Pre-hospital emergency treatment multi-device data acquisition method and pre-hospital emergency treatment multi-device data acquisition system

Abstract

The invention discloses a pre-hospital first-aid multi-device data acquisition method and system, and relates to the technical field of device data processing, and the method comprises the steps: obtaining a multi-device data stream; according to the multi-device data stream, performing protocol understanding degree evaluation on the data points to obtain a protocol understanding confidence score; performing time synchronization accuracy evaluation on the data points to obtain a time confidence score; performing data flow continuity evaluation on the data points to obtain a continuity confidence score; calculating an accumulated confidence score of the data point according to the protocol understanding confidence score, the time confidence score and the continuity confidence score; normalizing the data points according to the accumulated confidence score; and carrying out visualization processing on the data points after normalization processing. According to the method, the data points can be normalized in combination with various confidence scores, so that multi-device data acquisition is realized, and the accuracy and the reliability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of device data processing, and in particular to a pre-hospital emergency multi-device data acquisition method and system. BACKGROUND

[0002] In the complex environment of pre-hospital emergency, real-time and accurate acquisition of patient vital signs and key physiological parameters is the basis for timely and effective intervention. Existing systems acquire real-time state information of patients by integrating data from various medical devices. However, when new diagnostic devices are introduced, they often come with manufacturer-specific communication protocol details, which can cause unnoticed errors in data acquisition, processing and display, thereby affecting the judgment of emergency personnel, and the accuracy and reliability of data acquisition are low.

[0003] In summary, the technical problems in the related art need to be improved. SUMMARY

[0004] The main purpose of the embodiments of the present application is to provide a pre-hospital emergency multi-device data acquisition method and system, which can normalize data points by combining multiple confidence scores, to realize multi-device data acquisition and improve accuracy and reliability.

[0005] In one aspect, the present application provides a pre-hospital emergency multi-device data acquisition method, comprising the following steps: acquiring a multi-device data stream, the multi-device data stream comprising a plurality of data points; According to the multi-device data stream, the protocol understanding degree of the data points is evaluated to obtain a protocol understanding confidence score; The data points are evaluated for time synchronization accuracy to obtain a time confidence score; The data points are evaluated for data stream continuity to obtain a continuity confidence score; According to the protocol understanding confidence score, the time confidence score and the continuity confidence score, the cumulative confidence score of the data points is calculated; According to the cumulative confidence score, the data points are normalized; According to the cumulative confidence score, the normalized data points are visualized.

[0006] In another aspect, the present application provides a pre-hospital emergency multi-device data acquisition system, comprising: A data acquisition module is configured to acquire a multi-device data stream, the multi-device data stream comprising a plurality of data points; a protocol understanding evaluation module configured to evaluate protocol understanding degrees of the data points according to the multi-device data stream, and obtain protocol understanding confidence scores; a time synchronization evaluation module configured to evaluate time synchronization accuracies of the data points, and obtain time confidence scores; a data stream continuity evaluation module configured to evaluate data stream continuities of the data points, and obtain continuity confidence scores; a confidence score calculation module configured to calculate cumulative confidence scores of the data points according to the protocol understanding confidence scores, the time confidence scores and the continuity confidence scores; a normalization processing module configured to perform normalization processing on the data points according to the cumulative confidence scores; a visualization processing module configured to perform visualization processing on the normalized data points according to the cumulative confidence scores.

[0007] The embodiments of the present application have at least the following beneficial effects: the embodiments of the present application first acquire a multi-device data stream, then evaluate protocol understanding degrees of data points to obtain protocol understanding confidence scores, evaluate time synchronization accuracies to obtain time confidence scores, and evaluate data stream continuities to obtain continuity confidence scores, then calculate cumulative confidence scores of the data points according to the protocol understanding confidence scores, the time confidence scores and the continuity confidence scores, and finally perform normalization processing on the data points and perform visualization processing on the normalized data points, so that the data points can be normalized processed in combination with multiple confidence scores, multi-device data acquisition is realized, and the accuracy and reliability are improved.

[0008] Other features and advantages of the present application will be further described in the following description, and will be apparent from the description, or will be learned through implementation of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures specifically pointed out in the description and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced.

[0010] Figure 1 a flowchart of a pre-hospital emergency multi-device data acquisition method according to an embodiment of the present application; Figure 2 a structural schematic diagram of a pre-hospital emergency multi-device data acquisition system according to an embodiment of the present application. DETAILED DESCRIPTION

[0011] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments.

[0012] In the related art, in the complex environment of pre-hospital emergency, real-time and accurate collection of patient vital signs and key physiological parameters is the basis for timely and effective intervention. At present, a multi-device data collection method is widely used in ambulances, aiming to integrate data from various medical devices. However, in the context of the continuous development of medical technology, new and more advanced diagnostic devices are continuously introduced to improve emergency capabilities. These new devices often come with manufacturer-specific communication protocol details, posing a considerable challenge to existing data collection terminals, which may result in subtle defects in data collection, processing and display, thereby affecting the judgment of emergency personnel, and low data collection accuracy and reliability.

[0013] Specifically, a new generation of compact and fast-responding diagnostic devices, such as a portable blood lactate and electrolyte analyzer, are introduced into standard emergency medical kits. This device can provide key real-time information about the patient's metabolic state, which is crucial for guiding emergency treatment decisions. Although this new analyzer uses a certain widely accepted communication standard, such as a variant of the USB communication device class (USB-CDC) or a proprietary extension on top of the standard Bluetooth Low Energy (BLE) protocol, its specific data payload structure and command set contain manufacturer-specific enhancements. These enhancements may involve specialized data compression methods, custom error check fields, or a set of extended diagnostic markers beyond the scope of the published standard. Therefore, existing data collection terminals, which are designed to collect data from a series of established medical devices (such as electrocardiogram monitors, pulse oximeters, and automatic blood pressure meters), and whose internal parsing modules are strictly programmed according to published general communication protocols, lack specific firmware or software drivers to fully interpret the subtle changes in the data stream of this new analyzer. The collection terminal can establish a basic connection and identify the device type, but it cannot fully interpret the precise meaning of its proprietary extensions or all data fields. This incomplete understanding constitutes the initial technical obstacle faced by the multi-device data collection method.

[0014] Based on this incomplete protocol understanding, a subtle but critical issue arises with respect to time synchronization. Because the data acquisition terminal cannot fully understand the proprietary communication protocol of the new portable analyzer, it cannot properly negotiate or interpret the analyzer’s internal clock signal or timestamp conventions. For example, the acquisition terminal can send a standard network time protocol (NTP) or precision time protocol (PTP) synchronization request, but the new analyzer’s firmware can expect a slightly different command structure or data format for time alignment, so it can only partially process the request or respond with its own internal clock reference without fully adjusting to the terminal’s master time. This results in a persistent, albeit slight, time offset between the new analyzer’s internal clock and the data acquisition terminal’s system-level time reference. This offset can be in the range of tens to hundreds of milliseconds. In the dynamic and often chaotic environment of an emergency scene, medical devices are frequently connected, disconnected, and reconnected as patient conditions change or different diagnostic needs arise, making this subtle time drift particularly difficult to detect. The randomness of device connection order and the rapid succession of data points from multiple data sources mean that the acquisition terminal’s general time synchronization methods, which are typically designed to identify and correct larger, more obvious clock discrepancies, often miss this fine, persistent time misalignment. The system is aware of the connection, data flow, and general time alignment, but the precise time relationship between the new device data and other synchronized data streams still has a subtle flaw.

[0015] In the complex scenario of pre-hospital emergency care, where time is of the essence, information is incomplete, and the environment is variable, when the data acquisition terminal connects to a new medical device whose protocol details are not fully disclosed, it is necessary to overcome the cumulative effects of multiple hidden factors such as incomplete protocol understanding, subtle and unobservable time synchronization deviations between devices, small data packet silent loss caused by resource competition in non-hard real-time operating systems, and false positive data discontinuity due to faulty error handling logic, to accurately detect and characterize those small, silent, and persistent data integrity defects that affect the accuracy of derived physiological parameters, avoid misleading clinical decisions for emergency personnel, and ensure the correctness of patient treatment plans.

[0016] In view of this, the present application significantly improves the accuracy and reliability of pre-hospital emergency data collection by introducing a multi-dimensional confidence evaluation mechanism. First, by obtaining multi-device data streams, the present application can integrate raw data from different medical devices. Second, the data points are evaluated for protocol understanding, time synchronization accuracy and data stream continuity, respectively obtaining protocol understanding confidence scores, time confidence scores and continuity confidence scores. These evaluation steps can identify and quantify potential problems in data protocol analysis, time alignment and transmission integrity. Then, the cumulative confidence score of the data points is calculated according to these confidence scores, providing a comprehensive reliability indicator for each data point. Finally, the data points are normalized and visualized based on the cumulative confidence score. In the normalization stage, the filtering strength or correction algorithm can be dynamically adjusted according to the reliability of the data points to maximize the elimination of noise and bias. In the visualization stage, the display mode can be adjusted according to the reliability of the data points, such as color, line style or warning information to remind emergency personnel of the confidence level of the data, thereby avoiding misjudgment caused by data defects.

[0017] The embodiments of the present application will be specifically explained below in conjunction with the accompanying drawings: Figure 1 is an optional flowchart of a pre-hospital emergency multi-device data collection method provided by an embodiment of the present application, Figure 1 The method in can include but is not limited to steps S101 to S107.

[0018] Step S101, obtaining a multi-device data stream, the multi-device data stream including a plurality of data points; Step S102, according to the multi-device data stream, evaluating the data points for protocol understanding, obtaining protocol understanding confidence scores; Step S103, evaluating the data points for time synchronization accuracy, obtaining time confidence scores; Step S104, evaluating the data points for data stream continuity, obtaining continuity confidence scores; Step S105, calculating the cumulative confidence scores of the data points according to the protocol understanding confidence scores, the time confidence scores and the continuity confidence scores; Step S106, normalizing the data points according to the cumulative confidence scores; Step S107, visualizing the normalized data points according to the cumulative confidence scores.

[0019] The steps S101 to S107 shown in the embodiments of the present application can normalize the data points in combination with various confidence scores to realize multi-device data collection and improve accuracy and reliability.

[0020] In some embodiments, steps S101-S107 can first acquire a multi-device data stream, where the multi-device data stream includes a plurality of data points. The data acquisition module can be used for collection, which is configured with a variety of physical interfaces (such as USB, Bluetooth, Wi-Fi, serial port, etc.) to be compatible with different types of medical devices. When the device is connected, the data acquisition module continuously listens to and receives data from these devices. For example, a polling mechanism can be used, and the data acquisition module sends a data request to each connected medical device in turn at a preset frequency, receives the data packet returned by the device, and temporarily stores it in the internal buffer. Alternatively, the data acquisition module can be configured in an interrupt-driven mode, and when any connected medical device has new data available, the device will trigger an interrupt signal, and the data acquisition module will respond to the interrupt and read the corresponding data. It can be understood that the multi-device data stream refers to a collection of raw data from multiple medical devices (such as electrocardiogram monitors, oximeters, ventilators, etc.) in a pre-hospital emergency scene. These data streams usually contain multiple data points, each of which represents one or more physiological parameters or device status information collected by a specific device at a specific time point. For example, a data point can include heart rate, oxygen saturation, blood pressure, etc.

[0021] Then, according to the multi-device data stream, the data points are evaluated for protocol understanding degree to obtain a protocol understanding confidence score. For example, a database containing known medical device communication protocol features can be established in advance. When a new multi-device data stream is received, the system can try to match the features of the data stream with the protocol features in the database. The higher the matching degree, the higher the protocol understanding confidence score. Specifically, the received data stream can be parsed to extract features such as data packet header information, data field length, checksum, etc. These extracted features are compared with the preset protocol template. If the structure of the data stream is highly consistent with a known protocol template, it can be considered that the understanding degree of the protocol is high, and a higher protocol understanding confidence score is given.

[0022] The data points are then evaluated for time synchronization accuracy to obtain a time confidence score. For example, each connected medical device can be assigned an independent timer, and periodically synchronized with the system clock of the data acquisition terminal. By comparing the difference between the device timer and the system clock, the time deviation can be calculated. The smaller the time deviation, the higher the time confidence score. Specifically, standard time synchronization mechanisms such as Network Time Protocol (NTP) or Precision Time Protocol (PTP) can be used. The data acquisition terminal acts as a time server, periodically sends a time synchronization request to each medical device, and calculates the round-trip delay and time offset according to the time stamp returned by the device. By continuously monitoring these offsets, the accuracy of time synchronization can be evaluated.

[0023] A data stream continuity assessment is performed on the data points, resulting in a continuity confidence score. For example, the sequence numbers of data packets in the data stream can be monitored. If there is a jump or a missing data packet, it indicates that there is a discontinuity in the data stream. The more missing data packets, the lower the continuity confidence score. Specifically, each data packet can be assigned an incremental sequence number. When the data acquisition module receives a data packet, it checks whether its sequence number is consecutive to the expected sequence number. If a discontinuity in the sequence number is found, it is recorded as a data loss event. By counting the data loss rate within a certain time window, the continuity confidence score can be calculated.

[0024] Based on the protocol understanding confidence score, the time confidence score, and the continuity confidence score, a cumulative confidence score for the data points is calculated. For example, a weighted average method can be used. Each confidence score is assigned a weight, and then they are added together to get the cumulative confidence score. The weights can be adjusted according to the importance of different confidences in the emergency scenario. Specifically, a function can be defined that takes the three confidence scores as input and outputs a cumulative confidence score between 0 and 1.

[0025] Based on the cumulative confidence score, the data points are normalized. For example, when the cumulative confidence score is low, more stringent filtering can be applied to the data points to eliminate potential noise and outliers. When the cumulative confidence score is high, lighter filtering can be used to preserve the original characteristics of the data. Specifically, the parameters of the filter can be dynamically adjusted according to the cumulative confidence score. For example, for data points with low confidence, a stronger low-pass filter or median filter can be used to smooth the data; for data points with high confidence, a weaker filter or the original data can be used directly. In addition, if the cumulative confidence score is below a certain threshold, the system can also mark the data point as "suspicious" or "unreliable" and trigger an appropriate alarm.

[0026] Finally, based on the cumulative confidence score, the normalized data points are visualized. For example, when the cumulative confidence score is low, visual warnings can be added when displaying the data, such as marking the data points in red or displaying them in dashed lines, to remind emergency personnel that the reliability of the data may be low. When the cumulative confidence score is high, the data can be displayed in the usual way. Specifically, the display of the data can be adjusted according to the cumulative confidence score. For example, for data with high confidence, it can be displayed in solid lines, standard colors, and font sizes; for data with medium confidence, it can be displayed in dashed lines or lighter colors; for data with low confidence, it can be displayed in flashing, red, or with question marks, or even temporarily hidden until its confidence improves. In addition, the scaling ratio or time axis density of the chart can also be adjusted according to the cumulative confidence score to highlight reliable data.

[0027] By the above technical solutions, the embodiment can more comprehensively and accurately identify data defects by quantitatively evaluating protocol understanding degree, time synchronization accuracy and data flow continuity, and comprehensively calculating cumulative confidence scores. For example, when the protocol details of a new medical device are not fully disclosed, the protocol understanding evaluation module of the embodiment can identify the limitations of protocol understanding and assign a lower protocol understanding confidence score, so that the data is more carefully processed in subsequent processing. In addition, by normalizing and visualizing the data according to the cumulative confidence scores, the embodiment can dynamically adjust the data processing strategy and display mode to ensure that emergency personnel can always obtain the most reliable and accurate physiological parameter information. This adaptive processing capability makes the application have stronger robustness and adaptability when facing complex and variable pre-hospital emergency environments and constantly updated medical devices, significantly improving the medical safety and decision accuracy of pre-hospital emergency.

[0028] In some embodiments, in step S102, the protocol understanding degree evaluation of the data points is performed according to the multi-device data stream to obtain the protocol understanding confidence score, which can include but is not limited to the following steps: Low-level feature calculation is performed on the multi-device data stream to obtain a low-level feature statistical distribution, and the low-level feature statistical distribution includes byte entropy value, short sequence hash value and target byte pair occurrence frequency; A semantic fingerprint vector is generated according to the low-level feature statistical distribution; The semantic fingerprint vector is compared with a semantic fingerprint library to analyze the data field semantics, and a semantic fingerprint comparison result is obtained; The protocol understanding degree evaluation of the data points is performed according to the semantic fingerprint comparison result to obtain the protocol understanding confidence score.

[0029] In some embodiments, low-level feature calculation can be performed on the multi-device data stream to obtain a low-level feature statistical distribution, wherein the low-level feature statistical distribution includes byte entropy value, short sequence hash value and target byte pair occurrence frequency. By analyzing the byte sequence of the original data stream, statistical quantities that can reflect its structure and characteristics can be extracted. The byte entropy value is used to measure the randomness or complexity of the byte distribution in the data stream. A high entropy value may indicate that the data has been encrypted or compressed, while a low entropy value may indicate that the data has a repetitive pattern or fixed structure. The short sequence hash value is used to quickly identify the repeated short byte sequences in the data stream, which may represent protocol headers, delimiters or specific data fields. The target byte pair occurrence frequency focuses on the co-occurrence pattern of specific byte pairs in the data stream, which helps to identify specific instructions or data types in the protocol. Through the calculation of these low-level features, the internal structure information of the data stream can be comprehensively captured.

[0030] Then, according to the statistical distribution of low-level features, a semantic fingerprint vector is generated. The semantic fingerprint vector is an abstract representation of the data stream structure and semantic characteristics, which encodes complex low-level statistical information into a compact numerical vector to facilitate subsequent comparison and analysis. The vector can serve as an identity of the data stream, reflecting the unique properties of the protocol or data format it belongs to.

[0031] The semantic fingerprint vector is then compared with the semantic fingerprint library to analyze the semantics of the data fields and obtain the semantic fingerprint comparison result. The semantic fingerprint library stores typical semantic fingerprint vectors of known protocols or data formats. Through comparison, the semantics of the data fields can be analyzed, that is, the communication protocol or data structure that the data stream may follow can be identified, and the meanings of the various data fields can be understood. The comparison result will indicate the matching degree of the data stream with the known protocol in the library, thereby obtaining the semantic fingerprint comparison result.

[0032] Finally, according to the semantic fingerprint comparison result, the protocol understanding degree of the data points is evaluated, and a protocol understanding confidence score is obtained. The score quantifies the system's understanding of the current data stream protocol, with a higher score indicating more accurate identification and analysis of the protocol, and higher data quality and reliability.

[0033] The embodiment can effectively identify and understand complex data protocols from different pre-hospital emergency devices by analyzing the low-level features of multi-device data streams, constructing representative semantic fingerprint vectors, and comparing them with the semantic fingerprint library of known protocols. This method overcomes the problem of unknown or difficult-to-analyze protocols when dealing with diverse and non-standard data sources, laying a solid foundation for subsequent data processing.

[0034] Through the above technical solutions, the embodiment can significantly improve the system's protocol identification ability and understanding depth of multi-device data streams in pre-hospital emergency scenarios. This not only ensures accurate analysis of raw data and avoids data errors or information loss due to protocol misunderstanding, but also provides a high-confidence data source for subsequent time synchronization, data continuity evaluation, standardized processing, and visualization processing, thereby improving the reliability and effectiveness of pre-hospital emergency data acquisition and processing.

[0035] In some embodiments, the step S103 of evaluating the time synchronization accuracy of the data points to obtain a time confidence score can include but is not limited to the following steps: Obtain the local ambient temperature; According to the local ambient temperature, calculate the temperature change rate; According to the temperature change rate, calculate the instantaneous time deviation; According to the instantaneous time deviation, evaluate the deviation stability; According to the instantaneous time deviation and the deviation stability, the time synchronization accuracy of the data points is evaluated to obtain a time confidence score.

[0036] In some embodiments, due to the complex and changeable environment, such as temperature fluctuations, the crystal oscillator frequency of the internal clock of the device may drift, and a simple clock synchronization mechanism may not be able to fully cope with the influence of such external factors on the accuracy of the internal clock of the device, thereby causing the accuracy of time synchronization to decrease, affecting the reliability of subsequent data processing.

[0037] To this end, the local ambient temperature can be obtained. The ambient temperature of the device can be monitored in real time by deploying a temperature sensor near the data acquisition device or integrating it inside the device. The purpose is to obtain the key environmental parameters that affect the accuracy of the internal clock of the device.

[0038] Then, according to the local ambient temperature, the temperature change rate is calculated. The temperature rise or decrease amplitude per unit time can be obtained by differentiating or regression analyzing the continuously collected local ambient temperature data. For example, the difference between the temperature readings of two adjacent time points and the time interval can be calculated, or the short-term change trend of the temperature can be calculated using the moving average method. The purpose is to quantify the dynamic degree of the influence of the environmental temperature on the device clock.

[0039] Then, according to the temperature change rate, the instantaneous time deviation is calculated. Based on the pre-established relationship model between the device clock and the temperature change rate, the amount of instantaneous time drift that the device clock may produce under the current temperature change rate can be calculated. For example, the clock drift characteristic curve of the device under different temperature change rates can be used for table lookup or interpolation calculation. The purpose is to predict the time synchronization error caused by the dynamic change of temperature.

[0040] According to the instantaneous time deviation, the deviation stability is evaluated. The fluctuation of the instantaneous time deviation within a period of time can be analyzed to determine whether it is within an acceptable range or whether there is a persistent drift trend, so as to obtain the deviation stability. For example, the standard deviation, variance or trend analysis of the instantaneous time deviation can be calculated. The purpose is to understand the reliability and predictability of the time deviation.

[0041] Finally, according to the instantaneous time deviation and the deviation stability, the time synchronization accuracy of the data points is evaluated to obtain a time confidence score. The time stamp of each data point can be corrected and assigned a confidence score reflecting its time synchronization accuracy by considering the size and stability of the instantaneous time deviation. For example, the smaller the deviation and the higher the stability, the higher the confidence score. The purpose is to provide more accurate time reference and reliability indication for subsequent data processing.

[0042] To illustrate this technical solution more clearly, a specific example is used below. Suppose that during a pre-hospital emergency medical mission, the ambulance travels through different ambient temperature zones, and a multi-parameter monitor inside the vehicle is collecting the patient's physiological data. This monitor integrates a temperature sensor and is equipped with the time synchronization accuracy assessment module of this embodiment. First, this module continuously acquires the local ambient temperature of the monitor's location. For example, as the ambulance moves from outdoors to indoors or passes through different climate zones, the temperature sensor records temperature increases or decreases. Next, the system calculates the rate of temperature change based on these continuous temperature readings. For example, if the temperature rises from 20°C to 25°C within 5 minutes, the rate of temperature change is calculated to be 1°C per minute.

[0043] Then, based on the temperature characteristic curve of the crystal oscillator inside the monitor, the system calculates the instantaneous time deviation according to the rate of temperature change. For example, it is known that the crystal may produce an additional time drift of 0.5 milliseconds per hour at a heating rate of 1°C per minute. Simultaneously, the system continuously evaluates the stability of this instantaneous time deviation. If the deviation value fluctuates little over a period of time, it is considered to have high stability; if the fluctuation is drastic, the stability is considered low. Finally, the system comprehensively considers the calculated instantaneous time deviation and the evaluated deviation stability to correct the timestamp of each physiological data point collected by the monitor and generates a time confidence score. For example, if the instantaneous time deviation is small and the stability is high, the time confidence score of that data point will be high, indicating that its timestamp is very reliable; conversely, if the deviation is large or the stability is poor, the confidence score will decrease accordingly, indicating that there may be a large error in the time synchronization of that data point. In this way, even in complex scenarios with dynamically changing ambient temperatures, the time synchronization accuracy of multi-device data streams can be ensured, providing emergency personnel with reliable patient physiological data.

[0044] Through the above technical solution, this embodiment can effectively address the impact of ambient temperature changes on the clock synchronization accuracy of devices in pre-hospital emergency scenarios, significantly improving the time synchronization accuracy of multi-device data streams. By considering local ambient temperature and its rate of change, it can more accurately predict and compensate for instantaneous clock drift, thereby making the obtained time confidence scores more realistic and reliable. This not only improves the overall quality of data acquisition but also provides a more solid time foundation for subsequent data analysis, diagnosis, and decision-making, especially in physiological parameter monitoring and event recording that require high-precision time alignment, where its advantages are even more prominent.

[0045] In some embodiments, step S106, which normalizes the data points based on the cumulative confidence score, may include, but is not limited to, the following steps: Obtain key physiological parameters of personnel; According to the physiological parameter change trend, a physiological context correction factor is calculated. According to the cumulative confidence score, a filtering parameter is determined. According to the physiological parameter change trend, a physiological context correction factor is calculated. According to the physiological context correction factor, the filtering parameter is updated. According to the updated filtering parameter, the data points are normalized.

[0046] In some embodiments, since only the confidence score of the data itself is relied on for normalization, the real-time physiological state changes and individual differences of the patient may not be fully considered. If the physiological context of the patient is not taken into account, the normalized data may not fully reflect its clinical authenticity, which may affect the accuracy and effectiveness of subsequent decision support.

[0047] To this end, the key physiological parameters of the person can be obtained. The core physiological indicators of the patient can be collected in real time through various physiological sensors or manual input, such as heart rate, blood pressure, blood oxygen saturation, respiratory rate, body temperature, blood glucose level, etc. These parameters are the basis for evaluating the patient's vital signs and overall health status. According to the key physiological parameters of the person, the physiological parameter change trend is analyzed. Time series analysis can be performed on the obtained key physiological parameters to identify their patterns, rates and directions over time. For example, the instantaneous change rate of heart rate, the fluctuation range of blood pressure, or the long-term downward trend of blood oxygen saturation can be calculated. This analysis aims to capture the dynamic evolution of the patient's physiological state, rather than just focusing on a single instantaneous value.

[0048] Then, according to the cumulative confidence score, a filtering parameter is determined. The cumulative confidence score reflects the quality and reliability of the data itself, and the higher the score, the more reliable the data. The filtering parameter can include but is not limited to the type of filter (such as mean filter, median filter, Kalman filter, etc.), filter window size, cutoff frequency, or smoothing coefficient. When the cumulative confidence score is high, weaker filtering or smaller filter window can be used to preserve more details of the original data; when the score is low, stronger filtering or larger window may be needed to remove noise and outliers.

[0049] According to the physiological parameter change trend, a physiological context correction factor is calculated. The physiological context correction factor is a quantitative value that adjusts the filtering parameter to adapt to the current physiological state and change trend of the patient. For example, if the patient's heart rate is rapidly rising, indicating that they may be in a stress state, the correction factor may indicate that a more sensitive filtering strategy should be used for heart rate data to avoid smoothing out important physiological events. This correction factor aims to integrate the patient's clinical context into the data processing process.

[0050] According to the physiological context correction factor, the filtering parameters are updated. This means that the initial filtering parameters are dynamically adjusted according to the real-time physiological context of the patient. For example, if the physiological context correction factor indicates that the current physiological state fluctuates dramatically, requiring more detailed data, then the filtering parameters can be adjusted to allow more details to pass through; conversely, if the physiological state is stable, then stronger smoothing can be applied.

[0051] Finally, according to the updated filtering parameters, the data points are normalized, aiming to convert data of different sources and different dimensions into a unified format and range, so as to facilitate subsequent analysis and comparison. Through the use of physiological context corrected filtering parameters, the normalization process can more accurately reflect the real physiological condition of the patient, while effectively removing noise and artifacts, ensuring the clinical usability of the data.

[0052] In order to more clearly illustrate the technical solutions, specific examples are used in the following explanation. Suppose a patient's heart rate data stream is collected during pre-hospital emergency care. First, the system will obtain the patient's heart rate, blood pressure, blood oxygen saturation and other key physiological parameters. By analyzing these parameters, it is found that the patient's heart rate shows a rapid upward trend in a short period of time, and the blood pressure has decreased, which may indicate that the patient is in the early stage of stress or shock. At this time, according to the protocol understanding confidence score, time confidence score and continuity confidence score of the heart rate data stream, a moderate to low cumulative confidence score is calculated, and the system preliminarily determines a set of filtering parameters for the normalization of heart rate data, such as a moderate intensity low-pass filter.

[0053] However, due to the analysis of the rapid upward trend of heart rate, the system calculates a physiological context correction factor, which indicates that in the current stress situation, higher sensitivity needs to be maintained for the rapid changes in heart rate to avoid excessive smoothing. Based on this correction factor, the previously determined filtering parameters are updated, such as appropriately increasing the cutoff frequency of the low-pass filter or reducing the filtering window to allow the rapid upward trend of heart rate to be more clearly reflected in the normalized data. Finally, according to the updated filtering parameters, the heart rate data points are normalized, so that the processed heart rate data not only removes noise, but also accurately reflects the patient's physiological stress response, providing more clinically valuable information for emergency personnel.

[0054] Through the above technical solutions, the embodiment can dynamically adjust the filtering strategy by integrating key physiological parameters and physiological parameter change trends, thereby ensuring data quality while better preserving the authenticity and dynamics of the patient's physiological state. Thus, the normalized data can more accurately reflect the patient's clinical condition, providing more reliable and clinically meaningful data support for pre-hospital emergency personnel, significantly improving the practical value and accuracy of data in emergency medical decision-making.

[0055] In some embodiments, in step S107, the visualization of the normalized data points according to the cumulative confidence score can include but is not limited to the following steps: Step S201, obtaining key physiological parameters of the person; Step S202, performing trend analysis on the key physiological parameters of the person to obtain parameter change rate and long-term change direction; Step S203, identifying physiological indicator changes according to the parameter change rate and long-term change direction; Step S204, calculating the impact weight of the physiological indicator changes on the overall condition of the person according to the physiological interaction logic; Step S205, prioritizing the physiological indicator changes according to the impact weight to identify the main focus; Step S206, visualizing the normalized data points according to the main focus and the cumulative confidence score.

[0056] In some embodiments, since the visualization is only based on the confidence score of the data itself, it may not fully reveal the physiological meaning behind the data, especially in pre-hospital emergency scenarios that require quick and accurate judgment of patient condition, lack of deep analysis of key physiological indicator change trends and their impact on the overall condition of the person, which may lead to difficulty for emergency personnel to quickly identify the most urgent or important physiological abnormalities, thereby affecting decision-making efficiency and treatment effect.

[0057] Therefore, the key physiological parameters of the person can be obtained first. The core physiological indicators of the patient can be collected in real time through various physiological sensors or manual input, such as heart rate, blood pressure, blood oxygen saturation, respiratory rate, body temperature, blood glucose level, etc. These parameters are the basis for evaluating the patient's vital signs and overall health status. Trend analysis is performed on the key physiological parameters of the person to obtain parameter change rate and long-term change direction. The parameter change rate refers to the change speed and direction of a certain physiological parameter in a short period of time (e.g. several seconds to several minutes), such as the degree of sudden increase or decrease in heart rate; the long-term change direction refers to the overall trend of a certain physiological parameter in a longer time span (e.g. tens of minutes to hours), such as continuous increase or decrease in blood pressure. This can be calculated by sliding window average, linear regression, exponential smoothing, etc. statistical methods. The purpose is to capture the dynamic changes of physiological state, not just instantaneous values.

[0058] Then, physiological indicator changes are identified based on parameter change rates and long-term change directions. Significant, potentially clinically meaningful changes in a physiological parameter can be determined based on pre-set physiological thresholds, change rate thresholds, or pattern recognition algorithms. For example, a sudden increase in heart rate beyond a certain threshold or a sustained decrease in blood oxygen saturation. The goal is to convert raw numerical changes into physiological indicator change events.

[0059] Next, the impact weight of physiological indicator changes on the overall condition of the person is calculated based on physiological interaction logic. Considering that the human physiological system is a complex network, there are mutual correlations and influences between various physiological indicators. For example, a decrease in blood pressure can lead to a compensatory increase in heart rate. Physiological interaction logic can be built based on medical knowledge bases, clinical guidelines, or machine learning models to assess the potential impact of changes in a certain physiological indicator on the overall physiological stability and prognosis of the patient. For example, the impact weight of a sharp decrease in blood oxygen saturation on the overall condition is usually higher than a slight body temperature fluctuation. The goal is to quantify the importance of different physiological indicator changes.

[0060] Based on the impact weight, physiological indicator changes are prioritized to identify the main focus. All identified physiological indicator changes can be ranked in descending order of their impact weight on the overall condition of the person, thereby determining the physiological abnormalities that need the most attention at the moment. For example, in the presence of multiple physiological indicator abnormalities, the system will prioritize alerting about life-threatening conditions such as hypoxemia or arrhythmia. The goal is to help emergency personnel quickly focus on key issues in a complex information environment.

[0061] Finally, the normalized data points are visualized based on the main focus and cumulative confidence score. On the display interface, those physiological indicators identified as the main focus are presented in a more eye-catching and intuitive way, such as using different colors, flashing effects, larger fonts, or being placed in the center of the screen. At the same time, the cumulative confidence score can be used to adjust the intensity or reliability of these visual elements, for example, high-confidence data points can be displayed in a more certain way, while low-confidence data points may have warning signs indicating that the data may be uncertain. The goal is to provide a comprehensive visualization interface that highlights key information while reflecting data quality.

[0062] To make the technical solution clearer, specific examples are used for explanation below. Assume that a patient is sent to the hospital due to a traffic accident, and the emergency personnel use the multi-device data acquisition system on the scene. The system first acquires the patient's key physiological parameters such as heart rate, blood pressure, blood oxygen saturation, and respiratory rate. The system performs trend analysis on these parameters and finds that the patient's blood pressure is continuously decreasing, the heart rate is continuously increasing, and the blood oxygen saturation is also showing a downward trend. According to the parameter change rate and long-term change direction, the system identifies changes in physiological indicators such as "hypotension", "tachycardia", and "hypoxemia". Further, the system evaluates the impact weight of these physiological indicator changes on the patient's overall condition according to the preset physiological interaction logic. For example, hypoxemia and persistent hypotension are usually given a higher impact weight because they directly threaten life. The system prioritizes the identified physiological indicator changes according to these impact weights and identifies "hypoxemia" and "persistent hypotension" as the current main concerns.

[0063] Finally, on the emergency personnel's display device, when the system visualizes the normalized data points, it will particularly highlight the values and trend graphs of blood oxygen saturation and blood pressure, possibly using red or flashing effects to alert their abnormal state, and possibly displaying decision guidance prompt information such as "severe hypoxemia, oxygen supply immediately" and "blood pressure continuously decreasing, consider shock" in the center of the screen. At the same time, if the data quality of a certain sensor is low, its corresponding cumulative confidence score is low, and the system may display a small warning icon when displaying the data, indicating that the data may have uncertainty. In this way, the emergency personnel can immediately grasp the patient's most critical physiological condition and quickly take appropriate emergency measures according to the priority information provided by the system, such as immediate oxygen therapy and fluid resuscitation, thereby significantly improving emergency efficiency and patient survival rate.

[0064] Through the above technical solution, the present embodiment can significantly improve the intelligence and practicality of pre-hospital emergency multi-device data visualization processing. Instead of just facing raw data or simple confidence prompts, emergency personnel can obtain key information that has been analyzed in a physiological context and prioritized. This helps emergency personnel quickly identify the patient's most critical physiological abnormalities in a tense emergency environment, reducing cognitive load and improving decision-making efficiency and accuracy. In addition, visualization combined with cumulative confidence scores also ensures the reliability of the information presented, avoiding misleading due to low-quality data, thereby providing more timely and accurate treatment for patients.

[0065] In some embodiments, in step S206, visualizing the normalized data points according to the main concerns and the cumulative confidence scores can include but is not limited to the following steps: In step S301, personnel physiological state information, environmental light intensity, and personnel context information are acquired. The personnel physiological state information includes heart rate, pupil diameter, and eye movement trajectory. In step S302, personnel cognitive load is evaluated according to the personnel physiological state information. In step S303, environmental visual clarity is evaluated according to the environmental light intensity. In step S304, display parameters including font size, color contrast, flicker frequency, and information density are determined according to the personnel cognitive load and the environmental visual clarity. In step S305, decision guidance prompt information is generated according to the personnel context information. In step S306, the standardized data points are visualized according to the main focus of attention, the cumulative confidence score, the display parameters, and the decision guidance prompt information.

[0066] In some embodiments, if the visualization process does not adequately consider the real-time physiological state of the operator, environmental conditions, and the current context, it may result in unclear and difficult-to-understand information presentation, even increasing the cognitive load of the operator, thereby affecting their rapid and accurate decision-making.

[0067] To this end, personnel physiological state information, environmental light intensity, and personnel context information can be acquired first. The personnel physiological state information refers to data reflecting the current physiological condition of the operator, including heart rate, pupil diameter, and eye movement trajectory. Heart rate can reflect the operator's tension level or physical exertion; changes in pupil diameter may be related to cognitive load or lighting conditions; eye movement trajectory can reveal the operator's focus of attention and scanning pattern. These information can be acquired through physiological sensors worn or eye tracking systems integrated in the display device. The environmental light intensity refers to the brightness level of the environment where the operator is located, which can be measured in real time by an environmental light sensor. The personnel context information refers to background information related to the current emergency task, such as the severity of the patient's condition, the complexity of the scene, and available medical resources, which can be automatically acquired by the system or manually input by the operator.

[0068] Then, the cognitive load of the person is evaluated according to the personnel physiological state information. The environmental visual clarity is evaluated according to the environmental light intensity, which is used to determine the influence degree of the current light condition on visual information reception. The display parameters, including font size, color contrast, flicker frequency and information density, are determined according to the cognitive load of the person and the environmental visual clarity. The font size can be adjusted according to the cognitive load of the person to ensure the readability of information under different load levels. The color contrast can be optimized according to the environmental visual clarity to adapt to different light conditions. The flicker frequency and information density can be dynamically adjusted according to the cognitive load of the person to avoid information overload or distraction.

[0069] Finally, the decision guidance prompt information is generated according to the personnel situation information, aiming to provide targeted suggestions or next step operation guidance for the operator to assist him in making quick and accurate decisions in complex situations. The standardized data points are visualized according to the main focus, cumulative confidence score, display parameters and decision guidance prompt information.

[0070] In order to more clearly illustrate the technical scheme, specific examples are used for explanation below. Assume that an emergency medical technician is monitoring a patient in a moving ambulance. At this time, the light in the car is dim, and the emergency medical technician is performing a complex operation, resulting in an increased heart rate, dilated pupils, and eye movement trajectory showing high concentration of attention. The system first obtains these personnel physiological state information and evaluates that the cognitive load of the emergency medical technician is high. At the same time, the environmental light sensor detects that the light intensity in the car is low, and the system evaluates that the environmental visual clarity is general. Based on these evaluation results, the system dynamically adjusts the display parameters: for example, the font size on the display screen is increased by 20%, the color contrast is adjusted to high contrast mode (such as black background and white characters), and the flicker frequency and overall information density of non-critical information are reduced to ensure that the core data is still clear and readable under low light and high cognitive load. In addition, according to the critical situation information of the current patient (for example, the patient has arrhythmia), the system generates decision guidance prompt information such as "immediately prepare the defibrillator" or "check the drug dosage", and displays it on the screen in a prominent way to assist the emergency medical technician in rapid response. In this way, the emergency medical technician can more efficiently acquire and understand key information with lower cognitive load and get timely decision assistance in a complex and variable environment.

[0071] By the technical solution, the display parameters are dynamically adjusted to adapt to the cognitive load of the operator and the environmental visual clarity, effectively reducing the cognitive burden of the operator, and improving the readability and understandability of the information. At the same time, the decision guidance prompt information is generated in combination with the personnel situation information, real-time decision support is provided for the operator, and the response speed and decision accuracy of pre-hospital care are significantly improved. This human-centered and situation-aware visualization method greatly optimizes the work experience of pre-hospital care personnel and ultimately improves the overall emergency efficiency and patient treatment success rate.

[0072] In some embodiments, after obtaining the personnel physiological state information in step S301, the method can further include but is not limited to the following steps: Step S401, multi-sensor cross-validation is performed on the personnel physiological state information to obtain a data distortion and artifact recognition result; Step S402, a state change rate is calculated according to the personnel physiological state information; Step S403, a physiological reasonableness judgment is performed on the data distortion and artifact recognition result according to the state change rate to obtain a physiological reasonableness judgment result; Step S404, according to the physiological reasonableness judgment result, the distorted and artifact data in the personnel physiological state information is corrected, and the personnel physiological state information is updated.

[0073] In some embodiments, since the originally collected personnel physiological state information can be affected by various factors, such as sensor noise, environmental interference or physiological artifacts, the data may be distorted or inaccurate. If these data quality problems are not solved, cognitive load assessment and display parameter determination based on inaccurate physiological state information may result in poor visualization effect, and even mislead the decision of the emergency personnel.

[0074] Therefore, the personnel physiological state information can be first cross-validated by multiple sensors to obtain a data distortion and artifact recognition result. Multi-sensor cross-validation refers to comparing and verifying the data collected by different types or different positions of sensors with each other to identify abnormalities, inconsistencies or artifacts in the data. For example, when the heart rate data shows abnormal fluctuations, the blood oxygen saturation or respiratory rate data can be compared, and if other related indicators do not show synchronous abnormalities, the heart rate data may have artifacts. The data distortion and artifact recognition result refers to determining which data points or data segments have abnormalities, such as noise, drift, spikes or artifacts caused by motion, blinking and other non-physiological factors, after cross-validation.

[0075] Then, a state change rate is calculated according to the personnel physiological state information. The state change rate refers to the speed of change of the personnel physiological state information over time, such as the instantaneous change rate of heart rate, the contraction or expansion speed of pupil diameter, and aims to capture the dynamic characteristics of physiological activities. The physiological reasonableness judgment refers to judging whether the identified data distortion and artifact are consistent with the physiological possibility according to the known physiological laws, individual characteristics and current situation. For example, if the heart rate fluctuates sharply within a short time and exceeds the normal physiological range, it may be judged as unreasonable.

[0076] According to the state change rate, a physiological reasonableness judgment is made on the data distortion and artifact identification result to obtain a physiological reasonableness judgment result. The physiological reasonableness judgment result refers to the further confirmation or correction of the data distortion and artifact identification result after the physiological reasonableness judgment. In actual application, the correction of the distorted and artifact data specifically refers to the repair or replacement of the distorted or artifact data identified and judged as unreasonable, for example, interpolation, smoothing, filtering or model-based prediction methods can be used for correction, and the purpose is to restore the authenticity of the data.

[0077] Finally, according to the physiological reasonableness judgment result, the distorted and artifact data in the personnel physiological state information is corrected, and the personnel physiological state information is updated. Updating the personnel physiological state information refers to replacing the original data with the corrected data to form more accurate and reliable personnel physiological state information.

[0078] In order to more clearly illustrate the technical scheme, specific examples are used for explanation below. Assuming that in a pre-hospital emergency scene, a patient wears a heart rate monitor, a pupil diameter sensor and an eye tracking device. After obtaining these personnel physiological state information, first, the heart rate, pupil diameter and eye movement trajectory data are cross-verified by multiple sensors. For example, if an abnormally high spike suddenly appears in the heart rate data, but the pupil diameter and eye movement trajectory data do not show a physiological stress response (such as pupil dilation, increased eye movement frequency) related to the sharp heart rate change, the heart rate spike may be initially identified as an artifact. Then, the state change rates of heart rate, pupil diameter and eye movement trajectory are calculated. For example, the amplitude of change of heart rate within a short time is analyzed. Then, the identified heart rate spike is compared with the patient's historical physiological baseline data and the current emergency situation (such as the patient being moved, which may cause the sensor to shake), and a physiological reasonableness judgment is made. If it is judged that the spike is not reasonable in physiology and is consistent with the sensor shaking pattern, it is confirmed as an artifact. Finally, the heart rate spike data is corrected, for example, linear interpolation or median filtering of the data before and after is used for smoothing, so as to update the personnel physiological state information, and ensure that high-quality, undistorted physiological data is used for subsequent assessment of cognitive load by the evaluator.

[0079] By the technical solution, the embodiment can significantly improve the accuracy and reliability of the obtained personnel physiological state information, avoid deviation of subsequent cognitive load evaluation and visualization processing caused by original data quality problems. This makes the decision guidance prompt information based on physiological state information more accurate, thereby improving the effectiveness of information presentation and the reliability of decision support in the pre-hospital emergency scene. Especially in the high-pressure, complex and time-sensitive emergency environment, ensuring the authenticity of physiological data is crucial for accurately judging the patient's condition and developing a reasonable treatment plan.

[0080] In some embodiments, the multi-sensor cross-validation of the personnel physiological state information in step S401 to obtain the data distortion and artifact identification result can include but is not limited to the following steps: Obtaining physiological sensor raw data; According to the physiological sensor raw data, calculating the data change rate, data fluctuation range and data correlation; According to the data change rate and the data fluctuation range, identifying the noise characteristics and artifact patterns of the physiological sensor; According to the data correlation, evaluating the interaction between physiological indicators; According to the noise characteristics, artifact patterns and interaction between physiological indicators, the multi-sensor cross-validation of the personnel physiological state information is performed to obtain the data distortion and artifact identification result.

[0081] In some embodiments, the physiological sensor raw data can be obtained first. The untreated raw signal stream can be directly collected from various physiological sensors connected to the patient, such as electrocardiogram sensors, blood oxygen saturation sensors, blood pressure sensors, body temperature sensors, eye movement tracking sensors, etc. These raw data are the basis for subsequent analysis, and their quality directly affects the final identification result.

[0082] Then according to the physiological sensor raw data, the data change rate, data fluctuation range and data correlation are calculated, aiming to quantify the dynamic characteristics and mutual relationship of the data from different dimensions. The data change rate refers to the speed of change of the physiological parameter value with time within a certain time window, for example, by calculating the difference between adjacent data points or using the linear regression slope in the sliding window. The data fluctuation range refers to the difference between the maximum and minimum values of the physiological parameter value within a certain period of time, or the dispersion degree of the data is represented by statistical quantities such as standard deviation, variance. The data correlation refers to the degree of mutual dependence between different physiological sensor data in statistics, which can be quantified by Pearson correlation coefficient or mutual information, etc.

[0083] According to the data change rate and data fluctuation range, the noise characteristics and artifact patterns of the physiological sensor are identified. For example, when the data change rate is abnormally high and the fluctuation range exceeds the normal physiological threshold, it may indicate the presence of high-frequency noise or motion artifacts; when the data fluctuation range is too small or shows periodic regularity, it may indicate the presence of baseline drift or power interference noise. Through the preset threshold or machine learning model, these features can be pattern matched to identify specific noise types and artifact patterns.

[0084] According to the data correlation, the interaction between physiological indicators is evaluated. For example, under normal physiological conditions, there is a certain negative correlation between heart rate and blood pressure, and there is a coordinated change between blood oxygen saturation and respiratory rate. By analyzing whether the correlation between these physiological indicators conforms to the known physiological laws, it can be determined whether there is an abnormal sensor data that affects the overall physiological interaction pattern.

[0085] Finally, according to the noise characteristics, artifact patterns and interactions between physiological indicators, multi-sensor cross-validation of personnel physiological state information is performed to obtain data distortion and artifact recognition results. This process comprehensively utilizes the data characteristics (noise, artifacts) within a single sensor and the synergistic relationship (interaction) between multiple sensors. When the original data of a sensor exhibits specific noise characteristics or artifact patterns, and its data interaction with other related physiological indicators is also abnormal, it can be more reliably determined that the data of the sensor is distorted or has artifacts.

[0086] The embodiment first acquires the original data of the physiological sensor and performs multi-dimensional feature calculation, including data change rate, data fluctuation range and data correlation, thereby providing a quantitative basis for subsequent distortion and artifact recognition. Accurate calculation of these basic features enables the capture of data anomaly clues from a microscopic level. Subsequently, by analyzing the data change rate and fluctuation range, the noise characteristics and artifact patterns of a single physiological sensor can be effectively identified, which solves the problem of local anomalies that may exist in single sensor data. At the same time, by evaluating the interaction between physiological indicators, the embodiment can verify the logical consistency between different physiological parameters from a macroscopic level, thereby discovering systematic data distortion that is difficult to detect at the single sensor level. Finally, by comprehensively utilizing these noise characteristics, artifact patterns and interaction information between physiological indicators, the embodiment can perform multi-sensor cross-validation to ensure comprehensive and accurate evaluation of personnel physiological state information, effectively identifying and locating data distortion and artifacts.

[0087] By the technical solution, the embodiment introduces multiple dimensions such as data change rate, data fluctuation range and data correlation for analysis, and cross-verification is performed in combination with the interaction between physiological indexes, thereby significantly improving the accuracy and robustness of recognition. This helps to discover and correct damaged physiological data earlier and more accurately in a pre-hospital emergency scene, avoid misjudgment caused by data quality problems, and provide more reliable patient physiological state information for emergency personnel, support more timely and accurate treatment decisions, and ultimately improve the efficiency and success rate of pre-hospital emergency.

[0088] In some embodiments, in step S403, the physiological reasonableness judgment result of the data distortion and artifact recognition result is obtained according to the state change rate, which can include but is not limited to the following steps: Obtaining personnel personalized historical baseline data, emergency degree and operation complexity of the current task; Matching the state change rate with the physiological stress response pattern in the personnel personalized historical baseline data to obtain a matching degree; According to the matching degree, the emergency degree and the operation complexity, the physiological reasonableness judgment result of the data distortion and artifact recognition result is obtained.

[0089] In some embodiments, since the physiological response of personnel has high individual difference, and the physiological state is significantly affected by situational factors such as the emergency degree and the operation complexity of the current task. If only a single state change rate is used for judgment, these personalized and situational factors may not be fully considered, thereby limiting the accuracy and reliability of the judgment result, and misjudgment or omission may occur, affecting the effectiveness of subsequent data correction and visualization processing.

[0090] To this end, personnel individualized historical baseline data, urgency of the current task, and operation complexity can be obtained first. Historical physiological parameter data related to the target personnel can be stored in advance, such as the normal range, fluctuation characteristics of physiological indicators such as heart rate, blood pressure, respiratory rate, pupil diameter of the personnel in different physiological states (such as rest, light activity, moderate stress, etc.), and typical physiological stress response patterns under certain stimuli. These data are used to construct personnel individualized historical baseline data unique to the personnel, which aims to provide personalized reference for subsequent physiological reasonableness judgment. The urgency of the current task refers to the criticality of the pre-hospital emergency task, which can be classified into general, urgent, critical, etc. according to factors such as patient's vital signs, injury severity, consciousness state, etc., which aims to reflect the influence strength of the current situation on the personnel's physiological state. The operation complexity refers to the difficulty and delicacy of the current first aid operation, such as the number of operation steps, required skill level, environmental restrictions, etc., which aims to evaluate the impact of the operation itself on the physiological load of the emergency personnel.

[0091] Then the state change rate is matched with the physiological stress response pattern in the personnel individualized historical baseline data to obtain a matching degree. The currently monitored physiological state change rate can be compared with the physiological stress response pattern recorded in the historical data of the personnel under similar situations or stimuli. For example, pattern recognition algorithms, machine learning models or statistical methods can be used to calculate the similarity or deviation between the current state change and the historical pattern, so as to obtain a quantitative matching degree score, which aims to evaluate whether the current physiological change conforms to the typical physiological response rule of the personnel in a specific situation.

[0092] According to the matching degree, the urgency and the operation complexity, the physiological reasonableness judgment of the data distortion and artifact recognition result is performed to obtain the physiological reasonableness judgment result. The identified data distortion and artifact result can be finally verified for physiological reasonableness. For example, if the matching degree is high (i.e. the current change conforms to the typical stress pattern of the personnel), and the current task is high in urgency and complexity of operation (these situations themselves can cause significant physiological changes), even if the state change rate is large, it can be judged as physiological reasonableness, so that the previously identified distortion and artifact are considered to be not real artifacts, but real physiological response. On the contrary, if the matching degree is low, and the situation factors are not enough to explain the current change, it is more inclined to judge as data distortion or artifact. The purpose is to improve the accuracy and robustness of the judgment, and to avoid misjudging real physiological stress response as data anomaly, or misjudging real abnormal data as physiological reasonableness.

[0093] To make the technical solution clearer, specific examples are used for explanation below. Assume that an emergency medical technician is performing chest compression operation on a patient with cardiac arrest. In the basic scheme, if the heart rate of the technician suddenly rises from 80 beats per minute to 140 beats per minute, the system may identify it as data distortion or artifact due to the too fast heart rate change rate. However, the present embodiment first acquires the personalized historical baseline data of the technician, which may contain the physiological stress response pattern of the technician when performing high-intensity physical activity (such as intense exercise or simulated emergency operation), in which the heart rate can quickly rise to 150 beats per minute. At the same time, the system identifies that the current task is "cardiac arrest rescue", the criticality of which is "critical" and the operation complexity is "high" (which requires continuous high-intensity chest compression). At this time, the system matches the current heart rate change rate (from 80 beats per minute to 140 beats per minute) with the physiological stress response pattern of the technician in history, and finds that the matching degree is high. Combined with the criticality of the current task and the high operation complexity, the system will judge that the heart rate surge is a physiological reasonable stress response, rather than data distortion or artifact. Thus, the heart rate data will be regarded as valid data and used for subsequent physiological state assessment and decision guidance, avoiding important physiological information being ignored due to misjudgment.

[0094] Through the above technical solution, the present embodiment can significantly improve the intelligentization and personalization level of data quality assessment in the process of pre-hospital emergency multi-device data acquisition. The scheme can more accurately distinguish between real physiological changes and data distortion or artifact, reducing false positives and false negatives, thereby providing more reliable input for subsequent data correction, standardized processing and visualization processing. Thus, the physiological data relied on by the emergency medical technician can be more real and effective, thereby supporting more accurate clinical decision-making and more efficient emergency intervention, ultimately improving the overall efficiency of pre-hospital emergency and the success rate of patient treatment.

[0095] The beneficial effects of implementing the embodiments of the present application include that the embodiments of the present application first acquire a multi-device data stream, then perform protocol understanding degree assessment on the data points to obtain a protocol understanding confidence score, perform time synchronization accuracy assessment to obtain a time confidence score, and perform data stream continuity assessment to obtain a continuity confidence score, then calculate a cumulative confidence score of the data points according to the protocol understanding confidence score, the time confidence score and the continuity confidence score, finally perform standardized processing on the data points and perform visualization processing on the standardized processed data points, so that the data points can be standardized processed in combination with multiple confidence scores, to realize multi-device data acquisition, and improve accuracy and reliability.

[0096] As shown in Figure 2 The present embodiment further provides a pre-hospital emergency multi-device data acquisition system, which comprises: The data acquisition module 501 is configured to acquire a multi-device data stream, the multi-device data stream comprising a plurality of data points. The protocol understanding evaluation module 502 is configured to perform protocol understanding degree evaluation on the data points according to the multi-device data stream, to obtain a protocol understanding confidence score. The time synchronization evaluation module 503 is configured to perform time synchronization accuracy evaluation on the data points, to obtain a time confidence score. The data stream continuity evaluation module 504 is configured to perform data stream continuity evaluation on the data points, to obtain a continuity confidence score. The confidence score calculation module 505 is configured to calculate a cumulative confidence score of the data points according to the protocol understanding confidence score, the time confidence score and the continuity confidence score. The normalization processing module 506 is configured to perform normalization processing on the data points according to the cumulative confidence score. The visualization processing module 507 is configured to perform visualization processing on the normalized data points according to the cumulative confidence score.

[0097] The content in the method embodiments is applicable to the system embodiments, the system embodiments specifically implement the same functions as the method embodiments, and achieve the same beneficial effects as the method embodiments.

[0098] The embodiments described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

Claims

1. A method for collecting data from multiple devices in pre-hospital emergency care, characterized in that, Includes the following steps: Acquire multi-device data streams, wherein the multi-device data streams include multiple data points; Based on the multi-device data stream, the protocol understanding level of the data points is evaluated to obtain a protocol understanding confidence score; The time synchronization accuracy of the data points is evaluated to obtain a time confidence score; The data points are evaluated for data flow continuity to obtain a continuity confidence score; Based on the protocol, the confidence score, the time confidence score, and the continuity confidence score are understood, and the cumulative confidence score of the data point is calculated. The data points are normalized based on the cumulative confidence scores. Based on the cumulative confidence score, the normalized data points are visualized.

2. The method according to claim 1, characterized in that, The step of evaluating the protocol understanding level of the data points based on the multi-device data stream to obtain a protocol understanding confidence score includes: Low-level feature calculations are performed on the multi-device data stream to obtain a low-level feature statistical distribution, which includes byte entropy value, short sequence hash value, and frequency of occurrence of target byte pairs; Based on the statistical distribution of the low-level features, a semantic fingerprint vector is generated; The semantic fingerprint vector is compared with the semantic fingerprint database, and the semantics of the data fields are analyzed to obtain the semantic fingerprint comparison result; Based on the semantic fingerprint comparison results, the protocol understanding level of the data points is evaluated to obtain a protocol understanding confidence score.

3. The method according to claim 1, characterized in that, The time synchronization accuracy assessment of the data points, to obtain a time confidence score, includes: Obtain the local ambient temperature; Calculate the rate of temperature change based on the local ambient temperature; Calculate the instantaneous time deviation based on the stated rate of temperature change; The stability of the deviation is evaluated based on the instantaneous time deviation. Based on the instantaneous time deviation and the deviation stability, the time synchronization accuracy of the data points is evaluated to obtain the time confidence score.

4. The method according to claim 1, characterized in that, The normalization process for the data points based on the cumulative confidence score includes: Obtain key physiological parameters of personnel; Based on the key physiological parameters of the personnel, analyze the trends of physiological parameter changes; The filtering parameters are determined based on the cumulative confidence score. Calculate the physiological situation correction factor based on the changing trends of the physiological parameters; The filter parameters are updated based on the physiological context correction factor. The data points are normalized based on the updated filtering parameters.

5. The method according to claim 1, characterized in that, The step of visualizing the normalized data points based on the cumulative confidence score includes: Obtain key physiological parameters of personnel; Trend analysis was performed on the key physiological parameters of the personnel to obtain the rate of change and long-term direction of change of the parameters; Based on the rate of change of the parameters and the direction of long-term change, identify changes in physiological indicators; Based on the physiological interaction logic, the weight of the impact of the changes in the physiological indicators on the overall condition of the personnel is calculated. Based on the influence weights, the changes in the physiological indicators are prioritized to identify the main points of concern; Based on the main concerns and the cumulative confidence scores, the normalized data points are visualized.

6. The method according to claim 5, characterized in that, The visualization process for the normalized data points based on the main points of interest and the cumulative confidence score includes: Acquire information on the physiological state of the personnel, ambient light intensity, and personnel context information, wherein the physiological state information includes heart rate, pupil diameter, and eye movement trajectory; Assess the cognitive load of individuals based on their physiological state information; Assess the visual clarity of the environment based on the ambient light intensity. Based on the cognitive load of the person and the visual clarity of the environment, the display parameters are determined, including font size, color contrast, flicker frequency, and information density. Based on the personnel context information, decision guidance prompts are generated; Based on the main focus, the cumulative confidence score, the display parameters, and the decision guidance prompts, the normalized data points are visualized.

7. The method according to claim 6, characterized in that, After obtaining the physiological state information of the personnel, the method further includes: Multi-sensor cross-validation was performed on the physiological state information of the personnel to obtain the data distortion and artifact recognition results; Calculate the rate of change of state based on the physiological state information of the personnel; Based on the rate of state change, the physiological rationality of the data distortion and artifact recognition results is judged to obtain the physiological rationality judgment result; Based on the physiological rationality judgment results, the distortion and artifact data in the personnel physiological state information are corrected, and the personnel physiological state information is updated.

8. The method according to claim 7, characterized in that, The process of performing multi-sensor cross-validation on the physiological state information of the personnel to obtain data distortion and artifact recognition results includes: Acquire raw data from physiological sensors; Based on the raw data from the physiological sensors, calculate the data change rate, data fluctuation range, and data correlation. Based on the data change rate and the data fluctuation range, identify the noise characteristics and artifact patterns of the physiological sensor; Based on the correlation of the data, assess the interactions between physiological indicators; Based on the interaction between the noise characteristics, the artifact patterns, and the physiological indicators, the physiological state information of the personnel is cross-validated using multiple sensors to obtain data distortion and artifact recognition results.

9. The method according to claim 7, characterized in that, The step of making a physiological rationality judgment on the data distortion and artifact recognition results based on the state change rate, and obtaining the physiological rationality judgment result, includes: Acquire personalized historical baseline data of personnel, the urgency of the current task, and the complexity of the operation; The rate of state change is matched with the physiological stress response pattern in the individual's personalized historical baseline data to obtain the matching degree; Based on the matching degree, the urgency level, and the operational complexity, the physiological rationality of the data distortion and artifact recognition results is judged to obtain the physiological rationality judgment result.

10. A pre-hospital emergency multi-device data acquisition system, characterized in that, include: The data acquisition module is used to acquire multi-device data streams, wherein the multi-device data streams include multiple data points; The protocol understanding assessment module is used to assess the degree of protocol understanding of the data points based on the multi-device data stream, and obtain a protocol understanding confidence score. The time synchronization evaluation module is used to evaluate the time synchronization accuracy of the data points and obtain a time confidence score. The data flow continuity assessment module is used to assess the data flow continuity of the data points and obtain a continuity confidence score. The confidence score calculation module is used to understand the confidence score, the time confidence score, and the continuity confidence score according to the protocol, and to calculate the cumulative confidence score of the data point. The normalization processing module is used to normalize the data points based on the cumulative confidence score; The visualization processing module is used to visualize the normalized data points based on the cumulative confidence score.

Images