Hampshire filtering on the vital signs data after adaptive baseline compensation and calibration. A respiratory infection status feature extraction module is used to extract respiratory infection status features based on the multi-source data; A time-series normalization, confidence assessment, and gated weighted fusion module is used to perform time-series normalization, confidence assessment, and gated weighted fusion on the respiratory infection status characteristics. The respiratory infection status determination and closed-loop baseline update module is used to determine the respiratory infection status and update the closed-loop baseline based on the respiratory infection status characteristics.

10. An electronic device, characterized in that, The device includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the real-time monitoring method for respiratory infection status as described in any one of claims 1 to 8.

Images