Data transmission method and system for multi-functional vital sign monitoring system

Abstract

The present application relates to the technical field of data transmission, and more particularly to a data transmission method and system of a multifunctional vital sign monitoring system.The method comprises the following steps: collecting first physiological signal data through a sensor arranged in a precordial lead region; collecting second physiological signal data through a sensor arranged in a lower limb or back lead region; aligning the time stamps of the first physiological signal data and the second physiological signal data, and extracting a position calibration vector group; performing compression preprocessing according to the position calibration vector group to obtain a multi-dimensional vital sign compression feature flow; therefore, by constructing a "end-edge-cloud" collaborative vital sign data closed-loop optimization system, the present application solves the problems of spatial information disconnection, high abnormal false alarm rate and threshold rigidity in traditional vital sign monitoring, and improves the accuracy and timeliness of cardiovascular risk early warning in a complex motion scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data transmission, and in particular to a data transmission method and system of a multifunctional vital sign monitoring system. BACKGROUND

[0002] The existing vital sign monitoring system has significant limitations in data transmission: single transmission mode (relying on Bluetooth, Wi-Fi or cellular network) leads to narrow coverage, contradiction between power consumption and rate (low power consumption is insufficient for speed, and high speed consumes too much power), and high network dependency; multi-device cooperation is difficult, limited by the number of Bluetooth connections, difficult to compatible with heterogeneous sensor protocols, and insufficient time synchronization accuracy affects multi-parameter fusion analysis; real-time and reliability are poor, end-to-end delay is uncontrollable, vulnerable to 2.4GHz frequency band interference, and lack of effective data recovery mechanism; continuous high-power transmission and lack of intelligent scheduling severely restrict the endurance of wearable devices; security and privacy protection is weak, with insufficient encryption, authentication defects and the risk of missing end-to-end secure link; introducing additional gateways to solve coverage or connection problems increases system complexity and cost. SUMMARY

[0003] Therefore, it is necessary to provide a data transmission method and system of a multifunctional vital sign monitoring system to solve at least one of the above technical problems.

[0004] To achieve the above purpose, the data transmission method of the multifunctional vital sign monitoring system comprises the following steps:

[0005] Step S1: collecting first physiological signal data through a sensor deployed in the precordial lead area, and collecting second physiological signal data through a sensor deployed in the lower limb or back lead area;

[0006] Step S2: aligning the time stamps of the first physiological signal data and the second physiological signal data, and extracting a position calibration vector group; performing compression preprocessing according to the position calibration vector group to obtain a multi-dimensional vital sign compression feature flow;

[0007] Step S3: using a preset physiological threshold to judge the multi-dimensional vital sign compression feature flow, if it is higher than the preset physiological threshold, triggering an event label mechanism to obtain a current time stamp; searching for the position calibration vector group before compression based on the current time stamp, and packaging it as an abnormal data frame to upload to a preset cloud vital sign analysis platform;

[0008] Step S4: the preset cloud vital sign analysis platform accepts the abnormal data frame and performs abnormal segment return analysis to obtain vital sign threshold optimization data; adjusting the preset physiological threshold based on the vital sign threshold optimization data.

[0009] In the specification, a data transmission system of a multifunctional vital sign monitoring system is provided for performing the data transmission method of the multifunctional vital sign monitoring system described above, and the data transmission system of the multifunctional vital sign monitoring system comprises:

[0010] A multi-lead signal acquisition module is configured to acquire first physiological signal data through sensors deployed in a precordial lead area, and acquire second physiological signal data through sensors deployed in a lower limb or back lead area;

[0011] A feature fusion and compression construction module is configured to align timestamps of the first physiological signal data and the second physiological signal data, and extract a position calibration vector group; and perform compression preprocessing according to the position calibration vector group to obtain a multi-dimensional vital sign compression feature stream;

[0012] An event identification and data uploading module is configured to judge the multi-dimensional vital sign compression feature stream by using a preset physiological threshold value, and if the multi-dimensional vital sign compression feature stream is higher than the preset physiological threshold value, trigger an event label mechanism to obtain a current timestamp; search for the position calibration vector group before compression based on the current timestamp, and encapsulate the position calibration vector group as an abnormal data frame to upload to a preset cloud vital sign analysis platform;

[0013] A cloud abnormality analysis and parameter feedback module is configured to accept the abnormal data frame by the preset cloud vital sign analysis platform, perform abnormal segment return analysis, obtain vital sign threshold optimization data, and adjust the preset physiological threshold value based on the vital sign threshold optimization data.

[0014] The present application has the following beneficial effects: by constructing a multi-level data collaborative processing mechanism, significant breakthroughs are achieved in the aspects of data acquisition, transmission, decision-making and optimization:

[0015] The precordial lead (acquiring electrocardiogram / PPG, etc., first physiological signal) and the lower limb / back lead (acquiring motion acceleration / temperature, etc., second physiological signal) are separately deployed, high-precision timestamp alignment (error <1 ms) is used to solve the problem of lack of spatial information in traditional single-point monitoring, for example, the millisecond-level synchronization of the R wave of electrocardiogram and the peak value of leg acceleration is achieved, and the data foundation for subsequent spatial-physiological joint analysis is laid.

[0016] The position calibration vector group (including three-dimensional coordinates, sensor inclination angle, etc.) is introduced to perform compression preprocessing on the original physiological signal, and the generated multi-dimensional vital sign compression feature stream is reduced to less than 10% of the original data (such as 10 MB→0.9 MB), while retaining key space-time features (such as the gradual change of the leg inclination angle from 15° to 82° in a fall event), and significantly reducing the wireless transmission energy consumption.

[0017] The preset dynamic physiological threshold (such as electrocardio conduction <51 ms, temperature jump >2.2 DEG C) screens abnormal events in real time, and based on a time stamp, the precise backtracking of an uncompressed position calibration vector group (such as GPS coordinates [47.123 N, 11.456 E] and attitude data) is performed, and encapsulated as an encrypted abnormal frame for uploading, so that the cloud can obtain a reproducible complete on-site data chain.

[0018] The cloud analyzes a large number of abnormal data frames (such as 100 athlete falling events), extracts the median and dispersion of the group vital signs (such as the median of electrocardio conduction 46 ms ± 2.5 ms), iteratively generates vital sign threshold optimization data (new threshold = median + 2 sigma = 51 ms), and sends to the terminal device to update the local parameters, forming a closed-loop learning system of 'terminal detection-cloud optimization-strategy delivery'.

[0019] Therefore, by constructing the 'terminal-edge-cloud' collaborative vital sign data closed-loop optimization system, the problems of spatial information disconnection, high abnormal false alarm rate and threshold rigidity in traditional vital sign monitoring are solved, and the accuracy and timeliness of cardiovascular risk early warning in a complex sports scene are improved. BRIEF DESCRIPTION OF DRAWINGS

[0020] Fig. 1 It is a step flowchart of a data transmission method of a multifunctional vital sign monitoring system;

[0021] Fig. 2 It is a leg inclination change curve graph of a falling event;

[0022] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0023] The technical method of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0024] In addition, the accompanying drawings are only schematic illustrations of the present application, and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated description thereof will be omitted. Some block diagrams shown in the drawings are functional entities, which do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0025] It should be understood that, although the terms "first", "second", etc. can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the example embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the associated associated items.

[0026] To achieve the above object, please refer to Figs. 1-2 A data transmission of a multifunctional vital sign monitoring system, the method comprising the following steps:

[0027] In the embodiment of the present application, reference Fig. 1 As shown in the figure, it is a method step flow diagram of data transmission of a multifunctional vital sign monitoring system, in this example, the method of data transmission of the multifunctional vital sign monitoring system comprises the following steps:

[0028] Step S1: collecting first physiological signal data through sensors deployed in the precordial lead region; collecting second physiological signal data through sensors deployed in the lower limb or back lead region;

[0029] In this embodiment, the first physiological signal data and the second physiological signal data include:

[0030] Collecting first physiological signal data through sensors deployed in the precordial lead region, wherein the first physiological signal data includes first electrocardiogram waveform data, first electrical volume pulse wave data, first motion acceleration data and first body surface temperature data;

[0031] Collecting second physiological signal data through sensors deployed in the lower limb or back lead region, wherein the first physiological signal data includes second electrocardiogram waveform data, second electrical volume pulse wave data, second motion acceleration data and second body surface temperature data.

[0032] In the embodiment of the present application, a multi-modal sensor is arranged in the precordial lead area to synchronously collect a first electrocardiogram waveform (sampling rate ≥ 250 Hz), a first capacitive pulse wave (PPG, sampling rate ≥ 100 Hz), a first three-axis motion acceleration (range ± 8g, sampling rate ≥ 50 Hz) and a first body surface temperature (accuracy ± 0.1℃, sampling rate 1 Hz), to form a high-fidelity core vital sign dataset; meanwhile, an isomorphic sensor is arranged in the lower limb / back area to collect a second electrocardiogram waveform (sampling rate ≥ 125 Hz), a second PPG (sampling rate ≥ 50 Hz), a second three-axis motion acceleration (range ± 16g, sampling rate ≥ 50 Hz) and a second body surface temperature (accuracy ± 0.2℃, sampling rate 1 Hz), to form a spatially heterogeneous complementary dataset, and physiological signal spatial gradients (such as electrocardiogram R wave conduction time delay, PPG peak time difference, temperature gradient ΔT) and motion state differences of the trunk and limbs are captured through double-node time synchronization (clock error ≤ 1 ms).

[0033] Step S2: aligning the time stamps of the first physiological signal data and the second physiological signal data, and extracting a position calibration vector group; performing compression preprocessing according to the position calibration vector group to obtain a multi-dimensional vital sign compression feature flow;

[0034] Preferably, step S2 comprises the following steps:

[0035] Step S21: aligning the time stamps of the first physiological signal data and the second physiological signal data to form a synchronized data stream;

[0036] Step S22: extracting a space-physiology joint feature vector according to the synchronized data stream to construct a position calibration vector group, wherein the position calibration vector group comprises a limb relative displacement vector, precordial-lower limb lead data, precordial PPG data and precordial-lower limb body surface temperature difference data;

[0037] Step S23: differentially encoding the limb relative displacement vector, range compressing the precordial-lower limb body surface temperature difference data, and constructing a dimension reduction feature data; using the precordial-lower limb lead data and the precordial PPG data to eliminate motion interference of the position calibration vector group to obtain motion artifact suppression data;

[0038] Step S24: constructing a multi-dimensional vital sign compression feature flow from the dimension reduction feature data and the motion artifact suppression data.

[0039] In the embodiment of the present application, the multi-modal original data time stamps of the precordial and lower limb nodes are aligned by an interpolation synchronization algorithm (such as cubic spline interpolation) (synchronization error < 2ms when sampling rate difference ≤ 5%), to generate a synchronized data stream; a space-physiology joint feature vector is extracted based on the synchronized stream, specifically including:

[0040] 1) limb relative displacement vector (calculate the Euclidean distance difference value of the three-axis acceleration of the chest and lower limb, formula: ;

[0041] 2) chest-lower limb lead ECG R-wave time delay ( );

[0042] 3) chest-lower limb PPG peak time difference ( );

[0043] 4) chest-lower limb body surface temperature difference ( ); then first-order difference encoding is performed on the displacement vector (the change trend is retained), nonlinear range compression is performed on ΔT (such as logarithmic scaling when ΔT>1℃), and reduced dimension feature data is generated; meanwhile, the physiological propagation stability of Δt_ECG and Δt_PPG (still maintaining physiological correlation under motion interference) is used to eliminate the signal baseline drift caused by acceleration mutation through an adaptive filter, and motion artifact suppression data is output; finally, the reduced dimension feature and the suppression data are fused to construct a multi-dimensional signalmotion artifact suppression data.

[0044] In an implementation manner of the embodiment of the application, when a user runs quickly, the chest accelerometer records [1.2g, 0.8g, 2.1g], the lower limb accelerometer records [3.1g, 1.5g, 4.3g], and the following is calculated in real time:

[0045] Limb displacement vector:

[0046] Difference encoding: the current frame Δa increases by 0.6g compared with the previous frame → encoded as +0.6g

[0047] Temperature processing: chest temperature 36.7℃, calf temperature 34.0℃ → ΔT=2.7℃ → compressed to logarithmic value 1.08

[0048] Motion artifact suppression: use Δt_ECG (stabilized at 80ms) to correct the sawtooth noise in the chest PPG waveform caused by running

[0049] The final output compressed feature stream: [+0.6g, 1.08, 80ms, corrected PPG], which only occupies 8% of the storage space of the original data.

[0050] Preferably, step S22 comprises the following steps:

[0051] Extracting a space-physiological joint feature vector according to the synchronized data stream;

[0052] Calculating the angle change of the lower limb-trunk based on the difference value of the acceleration of the space-physiological joint feature vector, to obtain a limb relative displacement vector;

[0053] Based on the R wave time delay of the space-physiological joint feature vector, the conduction velocity of the electrical signal is calculated, and the precordial-lower limb lead data is output; based on the peak time difference of the space-physiological joint feature vector, blood flow wave propagation time measurement is carried out, and the precordial PPG data is obtained;

[0054] Based on the space-physiological joint feature vector, the peripheral circulation state measurement is carried out, and the precordial-lower limb surface temperature difference data is obtained.

[0055] In the embodiment of the application, three-axis acceleration data, dual-lead electrocardiogram (ECG) data, photoplethysmography (PPG) data and surface temperature data of the trunk and lower limbs are synchronously collected by multiple sensors, and a space-physiological joint feature vector (including time stamp, spatial coordinates and physiological signal value) aligned in time and space is constructed.

[0056] Lower limb-trunk angle calculation: the modulus change rate of the acceleration vector (such as trunk acceleration change rate 0.5g / s, lower limb 2.3g / s) is extracted, the difference value is calculated and integrated to obtain a relative displacement vector, and finally the joint angle is output through the inverse trigonometric function;

[0057] ECG conduction velocity measurement: the R wave peak time t1 of the precordial lead ECG and the R wave time t2 of the lower limb lead are identified, and the conduction velocity is calculated with the electrode spacing d=80cm (t2-t1=20ms, v=40m / s);

[0058] Blood flow wave propagation time measurement: the waveform peak time difference Δtp (such as Δtp=250ms) of the precordial PPG and the foot PPG is captured, and the pulse wave propagation speed v is calculated combined with the height h=170cm ;

[0059] Peripheral circulation evaluation: the difference ΔT=4.7℃ between the precordial temperature Tc=36.2℃ and the foot temperature Tf=31.5℃ is calculated, and when ΔT>4℃, it is determined that the peripheral circulation is weakened.

[0060] In one implementation manner of the embodiment of the application,

[0061] The acceleration sensor detects that the trunk acceleration change rate is 0.8g / s, and the lower limb is 2.5g / s, the system calculates the right leg swing angle θ=68° (normal walking reference value 55°±5°) in real time, and prompts the gait abnormality;

[0062] The ECG signal shows that the precordial R wave to the foot R wave delay is 22ms (electrode distance 80cm), the conduction velocity v=36.4m / s (lower than the healthy threshold 40m / s), and the conduction block risk is prompted;

[0063] PPG signal measured chest to foot pulse wave time difference Δtp=280ms (height 175cm), pulse wave velocity PWV=6.25m / s (higher than normal value 5m / s), suggesting a tendency of arteriosclerosis;

[0064] Temperature sensor records chest-foot temperature difference ΔT=5.2℃, the system determines that the lower limb circulation is poor, and generates a warning report based on the above data.

[0065] Preferably, step S2 further comprises:

[0066] In the process of obtaining the multi-dimensional sign compression feature flow, if a packet loss event caused by network fluctuation or device jitter is detected, then based on adjacent time stamps, a time interpolation method is used to reconstruct and repair the missing points, wherein the time interpolation method can use linear interpolation, local polynomial regression or Fourier transform extrapolation according to different data types.

[0067] In the embodiment of the application, by monitoring the time stamp continuity of the sign data stream in real time (such as the theoretical interval of 4ms under the ECG sampling rate of 250Hz), when the actual time stamp interval is detected to be abnormal (such as suddenly increasing from 4ms to 20ms), a packet loss event is determined; according to the data type, an interpolation algorithm is selected:

[0068] Linear interpolation: suitable for slowly changing body temperature data (such as missing points between 36.5℃ and 36.7℃, and reconstructing the intermediate value 36.6℃ according to the same slope);

[0069] Local polynomial regression (LOWESS): processing nonlinear waveform (such as missing peak points of PPG pulse wave, using adjacent 10 cycle data to fit a quadratic curve to predict the missing points);

[0070] Fourier transform extrapolation: reconstructing periodic signals (such as missing R wave segments of ECG, extracting the frequency spectrum features of the previous 1 second ECG to synthesize the filling segment)

[0071] In one implementation manner of the embodiment of the application, the network jitter event in patient remote monitoring:

[0072] Original data stream: the body temperature (0.1Hz sampling) is 36.2℃ at t=10:00:00, and should be 36.5℃ at t=10:00:10, but t=10:00:05 is missing due to packet loss;

[0073] Repair operation: the system detects that the 10-second interval is abnormal (normal 5 seconds), and linear interpolation generates t=10:00:05 body temperature=36.35℃ (calculation: (36.5-36.2) / 2 +36.2).

[0074] Motion acceleration data (100Hz sampling): t=12.3s to 12.31s missing 3 points (normal 0.01s / point), original sequence front point 1.2g, back point 0.8g;

[0075] Repair operation: adopt local polynomial regression (window width 0.05s), fit curve to reconstruct intermediate value: 12.305s=1.0g, 12.307s=0.9g, 12.309s=0.85g.

[0076] ECG signal (250Hz sampling): t=5.0s-5.016s missing 4 points (corresponding to R wave rising branch);

[0077] Repair operation: extract Fourier fundamental frequency component (such as main frequency 1.2Hz) of t=4.9s-5.0s complete cycle, extrapolate to generate missing segment waveform (reconstruct peak 0.85mV).

[0078] Step S3: using a preset physiological threshold to judge the multi-dimensional sign compression feature flow, if higher than the preset physiological threshold, triggering the event label mechanism, acquiring the current timestamp; searching the position calibration vector group before compression based on the current timestamp, and packaging as an abnormal data frame and uploading to the preset cloud end life sign analysis platform;

[0079] Preferably, step S3 includes the following:

[0080] Using a preset physiological threshold to judge the multi-dimensional sign compression feature flow, if the acceleration difference value of the limb relative displacement vector is greater than 3g, the chest-lower limb lead data is less than 50ms, the chest PPG data is greater than 200ms, or the single jump amplitude of the chest-lower limb body surface temperature difference data is higher than 2℃, the event label mechanism is triggered;

[0081] Acquiring the current timestamp;

[0082] Searching the position calibration vector group before compression based on the current timestamp, and packaging as an abnormal data frame;

[0083] Encrypting the abnormal data frame, and uploading to the preset cloud end life sign analysis platform.

[0084] In the embodiment of the application, the key parameters in the multi-dimensional sign compression feature flow are monitored in real time:

[0085] Acceleration difference value: calculate the acceleration change rate of adjacent frames of the limb relative displacement vector (such as 1.5g of the previous frame→4.8g of the next frame, difference value=3.3g);

[0086] ECG conduction time: detect the R wave interval of the chest-lower limb lead data (such as normal 70ms, abnormal value 45ms);

[0087] Pulse wave propagation time: analyze the peak time difference between chest PPG and foot PPG (e.g. normal 180ms, abnormal 220ms);

[0088] Body surface temperature jump: calculate the instantaneous change of temperature difference between chest and lower limb (e.g. temperature difference from 1.8℃ to 4.0℃ within 5 minutes, jump amplitude 2.2℃).

[0089] When any parameter exceeds the preset threshold (>3g, <50ms, >200ms, >2℃ jump):

[0090] Record the current timestamp (e.g. 2023-08-15 14:05:30.250)

[0091] Backtrack the uncompressed position calibration vector set (including three-dimensional coordinates / sensor orientation, etc. original data) before this moment

[0092] Packaged as an encrypted abnormal frame (structure: timestamp + original position data + abnormal parameter value)

[0093] Upload to the cloud platform after AES-256 encryption.

[0094] In an implementation manner of the embodiment of the application, sudden accident monitoring:

[0095] Acceleration difference value: leg sensor detects 2.9g→6.1g jump (difference value 3.2g > 3g threshold), determined as a fall impact;

[0096] ECG conduction time: chest-foot lead R wave interval is only 42ms (<50ms threshold), indicating abnormal conduction of cardiac electrical signals;

[0097] Pulse wave propagation time: chest-foot PPG time difference reaches 230ms (>200ms threshold), showing peripheral circulation obstruction;

[0098] Temperature jump: chest temperature 36.5℃ vs foot temperature 28.0℃, temperature difference 8.5℃ and jump 3.0℃ within 5 minutes (>2℃ threshold), indicating lower limb hypothermia.

[0099] Lock abnormal timestamp 2023-08-15 14:05:30.250

[0100] Extract the uncompressed position calibration vector set at this moment (e.g. GPS coordinates [47.123°N, 11.456°E], leg sensor inclination 82°)

[0101] Package abnormal frame data:

[0102] {

[0103] timestamp: "2023-08-15T14:05:30.250Z",

[0104] location: [47.123,11.456,2480m],

[0105] anomaly_data: {

[0106] acc_diff: 3.2g,

[0107] ecg_delay: 42ms,

[0108] ppg_delay: 230ms,

[0109] temp_delta: 8.5℃

[0110] }

[0111] }

[0112] Upload to the cloud after encryption, trigger the rescue system.

[0113] Preferably, searching for the pre-compressed location calibration vector group based on the current timestamp includes:

[0114] Taking a time window forward / backward from the current timestamp, generate adjacent window timestamps;

[0115] Based on the adjacent window timestamps and the current timestamp, search for the pre-compressed location calibration vector group respectively, to obtain the location calibration vector group of the previous timestamp, the location calibration vector group of the current timestamp, and the location calibration vector group of the next timestamp;

[0116] According to the location calibration vector group of the previous timestamp, the location calibration vector group of the current timestamp, and the location calibration vector group of the next timestamp as the frame header, frame body and frame tail respectively, and perform compression and packaging, to obtain the anomaly data frame.

[0117] When detecting a physical anomaly (such as cardiac arrest):

[0118] Define a time window: take a fixed time length (such as 1 second) forward / backward from the anomaly timestamp (such as 14:05:30.500) as the center, generate adjacent window timestamps (front window: 14:05:29.500-30.500, back window: 14:05:30.500-31.500);

[0119] Trace back the original data:

[0120] Extract from the uncompressed original database:

[0121] Position calibration vector group corresponding to the front window timestamp (such as [GPS: 47.123 ° N, 11.456 ° E; leg inclination: 75 °])

[0122] Position calibration vector group corresponding to the current abnormal timestamp (such as [GPS: 47.123 ° N, 11.457 ° E; leg inclination: 82 °])

[0123] Position calibration vector group corresponding to the rear window timestamp (such as [GPS: 47.124 ° N, 11.458 ° E; leg inclination: 5 °])

[0124] The leg inclination change as shown in Fig. 2

[0125] Encapsulating the abnormal frame: frame header = front window data (recording the motion state before the abnormality occurs); frame body = current data (recording the accurate position at the moment of the abnormality); frame tail = rear window data (recording the posture change after the abnormality)

[0126] The three are compressed into a single data packet (such as from 3MB original data to 300KB)

[0127] Preferably, uploading to the preset cloud vital sign analysis platform also includes:

[0128] Encrypting the abnormal data frame and uploading it to the preset cloud vital sign analysis platform while performing three-dimensional coding through the lead position information of the abnormal data frame to obtain coordinate point group data.

[0129] Preferably, step S4 includes:

[0130] The preset cloud vital sign analysis platform receives the abnormal data frame and performs abnormal segment return analysis to obtain abnormal sign data; uses the coordinate point group data to perform signal alignment of enhanced fusion features on the abnormal sign data, and iteratively optimizes the data based on the preset physiological threshold as a reference to obtain sign threshold optimization data;

[0131] Based on the sign threshold optimization data, the preset physiological threshold is adjusted, and a sign monitoring optimization log is constructed.

[0132] In the embodiment of the application, after the cloud platform receives the encrypted abnormal data frame (including the position calibration vector group of the front / middle / rear time window and the abnormal sign data):

[0133] Abnormal segment return analysis: decompressing the data frame to restore the original information (such as the three-stage change of leg inclination from 15 ° → 82 ° → 5 ° in a fall event), and extracting abnormal signs (such as ECG conduction time 42 ms);

[0134] ​Enhanced fusion alignment: Align position coordinates (e.g., GPS [47.123, 11.456] → [47.124, 11.458]) with physiological signal timeline: Match acceleration differential peak time (e.g., 14:05:30.500 speed drop point corresponds to ECG abnormal time) through speed change curve; Correct signal deviation caused by sensor delay (e.g., adjust foot PPG time delay from 230 ms to 215 ms);

[0135] Threshold iterative optimization: Statistic similar event data (e.g., 100 skiing fall records): Original threshold: ECG conduction < 50 ms is abnormal Actual data median: 45 ms (athletes generally faster) Calculate new threshold: 45 ms + 2 × standard deviation (3 ms) = 51 ms → Update to < 51 ms trigger;

[0136] Construct optimization log: Record original threshold 50 ms, new threshold 51 ms, effective time, etc., and issue to terminal device.

[0137] Preferably, step S4 further comprises:

[0138] The sign monitoring optimization log returned from the preset cloud vital sign analysis platform performs multi-time scale trend extraction on each channel in the compressed data frame, divides multiple levels of early warning levels according to the extracted trend abnormality degree, and generates a parameter adjustment suggestion with a level label accordingly.

[0139] In the embodiment of the application, after the terminal device receives the sign monitoring optimization log (including an updated threshold such as ECG conduction < 51 ms) issued by the cloud, the terminal device performs multi-time scale analysis on the locally stored compressed data frame (including ECG / acceleration / temperature channels): Short-term trend (10-minute window): Calculate the sliding average value of each channel data (e.g., ECG conduction value from 45 ms to 49 ms, slope + 0.4 ms / min); Medium-term trend (24-hour window): Extract daily data at the same period for comparison (e.g., leg acceleration peak value decreases by 5% per day at 14:00 every day); Long-term trend (7-day window): Fit the linear change rate (e.g., foot temperature weekly average decreases by 0.3°C per day);

[0140] Divide early warning levels based on trend abnormality degree:

[0141] First-level early warning: Single-channel trend exceeds preset limit value (e.g., ECG conduction weekly increase > 0.5 ms / day);

[0142] Second-level early warning: Dual-channel associated abnormality (acceleration peak value decrease + temperature decrease);

[0143] Third-level early warning: Three-channel deterioration and consistent with pathological model (e.g., ECG increase + temperature decrease + gait disorder);

[0144] Finally, the parameter adjustment suggestion with level label is generated (e.g., when the three-level warning, the ECG threshold is automatically relaxed by 2 ms and the sampling rate is increased).

[0145] Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, the scope of the application being defined by the appended claims and not by the above description, therefore all variations falling within the meaning and scope of the equivalent elements of the application file are intended to be included in the present application.

[0146] The foregoing is considered as merely illustrative of the principles of the application, and the application is to be regarded as not limited to such embodiments, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data transmission method for a multifunctional vital signs monitoring system, characterized in that, Includes the following steps: Step S1: Acquire first physiological signal data using sensors deployed in the precordial lead area; Second physiological signal data is collected by sensors deployed in the lead areas of the lower limbs or back. Step S2: Align the timestamps of the first and second physiological signal data and extract the position calibration vector group; perform compression preprocessing based on the position calibration vector group to obtain the multidimensional vital sign compressed feature stream; Step S2 is as follows: Step S21: Align the timestamps of the first physiological signal data and the second physiological signal data to form a synchronized data stream; Step S22: Extract the spatial-physiological joint feature vector based on the synchronized data stream and construct a position calibration vector group, which includes limb relative displacement vector, chest-lower limb lead data, chest PPG data and chest-lower limb surface temperature difference data. Step S23: Differential encoding is performed on the relative displacement vector of the limbs, the temperature difference data between the chest and lower limb surfaces is compressed, and dimensionality-reduced feature data is constructed. Motion artifact suppression data were obtained by eliminating motion interference from the position calibration vector group using precordial-lower limb lead data and precordial PPG data; Step S24: Construct a multidimensional feature compression flow from the dimensionality-reduced feature data and the motion artifact suppression data; Step S3: Use a preset physiological threshold to judge the multidimensional vital sign compressed feature stream. If it is higher than the preset physiological threshold, trigger the event labeling mechanism to obtain the current timestamp. Based on the current timestamp, search for the uncompressed location calibration vector group, encapsulate it into an abnormal data frame, and upload it to the preset cloud-based vital signs analysis platform; Step S4: The preset cloud-based vital signs analysis platform receives abnormal data frames and performs abnormal segment back transmission analysis to obtain vital sign threshold optimization data. The preset physiological thresholds are adjusted based on the optimized data of vital sign thresholds.

2. The data transmission method of the multifunctional vital signs monitoring system as described in claim 1, characterized in that, Step S22 includes the following: Extract spatial-physiological joint feature vectors from synchronized data streams; The change in the angle between the lower limb and trunk is calculated based on the difference in acceleration from the spatial-physiological joint feature vector, thus obtaining the relative displacement vector of the limb. Based on the R-wave delay of the spatial-physiological joint feature vector, the conduction velocity of the electrical signal is calculated, and the chest-lower limb lead data is output; based on the peak time difference of the spatial-physiological joint feature vector, the blood flow wave propagation time is measured to obtain the chest PPG data. Based on spatial-physiological joint feature vector measurement of peripheral circulation status, the temperature difference data between the chest and lower limb surfaces were obtained.

3. The data transmission method of the multifunctional vital signs monitoring system as described in claim 1, characterized in that, Step S2 also includes: In the process of obtaining the multidimensional vital sign compressed feature stream, if a packet loss event caused by network fluctuations or device jitter is detected, the missing points are reconstructed and repaired based on adjacent timestamps using time interpolation. The time interpolation method can be linear interpolation, local multinomial regression, or Fourier transform extrapolation, depending on the data type.

4. The data transmission method of the multifunctional vital sign monitoring system as described in claim 1, characterized in that, Step S3 includes the following: The multidimensional vital signs compression feature stream is judged using preset physiological thresholds. If the acceleration difference of the relative displacement vector of the limb is greater than 3g, the chest-lower limb lead data is less than 50ms, the chest PPG data is greater than 200ms, or the single jump amplitude of the chest-lower limb surface temperature difference data is greater than 2℃, the event labeling mechanism is triggered. Get the current timestamp; Search for the uncompressed location calibration vector group based on the current timestamp and encapsulate it into an abnormal data frame; The abnormal data frames are encrypted and uploaded to a pre-set cloud-based vital signs analysis platform.

5. The data transmission method of the multifunctional vital signs monitoring system as described in claim 4, characterized in that, The search for the uncompressed location calibration vector set based on the current timestamp includes: Take time windows forward and backward from the current timestamp to generate adjacent window timestamps; Based on the adjacent window timestamps and the current timestamp, search for the position calibration vector group before compression to obtain the position calibration vector group of the previous timestamp, the position calibration vector group of the current timestamp, and the position calibration vector group of the next timestamp. The position calibration vector groups of the previous timestamp, the current timestamp, and the next timestamp are used as the frame header, frame body, and frame tail, respectively, and then compressed and packaged to obtain the abnormal data frame.

6. The data transmission method of the multifunctional vital signs monitoring system as described in claim 5, characterized in that, The data uploaded to the pre-set cloud-based vital signs analysis platform also includes: While encrypting abnormal data frames and uploading them to a pre-set cloud-based vital signs analysis platform, the system performs three-dimensional encoding using the lead position information of the abnormal data frames to obtain coordinate point group data.

7. The data transmission method of the multifunctional vital signs monitoring system as described in claim 6, characterized in that, Step S4 includes: The pre-set cloud-based vital sign analysis platform receives abnormal data frames and performs abnormal segment back transmission analysis to obtain abnormal vital sign data; it uses coordinate point group data to enhance the signal alignment of the fusion features of the abnormal vital sign data, and iteratively optimizes the data based on the pre-set physiological threshold to obtain vital sign threshold optimized data. Based on the vital sign threshold optimization data, the preset physiological thresholds are adjusted, and a vital sign monitoring optimization log is constructed.

8. The data transmission method of the multifunctional vital signs monitoring system as described in claim 7, characterized in that, Step S4 also includes: The system extracts multi-timescale trends from the vital signs monitoring optimization logs returned from the pre-set cloud-based vital signs analysis platform, classifies the extracted trends into multiple warning levels based on their degree of abnormality, and generates parameter tuning suggestions with level labels accordingly.

9. A data transmission system for a multifunctional vital signs monitoring system, characterized in that, A data transmission method for performing the multifunctional vital signs monitoring system as described in claim 1 includes: The multi-lead signal acquisition module is used to acquire first physiological signal data through sensors deployed in the anterior chest lead area; and to acquire second physiological signal data through sensors deployed in the lower limb or back lead area. The feature fusion and compression module is used to align the timestamps of the first and second physiological signal data and extract the position calibration vector group; the compression preprocessing is performed based on the position calibration vector group to obtain the multidimensional vital sign compressed feature stream; The event recognition and data upload module is used to judge the multidimensional vital sign compressed feature stream using a preset physiological threshold. If it is higher than the preset physiological threshold, the event tagging mechanism is triggered to obtain the current timestamp. Based on the current timestamp, the module searches for the location calibration vector group before compression and encapsulates it into an abnormal data frame and uploads it to the preset cloud vital sign analysis platform. The cloud-based anomaly analysis and parameter feedback module is used by the preset cloud-based vital sign analysis platform to receive abnormal data frames and perform abnormal segment feedback analysis to obtain vital sign threshold optimization data; and to adjust the preset physiological thresholds based on the vital sign threshold optimization data.

Images