Method for collecting and analyzing full information of medical record data based on ICU robot

Abstract

The present application relates to medical data analysis technical field, specifically to the medical record data full information acquisition synchronization and analysis method based on ICU robot, including the following steps: obtaining multiple monitoring and device parameters and corresponding advancing according to time, combining cross-device running state and monitoring project pairing, supplementing key parameters such as oxygen partial pressure in trend comparison, forming continuous chain of respiratory cycle and systemic state, in the present application, by corresponding multiple source monitoring and testing data according to time in the acquisition process, keeping the consistency of differentiated source information in the advancement, through the separation and reorganization of cross-device data, physiological changes and running state are synchronously displayed, missing parameters are supplemented in trend comparison, the continuous link of respiratory cycle and systemic state is constructed, and medical orders, testing and imaging data are embedded in the data chain according to time, so that text and image are accumulated synchronously with physiological information, forming multidimensional information archives, enhancing the coherence and traceability of disease analysis.

Description

Technical Field

[0001] This invention relates to the field of medical data analysis technology, and in particular to a method for the full information collection, synchronization and analysis of medical record data based on an ICU robot. Background Technology

[0002] The field of medical data analytics involves the collection, integration, storage, and analysis of multi-source data related to medical processes and patient health status. Core aspects include medical record information acquisition, clinical test data integration, imaging data analysis, physiological signal interpretation, and simultaneous processing of multiple data sources. Through information technology, it achieves unified management and structured presentation of different types of medical data, providing usable foundational data support for medical decision-making, scientific research analysis, and health management. This technical field covers the entire process from collecting multi-dimensional data from medical institutions, standardizing processing on a unified platform, to logically linking and multi-dimensionally statistically analyzing the data, involving aspects such as interface integration with hospital information systems, standardized data coding, and cross-system data synchronization. The traditional method for collecting, synchronizing, and analyzing full information of medical record data based on ICU robots refers to using mobile or fixed ICU robots in an intensive care environment to acquire real-time patient medical record data from multiple information sources, such as monitoring equipment, testing equipment, and medical staff input terminals. The collected data is then synchronized and converted in time on a unified information processing platform. Subsequently, structured data parsing is used to perform correlation analysis on various types of information, such as medical records, physiological parameters, and test results. This is generally accomplished through means such as equipment interface acquisition, communication protocol parsing, and data standard mapping.

[0003] In the process of collecting and processing multi-source medical records and monitoring data, existing technologies rely on the synchronization accuracy between devices and systems to ensure the time correspondence and sequential connection of different data sources. When collecting data across devices, there may be inconsistencies in timestamps or missing data entries, which can easily cause the monitoring information and test results to lose their correspondence during analysis. Some physiological changes cannot be recorded continuously, and unstructured information such as medical orders and images are difficult to integrate synchronously with physiological data. In the judgment of a condition that requires the support of multi-dimensional information, this can lead to incomplete trend judgments and interruptions in the clinical reference chain. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for the full information collection, synchronization, and analysis of medical record data based on an ICU robot, comprising the following steps: To achieve the above objectives, the present invention adopts the following technical solution: a method for the full information collection, synchronization, and analysis of medical record data based on an ICU robot, comprising the following steps: S1: Acquire heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation from ICU monitors, tidal volume and oxygen concentration from ICU ventilators, and flow rate from ICU micro-infusion pumps. Identify the data source devices, extract data entries at the same time points, analyze whether the acquisition time is consistent, and advance the unbiased data into the process in sequence to obtain the vital signs time chain. S2: Based on the vital signs time chain, call the bedside blood purification rate, ECMO blood flow and oxygenation ratio, split the data of the same time period according to the equipment type, analyze the relationship between the collection time and the sodium, potassium and chloride ion detection time, and advance the corresponding time data in the original order to obtain the cross-equipment operation status time chain. S3: Based on the cross-device operation status time chain, call the ICU monitor's blood oxygen saturation, mean arterial pressure, central venous pressure and bedside lactate concentration, pair the monitoring items and operation status, and push the paired data to the node to obtain the monitoring parameter chain; S4: Based on the monitoring parameter chain, call the values ​​of tidal volume and respiratory rate at the same time point, compare the trend direction, extract the oxygen partial pressure content according to the data with the same direction corresponding to the time position, and add it to the corresponding position in time order to obtain the respiratory and circulatory monitoring chain.

[0005] As a further aspect of the present invention, the vital signs timeline includes heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation, tidal volume, oxygen concentration, and microinfusion pump flow rate; the cross-device operation status timeline includes bedside blood purification rate, ECMO blood flow rate, oxygenation ratio, sodium ion concentration, potassium ion concentration, and chloride ion concentration; the monitoring parameter chain includes blood oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration; and the respiratory and circulatory monitoring chain includes tidal volume, respiratory rate, and partial pressure of oxygen.

[0006] As a further aspect of the present invention, the time point data entry refers to comparing parameters from different devices according to timestamps during the multi-device monitoring process in the ICU, filtering data items with completely consistent collection times, and sequentially advancing them to the processing stage at the same time position. The aforementioned detection time relationship refers to the process of comparing the electrolyte detection time with the bedside blood purification rate, ECMO blood flow, and oxygenation ratio device acquisition time during cross-device data processing, analyzing whether the timestamp differences are within a preset threshold, and then advancing the detection data to the corresponding time series position.

[0007] As a further aspect of the present invention, the method further includes: in the processing of monitoring parameters, corresponding the mean arterial pressure with the data of blood oxygen saturation, central venous pressure, and lactate concentration at the same time position, and comparing and analyzing the trend direction with the ventilation data; The oxygen partial pressure content refers to the oxygen partial pressure value extracted from the blood gas analysis results during the respiratory and circulatory monitoring process, which is exactly the same as the ventilation data timestamp, and inserted into the corresponding monitoring data position in chronological order.

[0008] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Acquire the heart rate, systolic blood pressure, diastolic blood pressure, and blood oxygen saturation of the ICU monitor, as well as the tidal volume, oxygen concentration, and flow rate of the micro-infusion pump of the ICU ventilator. Identify the data source, corresponding device type and device number, compare with the attached timestamp information, select data items with consistent time content, and obtain the time synchronization monitoring dataset. S102: Based on the time-synchronized monitoring dataset, extract the values ​​of the device data at the same point in time, advance them into the data stream according to the order of data collection, and advance each item according to time to obtain a continuous monitoring sequence; S103: Based on the continuous monitoring sequence, each monitoring data in the sequence is sequentially connected in time. According to the time of the differentiated equipment data corresponding to the progress rhythm, the values ​​are pushed into the time structure channel in sequence to obtain the vital signs time chain.

[0009] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Based on the vital signs time chain, call the data content of bedside blood purification rate, ECMO blood flow and oxygenation ratio, extract the values ​​of the source devices in the same time period, separate the data corresponding to each type of device according to the device number, extract the data items of the device, and obtain the device classification monitoring data. S202: Based on the equipment classification monitoring data, obtain the detection time of sodium, potassium and chloride ions, the corresponding equipment data acquisition time, filter data items with consistent time, extract the numerical content within the same time point, and obtain the time-corresponding electrolyte data sequence. S203: Based on the time-corresponding electrolyte data sequence, according to the order of the data items in the vital signs time chain, advance each group of data to the corresponding position in the original sequence according to the time point, continue the data structure within the timeline, and obtain the cross-device operating status time chain.

[0010] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the cross-device operation status time chain, extract the numerical content of the monitoring items within the same time period, filter the data items with consistent time points, and match each data item with the source device to obtain a time node paired dataset. S302: Based on the time node paired dataset, at each corresponding time position, the successfully paired data is advanced according to the original order of the running status time chain, and the monitoring items are connected to the next monitoring item in chronological order to obtain the monitoring parameter chain.

[0011] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Based on the monitoring parameter chain, call the tidal volume and respiratory rate values ​​at the same time point, extract the time period and direction of change of each data, compare the trend direction of respiratory data with the trend direction of blood oxygen saturation, mean arterial pressure, central venous pressure and lactate concentration at the corresponding time point, analyze the trend of change by comparing the values ​​at the current time point with those at the previous time point, and filter out tidal volume and respiratory rate with the same direction to obtain a ventilation data sequence with consistent trend. S402: Based on the trend-consistent ventilation data sequence, according to the time point content, the tidal volume and respiratory rate values ​​are respectively mapped to the time point of the monitoring parameters, and the data are connected to the corresponding positions according to the time relationship to obtain the ventilation-related monitoring sequence. S403: Based on the ventilation-related monitoring sequence, call the oxygen partial pressure value in the blood gas analysis, extract the oxygen partial pressure content that is consistent with the ventilation data time, insert the oxygen partial pressure data into each group of data in chronological order, and advance it into the overall structure to obtain the respiratory and circulatory monitoring chain.

[0012] As a further aspect of the present invention, the method further includes: S5: Based on the respiratory and circulatory monitoring chain, call up medical order information, test results and image data with the same timestamp, compare the time and physiological data nodes of each type of information, advance the text and image content to the corresponding position according to time, and attach it to the existing data to obtain the medical record data analysis file; The medical record data analysis archive includes medical orders, test results, and imaging data.

[0013] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Based on the respiratory and circulatory monitoring chain, call up the medical order information, test results and image data within the same time period, extract the time field corresponding to each type of information, compare it with the time point of physiological monitoring data, filter the data items with the same time content, and obtain the time-corresponding non-physiological information set. S502: Based on the time-corresponding non-physiological information set, according to the time order of each type of data, the text information and image data are respectively mapped to the adjacent positions of the time point where the monitoring data is located, and the data is mapped on the time line to obtain the sequence correspondence information structure; S503: Based on the sequence correspondence information structure, extract the medical orders, test items and imaging data that are not at the same time point, and append them one by one to the end of the monitoring sequence in chronological order, advancing the information into a continuous structure to obtain a medical record data analysis archive.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by corresponding multi-source monitoring and test data according to time during the collection process, the consistency of information from different sources is maintained in the process. Through the separation and recombination of cross-device data, physiological changes and operating status are displayed synchronously. Missing parameters are supplemented in trend comparison, and a continuous link between respiratory circulation and overall body status is constructed. Furthermore, medical orders, test results, and imaging data are embedded into the data chain according to time, so that text and images are accumulated synchronously with physiological information, forming a multi-dimensional information archive, which enhances the coherence and traceability of disease analysis. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a method for the full information collection, synchronization, and analysis of medical record data based on an ICU robot, including the following steps: S1: Acquire heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation from ICU monitors, tidal volume and oxygen concentration from ICU ventilators, and flow rate from ICU micro-infusion pumps. Identify the source devices corresponding to the data, extract data items at the same time point, analyze whether the acquisition time is completely consistent, connect data with no deviation in acquisition time according to the predetermined order between devices, and advance each data item to the continuous processing flow in sequence to obtain the vital signs time chain. S2: Based on the vital signs time chain, the bedside blood purification rate, ECMO blood flow and oxygenation ratio are called. The data obtained by the source devices in the same time period are split according to the device type. The collection time is analyzed to see whether the collection time is consistent with the sodium, potassium and chloride ion detection time. The data items with related time points are advanced according to their order in the original data sequence to obtain the cross-device operation status time chain. S3: Based on the cross-device operation status time chain, call the blood oxygen saturation, mean arterial pressure, central venous pressure and bedside lactate concentration of the ICU monitor, match the monitoring item values ​​in the corresponding time period with the arranged operation status, and place the data of each successfully matched item into the next node of the chain structure according to the original time order. According to the continuous time structure, the monitoring parameter chain is obtained. S4: Based on the monitoring parameter chain, call the values ​​of tidal volume and respiratory rate at the same time point, compare the trend direction of each item with the trend direction of the corresponding monitoring parameter, and match the monitoring values ​​with the same direction to the parallel positions in chronological order. Extract the oxygen partial pressure content from the blood gas analysis according to the same time point, and add the oxygen partial pressure values ​​to the aforementioned position sequence in sequence to obtain the respiratory and circulatory monitoring chain. S5: Based on the respiratory and circulatory monitoring chain, call up medical orders, test results and imaging data with the same timestamp, compare the time field of each type of information with the corresponding time point of physiological data, and advance the text information and image content to the node in sequence along the time order. Add non-physiological data to the end of the monitoring data sequence one by one to obtain the medical record data analysis file.

[0023] The vital signs timeline includes heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation, tidal volume, oxygen concentration, and microinfusion pump flow rate. The cross-device operation timeline includes bedside blood purification rate, ECMO blood flow, oxygenation ratio, sodium ion concentration, potassium ion concentration, and chloride ion concentration. The monitoring parameter chain includes blood oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration. The respiratory and circulatory monitoring chain includes tidal volume, respiratory rate, and partial pressure of oxygen. The medical record data analysis archive includes medical orders, test results, and imaging data.

[0024] Please see Figure 2 The specific steps of S1 are as follows: S101: Acquire the heart rate, systolic blood pressure, diastolic blood pressure, and blood oxygen saturation of the ICU monitor, as well as the tidal volume, oxygen concentration, and flow rate of the micro-infusion pump of the ICU ventilator. Identify the data source, corresponding device type and device number, compare with the attached timestamp information, select data items with consistent time content, and obtain the time synchronization monitoring dataset. First, extract the raw data files from the real-time output interfaces of each monitoring device, and identify the physical meaning of each value and its data source device. During this process, each monitoring parameter should be matched with its device number to clarify its device function type and the monitoring dimension it belongs to. For example, heart rate values ​​come from the ECG channel of the monitor, systolic and diastolic blood pressure are detected by non-invasive or invasive blood pressure modules, blood oxygen saturation is obtained through a pulse oximeter, while tidal volume and oxygen concentration are usually recorded by the ventilator at the end of expiration or the beginning of inspiration, and the flow rate of the micro-infusion pump is output in milliliters per minute by its flow control unit. To ensure consistency in subsequent data processing, all raw data needs to be extracted into a processing list in a unified data format according to the monitoring items. Then, read the time tags attached to each data item and extract the time field from the data. After extraction, all time fields need to be standardized. The process involves several steps, such as unifying hours, minutes, and seconds into numerical values ​​at the same time granularity to avoid matching failures caused by inconsistent formats. Then, each monitoring data point from each device is compared line by line to determine if there are any precisely identical time points. For this type of comparison, entire rows of values ​​should be used for screening. For example, if the heart rate is 84 at 08:35:12, the tidal volume of the ventilator is 520 at 08:35:12, and the flow rate of the microinfusion pump is 4.0 at 08:35:12, then these three data points can be considered to meet the time consistency condition and thus included in the time synchronization range. Fuzzy interval matching should not be used in the time consistency judgment. For example, if 08:35:12 and 08:35:13 do not constitute a consistent relationship, records that do not meet the complete matching condition should be directly removed. After obtaining multiple sets of records that meet the consistency condition, the time synchronization monitoring dataset is obtained.

[0025] S102: Based on the time-synchronized monitoring dataset, extract the values ​​of the equipment data at the same point in time, advance them into the data stream according to the order of data collection, and advance each item according to time to obtain a continuous monitoring sequence; First, extract the data from each time point separately according to the device number. For example, extract heart rate of 84 beats per minute, systolic blood pressure of 128 mmHg, tidal volume of 520 ml, oxygen concentration of 40%, and infusion pump flow rate of 4.0 ml / min. Use these values ​​as the joint observation values ​​for that time point. To ensure the consistency of the data order, the data should be arranged item by item according to the original acquisition order. The acquisition order can be determined by the device data writing order or acquisition number. When the acquisition order information is missing, the data receiving order can be used for time alignment. For example, after the monitor data is written to the ECG channel, it should be immediately connected to the blood pressure channel, and then the ventilator tidal volume data should be connected. In this case, it can be assumed that heart rate precedes tidal volume, and tidal volume precedes infusion pump data. To avoid data overwriting or misalignment due to chaotic acquisition order, the sequence number or channel index should be used to limit the order of acquisition. Then, using time as a guide, the values ​​at the same time are sequentially pushed into the data stream to form continuous observations on the timeline. If the next data record has the same time, the above extraction and sorting actions are repeated. If the time point changes, a new round of data combination process is started to advance the subsequent data so that they are arranged continuously after the data at the previous time point. For example, if the previous set of data corresponds to 08:35:12 and the next set is 08:35:15, then all values ​​at 08:35:15 will be sequentially continued after the previous set is fully advanced. During the process, content that does not belong to the time point should not be inserted. At the same time, for missing data, null placeholders should be used to clearly mark the missing data. For example, when tidal volume is missing, its position is marked with "—" and the subsequent data continues to be advanced, thereby completing the serial processing of the data content at each time point and obtaining a continuously advancing monitoring sequence.

[0026] S103: Based on the continuous monitoring sequence, each monitoring data in the sequence is sequentially connected in time. According to the rhythm of the advancement, the data of the differentiated equipment are pushed into the time structure channel in sequence to obtain the vital signs time chain. First, the timeline of each value in the push sequence is analyzed to determine if there are any discrepancies in the order of records across different devices. Especially considering the acquisition delay characteristics of different devices, the writing rhythm of values ​​from each device needs to be checked. For example, tidal volume data collected by a ventilator usually follows immediately after inspiration, while medication flow rate recorded by a microinfusion pump may be updated every fixed sampling period, and heart rate and blood pressure changes recorded by a monitor may be updated dynamically at high frequencies. Therefore, during the push process, data cannot be directly processed side-by-side according to device category. Instead, it must be uniformly connected according to the original order of each data point in the push sequence. For example, if a heart rate value of 82 is found before a systolic blood pressure of 125, and a tidal volume of 510 is found after blood pressure, these three values ​​should be pushed sequentially to the nth, n+1th, and n+2th positions of the continuous timeline, respectively. Simultaneously, it should be determined whether there are any abrupt changes in the update frequency of the device to which they belong. If the same device appears in multiple consecutive... If a data point is updated only once, the data from that device will not be used for time jump filling. The process continues with the next set of data. For each value, the process is then assessed based on the original acquisition cycle of the device to which it belongs to determine if there is a potential time shift. For example, if respiratory rate appears every three sets in the sequence, while heart rate is recorded in every set, the respiratory rate position should be left blank while continuously pushing heart rate data to avoid creating a false continuous trend. Then, all processed device values ​​are pushed into the time stream in the order of advancement until all data recording points are traversed, completing a full timeline processing flow. For example, if the following data set appears in a sequence: heart rate 84, tidal volume 520, oxygen concentration 38%, microinfusion pump rate 4.0, systolic blood pressure 130, diastolic blood pressure 85, and blood oxygen saturation 96%, it can be pushed to the same position in the continuous data stream according to the current time, constructing a complete record of all vital signs data within the entire time period, ultimately obtaining the vital signs timeline.

[0027] Please see Figure 3 The specific steps of S2 are as follows: S201: Based on the vital signs time chain, call the data content of bedside blood purification rate, ECMO blood flow and oxygenation ratio, extract the values ​​of the source equipment in the same time period, separate the data corresponding to each type of equipment according to the equipment number, extract the data items of the equipment, and obtain the equipment classification monitoring data. To retrieve data on bedside blood purification rate, ECMO blood flow rate, and oxygenation ratio, in practice, it's essential to first connect to the data interfaces or data export paths of each device. This involves obtaining the data record sheets from the bedside blood purification unit, the blood flow parameter list from the ECMO unit, and the historical oxygenation ratio records from the blood gas analyzer. During extraction, the data format exported from each device must be standardized, presenting all parameters numerically along with the device identification number for subsequent identification and classification. Each record in the extracted raw data should include a numerical value, device identification number, and acquisition time. For example, if the bedside purification unit outputs a rate value of 100, the device identification number is A01; if the ECMO unit outputs a blood flow rate value of 4.5, the device identification number is B01; and if the oxygenation ratio is 0.95, the device identification number is C01. In this case, the data needs to be separated by identification number. By reading the device identification number field corresponding to each data entry, the data with identification number A01 is categorized as purification data. Equipment data, numbered B01, is categorized as ECMO data, while data numbered C01 is categorized as blood gas-related data. After classification, the time field should be used to determine whether the data from each device falls within the same time period. For example, if the purification rate time is 08:42:10, the ECMO blood flow time is 08:42:12, and the oxygenation ratio time is 08:42:13, these values ​​can be grouped into a unified monitoring time period. Due to slight differences in the collection cycles of different devices, a maximum interval threshold should be set. For example, a time interval of no more than 3 seconds should be considered as the same time period. This judgment can be used to group data with similar times into the same group, ultimately forming a classified data list indexed by time period and categorized by device. Based on this, the numerical content of each type of device within the corresponding time period is extracted, and the values ​​of blood purification rate, ECMO blood flow, and oxygenation ratio are included in separate columns for subsequent cross-device data relationship analysis, thus obtaining the equipment classification monitoring data.

[0028] S202: Based on equipment classification monitoring data, obtain the detection time of sodium, potassium and chloride ions, the corresponding equipment data acquisition time, filter data items with consistent time, extract the numerical content within the same time point, and obtain the time-corresponding electrolyte data sequence. First, electrolyte test information is extracted from the laboratory data interface or electronic medical record platform. Each electrolyte test report typically includes the test item, test value, sample number, and test time field. Sodium, potassium, and chloride values ​​are usually listed in mmol / L. For example, the report might show a sodium concentration of 139, a potassium concentration of 4.3, and a chloride concentration of 101, corresponding to a specific data point. Next, the collection times for bedside blood purification rate, ECMO blood flow, and oxygenation ratio are extracted from the equipment monitoring data. Each piece of equipment data should include a time field and equipment number. For example, a purification rate record of 110 corresponds to time A, an ECMO blood flow record of 4.6 corresponds to time B, and an oxygenation ratio record of 0.92 corresponds to time C. To ensure accurate time interpretation, the format of all time fields should be standardized, and alignment should be performed according to hours, minutes, and seconds. After ensuring format consistency, each electrolyte test time needs to be compared line by line to determine if there are any completely identical equipment data points. If the sodium ion detection time is 08:26:42, which is exactly the same as the time field of a blood flow data point from the ECMO device, it is considered a time-consistent item. If the device acquisition time and detection time have only a 1 to 2 second error, it can be defined as a valid corresponding time according to the tolerance threshold set by different institutions. In this task, "complete consistency" is used as the screening criterion. After excluding any non-precise matching items, the monitoring values ​​and electrolyte detection values ​​appearing at the exact same time point are retained to complete the establishment of a correspondence. Then, the purification rate, ECMO blood flow, and oxygenation ratio values ​​at that time point are extracted and stored in the waiting queue along with the corresponding sodium, potassium, and chloride concentrations. For example, if the time point is 08:26:42, the corresponding purification rate is 105, the ECMO blood flow is 4.8, the oxygenation ratio is 0.93, and the sodium, potassium, and chloride are 138, 4.5, and 102, respectively, then the above six values ​​are regarded as joint data at the same time point. All combinations that meet the conditions are extracted in sequence to obtain the time-corresponding electrolyte data sequence.

[0029] S203: Based on the time-corresponding electrolyte data sequence, according to the order of the data items in the vital signs time chain, each group of data is advanced to the corresponding position in the original sequence according to the time point, and the data structure within the timeline is continued to obtain the cross-device operating status time chain. First, the time points associated with each data set in the data sequence are precisely extracted and used as the location basis. Then, their positions within the vital signs timeline are matched one-to-one. For example, one data set includes sodium concentration 138, potassium concentration 4.2, chloride concentration 102, purification rate 110, ECMO blood flow 4.6, and oxygenation ratio 0.91, corresponding to the time 08:52:26. Within this time point, the same time marker should be searched for within the vital signs timeline, and the time should be confirmed. The time point is located in the sequence number of the time chain, for example, the 47th data group. After confirming no duplicates, the complete data of this group is advanced to the position numbered 47. If the time point does not exist in the vital signs time chain, the data group is skipped without insertion to avoid logical deviations caused by time mismatch. If the time point exists but the original data at that position is incomplete, the electrolyte data corresponding to the current time is directly added to that position without overwriting. If other data, such as heart rate, blood oxygen saturation, tidal volume, etc., already exist in the vital signs time chain, they are also... The electrolyte data should be stored alongside this data at this location to maintain the independence of each data source. Simultaneously, the next set of electrolyte data should be processed sequentially along the timeline. Then, the time points in the next sequence should be extracted, and the above search and location process should be repeated. Multiple data times should be matched one by one to achieve time consistency integration across different device types of data. For example, three consecutive time points are 08:52:26, ​​08:53:14, and 08:53:45, corresponding to electrolyte and device parameters of group one (138, 4.2, 102, 110, ...). Group 4.6, 0.91), Group 2 (137, 4.1, 101, 108, 4.8, 0.89), and Group 3 (139, 4.3, 103, 115, 4.5, 0.93) should be pushed to positions 47, 48, and 49 respectively according to the three time sequences to form a stable time progression rhythm. No skipping insertions should be made, and the original sequence sorting method should not be disrupted. Finally, the supplementary data from electrolyte detection and other bedside equipment should be added to the basic vital signs data sequence to obtain a cross-equipment operating status time chain.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the cross-device operating status time chain, extract the numerical content of the monitored items within the same time period, filter the data items with consistent time points, and match each data item with the source device to obtain the time node paired dataset. First, a full traversal of the operational timeline of the participating devices is required. The monitoring data of each device under its operational state is arranged chronologically. During execution, data records for each monitoring item are retrieved from each device. Each monitoring item is processed individually during data extraction. For example, an ECG monitoring device records a heart rate of 76 beats / min, a respiratory monitoring device records a respiratory rate of 18 breaths / min, and an arterial blood pressure monitoring device records a systolic blood pressure of 122 mmHg and a diastolic blood pressure of 78 mmHg. Each data record is labeled with the device identifier and the collection timestamp. During processing, the timestamps are compared to determine if there are records with completely identical timestamps. This matching must be based on millisecond-level precision. If device A records a time of 14:35:20.123 and device B records a time of 14:35:20.456, they are considered inconsistent. If both are 14:35:20.000, they are marked as consistent timestamps. Monitoring data from all devices are extracted at these consistent timestamps. The filtering process is based on whether the data is consistent with the timetamp. Based on the condition of complete time point matching, only time records where all monitoring devices output data at that time point are retained. Further, each data point at that time point is labeled and bound to its source device. During execution, it is necessary to determine whether the device number corresponding to each data point corresponds to the device identification table. The device identification must be composed of a preset unique identification code. For example, device A is numbered ECG001, corresponding to heart rate data, and device B is numbered RESP002, corresponding to respiratory rate data. A dual binding method of data content and device identification is used during the pairing process. That is, only when both the time and device number correspondence are met is the data considered a valid pairing data item. If only some devices record data at a certain time point, all data corresponding to that time point are discarded. In this operation, it is also necessary to determine the reasonable value range of each data point. For example, the heart rate value should be between 30-180 beats / min. If it is less than 30 or greater than 180, it is an invalid data item and proceeds to the subsequent pairing stage. Through the above operations, the final time point paired dataset is obtained.

[0031] S302: Based on time node paired datasets, at each corresponding time position, the successfully paired data is advanced according to the original order of the running status time chain, and the monitoring items are connected to the monitoring parameters in chronological order to obtain the monitoring parameter chain. First, the order of each time point in the operational timeline must be clearly defined and locked to ensure that any subsequent data additions do not affect the existing timeline position. Based on this, extract each time point and corresponding monitoring item from the time node pairing dataset. Identify the four categories of data in the data structure: blood oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration, and extract their original values ​​and source identifiers. For example, if a time point is 08:21:16, and the successfully paired values ​​are SpO2 95, MAP 82, CVP 10, and Lac 2.3, then this time point should be used as the anchor point. Find the time period corresponding to 08:21:16 in the operational timeline, and then add the above four data items sequentially to the content group under that time point. During the process, the order of the monitoring items must not be changed, and no changes should be made to the data. Instead of replacing existing data items, new monitoring value data fields are directly added to them. If there is an overlap in time points but some identical monitoring data already exists, data marking and identification are required to ensure that only missing items are added and no duplicates are added. The next set of data is then added in chronological order. When the time point is 08:22:12, the data items are 96, 85, 9, and 2.1 respectively. After being added to this position, it is marked that all monitoring items have been added at this time point. Then, the above processing logic is repeated according to the next set of time points. No skipping steps are performed, and no missing data is added for unpaired data. The addition operation is always maintained in chronological order. When the processing scope covers the entire paired dataset, all identifiable monitoring data has been completely added to its time position, forming a continuously distributed multi-parameter monitoring value set, and finally the monitoring parameter chain is obtained.

[0032] Please see Figure 5 The specific steps of S4 are as follows: S401: Based on the monitoring parameter chain, call the tidal volume and respiratory rate values ​​at the same time point, extract the time period and change direction of each data, compare the trend direction of respiratory data with the trend direction of blood oxygen saturation, mean arterial pressure, central venous pressure and lactate concentration at the corresponding time point, analyze the change trend by comparing the values ​​at the current time point with the previous time point, filter the tidal volume and respiratory rate with the same direction, and obtain the ventilation data sequence with consistent trend. First, the tidal volume and respiratory rate values ​​at the same time point need to be retrieved. During the process, each tidal volume record is extracted. Tidal volume refers to the volume of gas inhaled or exhaled by the patient in a single breath. For example, if it is set to 500ml, and the tidal volume at the previous time point was 480ml, the change is judged as increasing. At the same time, the respiratory rate value needs to be extracted point by point. For example, if the current record is 18 breaths / min and the previous time point was 20 breaths / min, the change is decreasing. During the process, the current value of tidal volume and respiratory rate are compared with the previous value to obtain the trend of their change. Then, the corresponding blood oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration at the same time point are retrieved. The process involves sequentially calling up each parameter, comparing its current value with its previous value. For example, a decrease in blood oxygen saturation from 96% to 94% indicates a decrease; an increase in mean arterial pressure from 85 mmHg to 90 mmHg indicates an increase; an increase in central venous pressure from 8 mmHg to 10 mmHg indicates an increase; and an increase in lactate concentration from 2.1 mmol / L to 2.4 mmol / L indicates an increase. During the comparison, the direction of each parameter must correspond one-to-one with the trend direction of tidal volume and respiratory rate to determine their consistency at the same time point. If tidal volume increases while respiratory rate decreases, it is marked as inconsistent; if both increase, it is marked as consistent. The trend should be consistent, and then compared with the direction of parameters such as blood oxygen saturation. Clear interval divisions should be used during execution to avoid ambiguous judgments. For example, a respiratory rate greater than or equal to 16 breaths / min and less than or equal to 20 breaths / min is defined as the normal range; more than 20 breaths / min is defined as an increasing trend; and less than 16 breaths / min is defined as a decreasing trend. A tidal volume less than 400ml is defined as a low range; between 400ml and 600ml is the normal range; and greater than 600ml is the increasing range. Using these interval standards to compare trend directions avoids ambiguous terminology. In actual calculations, the above trend directions should be uniformly arranged. For example, a tidal volume of 480ml to 500ml is considered increasing, and a respiratory rate... A decrease in tidal volume is defined as a rate of 20 breaths / min to 18 breaths / min; a decrease in blood oxygen saturation is defined as a rate of 96% to 94%; an increase in mean arterial pressure is defined as a rate of 85 mmHg to 90 mmHg; an increase in central venous pressure is defined as a rate of 8 mmHg to 10 mmHg; and an increase in lactate concentration is defined as a rate of 2.1 mmol / L to 2.4 mmol / L. During this process, segments where tidal volume and respiratory rate are consistent are selected. For example, if the tidal volume increases from 480 ml to 520 ml and the respiratory rate increases from 18 breaths / min to 22 breaths / min at a certain time point, then the trend at that time point is consistent, and it is included in the trend-consistent ventilation data sequence. Otherwise, inconsistent segments are removed, resulting in the trend-consistent ventilation data sequence.

[0033] S402: Based on the trend-consistent ventilation data sequence, according to the content of the time point, the tidal volume and respiratory rate values ​​are respectively mapped to the time point of the monitoring parameters, and the data are connected to the corresponding positions according to the time relationship to obtain the ventilation-related monitoring sequence. First, identify the time point information corresponding to tidal volume and respiratory rate in each ventilation data point, and extract the time position index of each monitoring parameter in the original monitoring parameter chain at that time point. Then, determine the specific location where the ventilation data needs to be connected based on this index, using tidal volume and respiratory rate as the main identifiers. Read records one by one from all ventilation data with consistent trends. The read content must include key information such as numerical fields, equipment source identifiers, and time fields. Next, for each record, match the same time marker in the original monitoring parameter chain according to its time field. After matching, insert tidal volume and respiratory rate as independent fields into the data record at that location. For example, at a certain monitoring time point, there may already be a blood oxygen saturation of 97, a mean arterial pressure of 86, and a central venous pressure of [missing information]. 10. If the lactate concentration is 2.1 and ventilation data with a tidal volume of 480 and a respiratory rate of 20 are matched, these two values ​​need to be appended to the original record in sequence to form a complete record of seven monitoring parameters. Then, proceed to the next data record. During the insertion process, it is necessary to determine whether ventilation data already exists at the time point before insertion. If it does, multiple sets of ventilation data with the same time should be retained and listed after the monitoring parameter field according to the access order. For example, if there are two sets of tidal volume data of 480 and 490 at a certain time point, they need to be inserted into the monitoring entries at that time point in sequence. After all ventilation data has been inserted, the entire monitoring chain is checked in chronological order to ensure that all data items are not lost or misplaced, thus obtaining the ventilation-related monitoring sequence.

[0034] S403: Based on the ventilation-related monitoring sequence, call the oxygen partial pressure value in the blood gas analysis, extract the oxygen partial pressure content that is consistent with the ventilation data time, insert the oxygen partial pressure data into each data group in chronological order, and push it into the overall structure to obtain the respiratory and circulatory monitoring chain. First, the time points in each ventilation record need to be identified as the basis for subsequent data insertion. Next, the partial pressure of oxygen (MPO) data are read from the blood gas analysis, and their corresponding time and numerical fields are extracted. For example, a record with an MPO of 87 corresponds to a specific time. Then, by comparing the time fields of each monitoring item in the ventilation sequence, MPO data is matched one-to-one with records of the same time. If a data point in a ventilation record has the same time marker, an MPO of 87 is inserted as a new field into that record, expanding the original monitoring items from tidal volume, respiratory rate, oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration to seven data points. During this process, the original ventilation data needs to be retrieved line by line. Each time field is checked individually. Once a match is found, the field expansion and insertion are completed immediately. If there is no corresponding oxygen partial pressure data for a certain time point, the empty field can be retained or filled with a missing value marker to maintain structural integrity. Then, the structure of the dataset that has been paired and inserted is sorted out, and the newly added field column is positioned after the ventilation data to ensure that the field order is consistent. Finally, the sequential access of all oxygen partial pressure data is achieved. For example, if there are three sets of time in the monitoring sequence, t1, t2, and t3, which correspond to oxygen partial pressures of 87, 92, and 90 respectively, these three sets of values ​​need to be inserted into the data entries corresponding to their respective time positions in sequence. The inserted data is continuously advanced until all oxygen partial pressure data is matched and embedded into the ventilation sequence, and finally, the respiratory and circulatory monitoring chain is obtained.

[0035] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the respiratory and circulatory monitoring chain, call up medical orders, test results and imaging data within the same time period, extract the time field corresponding to each type of information, compare it with the time point of physiological monitoring data, filter data items with consistent time content, and obtain the time-corresponding non-physiological information set. First, the time fields included in the chain need to be clearly defined as the comparison benchmarks. For example, a set of monitoring data such as respiratory rate, partial pressure of oxygen, and blood oxygen saturation in the record is marked as a certain time point T1. This time point serves as a reference for subsequent screening of non-physiological information. Next, the data retrieval stage is entered, extracting time fields such as the execution time of medical orders, the sampling or issuance time in test reports, and the shooting time in imaging data. These time fields correspond to the specific time markers extracted for each type of information. For example, a blood routine test report shows a sampling time of T1, and a CT imaging report shows an shooting time of T2. Then, these time fields are compared one by one with the time points recorded in the respiratory and circulatory monitoring chain. Each comparison uses a line-by-line matching strategy, that is, when the time field of the test record matches a certain record in the monitoring chain... When the time fields are completely consistent, the non-physiological data is confirmed as a time content consistent item. The data content is then extracted from the original information table, while retaining its original information identifier, time identifier, and category field. If the time field of a medical order record or image file is T3, but there is no record with the time mark T3 in the monitoring chain, the inconsistent item is skipped and not processed. After performing time comparison operations on all non-physiological information one by one, multiple information records consistent with the time point of the monitoring chain can be obtained. For example, if the lactate test result time is T1, the ventilation record in the monitoring chain at time T1 is retained. Repeating this process can form a set of non-physiological information containing only consistent time fields, and finally obtain the time-corresponding non-physiological information set.

[0036] S502: Based on the time-corresponding non-physiological information set, according to the time order of each type of data, the text information and image data are respectively mapped to the adjacent positions of the time point where the monitoring data is located, and the data is mapped on the time line to obtain the sequence correspondence information structure; First, the time fields in the extracted medical orders, test descriptions, and imaging data need to be sorted separately, arranged into two sequences from chronological order: one is a plain text information sequence, and the other is an image data sequence. The source identifier and original location number are retained for each. Simultaneously, the monitoring time points in the monitoring parameter chain are organized, for example, multiple time markers such as T1, T2, and T3, corresponding to different respiratory and circulatory monitoring data groups. Next, the time points of various non-physiological information are read sequentially and matched with each time point in the monitoring parameter chain to determine if there are any cases where the time fields are completely identical. If an image record was taken at time T2, and the time of the T2 record in the monitoring chain is completely identical, then this image data is linked to the monitoring data record corresponding to T2 in chronological order via image linking or image preview. If there is a slight discrepancy between the image time and the monitoring data time, the data is considered identical. For offsets, such as time records differing by 1 to 2 minutes but being adjacent time periods, the original time is retained while arranging the data close to the corresponding time period in the display structure. This ensures a compact arrangement with adjacent physiological data, avoiding time jumps and insertion into non-corresponding time positions. Text information is processed in the same way. For example, if a medical order is signed at time T3, which matches the T3 time record in the monitoring chain, the order will be organized into a time-corresponding description segment and inserted before the corresponding monitoring data at T3 to form a nearby display. To avoid time overlap and confusion, the insertion position of non-physiological information is set to maintain consistency with its type. For example, laboratory test results take precedence over image data, and image data takes precedence over medical order descriptions. If multiple non-physiological information entries of the same type exist at the same time point, they are sequentially arranged according to their numerical order, ultimately forming an extended structure that corresponds one-to-one with the physiological monitoring data, resulting in a sequence-corresponding information structure.

[0037] S503: Based on the sequence correspondence information structure, extract medical orders, test items and imaging data that are not at the same time point, and append them one by one to the end of the monitoring sequence in chronological order, advancing the information into a continuous structure to obtain medical record data analysis archives; First, the matched medical orders, test results, and imaging data records are retrieved and identified by record number. Then, all medical order texts, test reports, and imaging data are iterated again to find remaining information items not yet assigned to any monitoring time point. These remaining information items are initially grouped by source category: medical order information is arranged in the order of issuance, test information in the order of detection, and imaging information in the order of recording. Each group retains the complete original time field of each item. For example, if an unmatched medical order is "fasting for 6 hours" issued at time point T4+1 after T4, this item is recorded in the unmatched medical order group. Next, the tail node time point of the constructed sequence correspondence information structure is read and used as the starting point of the additional time period. For example, if the tail monitoring record ends at time point T6, all unmatched non-physiological information will be assigned according to its own time field. The data is arranged sequentially with the time sequence following T6. For example, a chest X-ray taken on T7 will be linked to the first gap after T6, and a new node consisting of this image will be added to the end of the structure, appended to the end of the sequence, and marked with its time as T7. Similarly, if a lab report is issued in time T9 and the corresponding content is not synchronized with any monitoring point, the corresponding items and specific values ​​of the test content will be extracted and inserted into the end of the structure according to the text structure, with the words "T9-Test:" added before it for time archiving. All data not included in the main monitoring time point will not be inserted in the middle of the sequence to avoid structural disorder. At the same time, the time advancement interval is set to not be reversed. All newly added nodes must be connected in the order of the first occurrence before the last generation, ultimately forming a continuous and extended time-linked data structure, so that no unmatched content is omitted or mixed, resulting in a medical record data analysis archive.

[0038] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for the full information collection, synchronization, and analysis of medical record data based on an ICU robot, characterized in that: Includes the following steps: S1: Acquire heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation from ICU monitors, tidal volume and oxygen concentration from ICU ventilators, and flow rate from ICU micro-infusion pumps. Identify the data source devices, extract data entries at the same time points, analyze whether the acquisition time is consistent, and advance the unbiased data into the process in sequence to obtain the vital signs time chain. S2: Based on the vital signs time chain, call the bedside blood purification rate, ECMO blood flow and oxygenation ratio, split the data of the same time period according to the equipment type, analyze the relationship between the collection time and the sodium, potassium and chloride ion detection time, and advance the corresponding time data in the original order to obtain the cross-equipment operation status time chain. S3: Based on the cross-device operation status time chain, call the ICU monitor's blood oxygen saturation, mean arterial pressure, central venous pressure and bedside lactate concentration, pair the monitoring items and operation status, and push the paired data to the node to obtain the monitoring parameter chain; S4: Based on the monitoring parameter chain, call the values ​​of tidal volume and respiratory rate at the same time point, compare the trend direction, extract the oxygen partial pressure content according to the data with the same direction corresponding to the time position, and add it to the corresponding position in time order to obtain the respiratory and circulatory monitoring chain.

2. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The vital signs timeline includes heart rate, systolic blood pressure, diastolic blood pressure, blood oxygen saturation, tidal volume, oxygen concentration, and microinfusion pump flow rate. The cross-device operation status timeline includes bedside blood purification rate, ECMO blood flow, oxygenation ratio, sodium ion concentration, potassium ion concentration, and chloride ion concentration. The monitoring parameter chain includes blood oxygen saturation, mean arterial pressure, central venous pressure, and lactate concentration. The respiratory and circulatory monitoring chain includes tidal volume, respiratory rate, and partial pressure of oxygen.

3. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The time point data entries refer to the comparison of parameters from different devices according to timestamps during the multi-device monitoring process in the ICU, and the selection of data items with completely consistent collection times, which are then sequentially advanced to the processing stage at the same time position. The aforementioned detection time relationship refers to the process of comparing the electrolyte detection time with the bedside blood purification rate, ECMO blood flow, and oxygenation ratio device acquisition time during cross-device data processing, analyzing whether the timestamp differences are within a preset threshold, and then advancing the detection data to the corresponding time series position.

4. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The method further includes: in the processing of monitoring parameters, corresponding the mean arterial pressure with the data of blood oxygen saturation, central venous pressure, and lactate concentration at the same time position, and comparing and analyzing the trend direction with the ventilation data; The oxygen partial pressure content refers to the oxygen partial pressure value extracted from the blood gas analysis results during the respiratory and circulatory monitoring process, which is exactly the same as the ventilation data timestamp, and inserted into the corresponding monitoring data position in chronological order.

5. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Acquire the heart rate, systolic blood pressure, diastolic blood pressure, and blood oxygen saturation of the ICU monitor, as well as the tidal volume, oxygen concentration, and flow rate of the micro-infusion pump of the ICU ventilator. Identify the data source, corresponding device type and device number, compare with the attached timestamp information, select data items with consistent time content, and obtain the time synchronization monitoring dataset. S102: Based on the time-synchronized monitoring dataset, extract the values ​​of the device data at the same point in time, advance them into the data stream according to the order of data collection, and advance each item according to time to obtain a continuous monitoring sequence; S103: Based on the continuous monitoring sequence, each monitoring data in the sequence is sequentially connected in time. According to the time of the differentiated equipment data corresponding to the progress rhythm, the values ​​are pushed into the time structure channel in sequence to obtain the vital signs time chain.

6. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the vital signs time chain, call the data content of bedside blood purification rate, ECMO blood flow and oxygenation ratio, extract the values ​​of the source devices in the same time period, separate the data corresponding to each type of device according to the device number, extract the data items of the device, and obtain the device classification monitoring data. S202: Based on the equipment classification monitoring data, obtain the detection time of sodium, potassium and chloride ions, the corresponding equipment data acquisition time, filter data items with consistent time, extract the numerical content within the same time point, and obtain the time-corresponding electrolyte data sequence. S203: Based on the time-corresponding electrolyte data sequence, according to the order of the data items in the vital signs time chain, advance each group of data to the corresponding position in the original sequence according to the time point, continue the data structure within the timeline, and obtain the cross-device operating status time chain.

7. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the cross-device operation status time chain, extract the numerical content of the monitoring items within the same time period, filter the data items with consistent time points, and match each data item with the source device to obtain a time node paired dataset. S302: Based on the time node paired dataset, at each corresponding time position, the successfully paired data is advanced according to the original order of the running status time chain, and the monitoring items are connected to the next monitoring item in chronological order to obtain the monitoring parameter chain.

8. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the monitoring parameter chain, call the tidal volume and respiratory rate values ​​at the same time point, extract the time period and direction of change of each data, compare the trend direction of respiratory data with the trend direction of blood oxygen saturation, mean arterial pressure, central venous pressure and lactate concentration at the corresponding time point, analyze the trend of change by comparing the values ​​at the current time point with those at the previous time point, and filter out tidal volume and respiratory rate with the same direction to obtain a ventilation data sequence with consistent trend. S402: Based on the trend-consistent ventilation data sequence, according to the time point content, the tidal volume and respiratory rate values ​​are respectively mapped to the time point of the monitoring parameters, and the data are connected to the corresponding positions according to the time relationship to obtain the ventilation-related monitoring sequence. S403: Based on the ventilation-related monitoring sequence, call the oxygen partial pressure value in the blood gas analysis, extract the oxygen partial pressure content that is consistent with the ventilation data time, insert the oxygen partial pressure data into each group of data in chronological order, and advance it into the overall structure to obtain the respiratory and circulatory monitoring chain.

9. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 1, characterized in that, The method further includes: S5: Based on the respiratory and circulatory monitoring chain, call up medical order information, test results and image data with the same timestamp, compare the time and physiological data nodes of each type of information, advance the text and image content to the corresponding position according to time, and attach it to the existing data to obtain the medical record data analysis file; The medical record data analysis archive includes medical orders, test results, and imaging data.

10. The method for full information acquisition, synchronization, and analysis of medical record data based on an ICU robot according to claim 9, characterized in that, The specific steps of S5 are as follows: S501: Based on the respiratory and circulatory monitoring chain, call up the medical order information, test results and image data within the same time period, extract the time field corresponding to each type of information, compare it with the time point of physiological monitoring data, filter the data items with the same time content, and obtain the time-corresponding non-physiological information set. S502: Based on the time-corresponding non-physiological information set, according to the time order of each type of data, the text information and image data are respectively mapped to the adjacent positions of the time point where the monitoring data is located, and the data is mapped on the time line to obtain the sequence correspondence information structure; S503: Based on the sequence correspondence information structure, extract the medical orders, test items and imaging data that are not at the same time point, and append them one by one to the end of the monitoring sequence in chronological order, advancing the information into a continuous structure to obtain a medical record data analysis archive.

Images