An electrocardiogram signal analysis and processing system

Abstract

The embodiment of the present invention relates to an electrocardiogram signal analysis and processing system, which includes: a forwarding module, a first cache module, a second cache module and an analysis module; the system is connected to an acquisition device. The forwarding module transfers the uploaded data packet of the acquisition device to the first cache module; the first cache module sends a first record sequence consisting of a specified number of cache records to the analysis module; the analysis module performs multi-lead long signal splicing, electrocardiogram signal analysis and warning data analysis to generate new warning data cache records and send them to the second cache module; the second cache module stores the new records and extracts the previous and current records and pushes them to the forwarding module; the forwarding module performs warning data deduplication on the current record based on the previous record to generate corresponding downlink data packets and send them to the acquisition device. The system of the present invention can solve the problem that bedside monitors cannot be deployed universally, and can also solve the problem that existing acquisition devices do not have complex analysis capabilities.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an electrocardiogram signal analysis and processing system. Background Art

[0002] Large medical institutions are equipped with bedside monitors, which are high-end multifunctional devices that can simultaneously handle the collection, analysis and early warning of ECG signals. However, such multifunctional bedside monitors are generally complex to operate, large in size and expensive, making them difficult to deploy in wards, and even more impossible to provide them as home monitoring devices to some older, mobility-impaired users who need long-term ECG monitoring.

[0003] There are also some ECG collection components or devices on the market that are simple to operate, low-priced, small in size and easy to carry. Among them, the ECG collection components can cooperate with personal mobile phones and computer terminals to store the collected short-term ECG signals (for example, 1 minute) on the mobile phone, and the mobile phone APP can perform simple analysis of the short-term ECG signals such as heart rate measurement; ECG collection equipment can be used for long-term (for example, 24 hours) portable ECG signal collection for users, and after the collection is completed, it is transmitted to the background of the medical institution through a wired or wireless network for professional personnel to read and analyze the information. Although these ECG collection components or devices can be deployed in a popular manner, they do not have or the local devices connected to them do not have complex ECG signal analysis capabilities, and cannot achieve the effect of real-time analysis and early warning. Summary of the invention

[0004] The purpose of the present invention is to provide an ECG signal analysis and processing system to address the defects of the prior art. The system can be connected to any ECG acquisition component or device. The system includes: a forwarding module, a first cache module, a second cache module and an analysis module. The above-mentioned ECG acquisition components or devices are collectively referred to as acquisition devices, and the forwarding module is responsible for receiving the upload data packets sent by the acquisition device and forwarding them to the cache queue of the first cache module for storage; the first cache module is responsible for regularly forming a first record sequence with a nearly specified number of cache records and pushing it to the analysis module; in order to improve the analysis accuracy, the analysis module performs long frame signal splicing based on the first record sequence, and performs ECG signal feature analysis and heart beat morphology feature analysis based on multi-lead long frame signals, and performs warning event and warning level identification based on the multi-lead feature set obtained by analysis to generate corresponding warning data cache records, and sends the warning data cache records to the cache queue of the second cache module for storage; the second cache module queues and stores each new warning data cache record received, and extracts the last warning data cache record from the queue and pushes it to the forwarding module together with the current warning data cache record in real time; after receiving the last and current warning data cache records, the forwarding module filters out the overlapping warning information in the current record that overlaps with the last record in the overlapping period, and sends the deduplicated warning data cache record as a downlink data packet to the acquisition device. By combining the system of the present invention with existing ECG acquisition components or devices on the market, the problem that bedside monitors cannot be widely deployed can be solved, and the problem that existing ECG acquisition components or devices do not have the ability to analyze complex ECG signals can also be solved.

[0005] To achieve the above-mentioned purpose, an embodiment of the present invention provides an electrocardiogram signal analysis and processing system, the system comprising: a forwarding module, a first buffer module, a second buffer module and an analysis module; the system is connected to a collection device;

[0006] The forwarding module is connected to the acquisition device, the first cache module, the second cache module and the analysis module respectively; the forwarding module is used to receive a first uplink data packet sent by the acquisition device; and extract a first data type and a first data body from the first uplink data packet; and send the first data body to the analysis module when the first data type is the first type; and send the first data body to the first cache module when the first data type is the second type; the forwarding module is also used to receive the current warning data cache record and the previous warning data cache record pushed by the second cache module; and perform warning data deduplication processing on the current warning data cache record according to the previous warning data cache record to generate a corresponding first downlink data packet; and send the first downlink data packet to the acquisition device;

[0007] The first cache module is connected to the analysis module; the first cache module is used to store the ECG data cache queue; the first cache module is also used to receive the first data body sent by the forwarding module; and store the first data body in the ECG data cache queue as the latest first ECG data cache record; and obtain the latest specified number of the first ECG data cache records from the ECG data cache queue to form a corresponding first record sequence and push it to the analysis module;

[0008] The second cache module is connected to the analysis module; the second cache module is used to store the warning data cache queue; the second cache module is also used to receive the current warning data cache record sent by the analysis module; and store the current warning data cache record into the warning data cache queue as the latest first warning data cache record; and extract the previous record of the current warning data cache record from the warning data cache queue as the corresponding previous warning data cache record; and push the current warning data cache record and the previous warning data cache record to the forwarding module;

[0009] The analysis module is used to receive the first data body sent by the forwarding module; and extract the warning parameter set from the first data body and store it locally; the analysis module is also used to receive the first record sequence pushed by the first cache module; and perform multi-lead long signal splicing processing according to the first record sequence to generate a corresponding multi-lead long frame signal record; and perform electrocardiogram signal analysis processing according to the multi-lead long frame signal record to generate a corresponding multi-lead feature set; and perform warning data analysis processing according to the multi-lead feature set and the warning parameter set to generate the corresponding warning data cache record of the time; and send the warning data cache record of the time to the second cache module.

[0010] Preferably, the forwarding module is connected to the acquisition device via a wired or wireless method;

[0011] The wired mode at least includes a first wired mode based on a network cable, a second wired mode based on a coaxial data cable, a third wired mode based on a serial communication data cable, and a fourth wired mode based on a parallel communication data cable;

[0012] The wireless methods include a first wireless method based on 3G / 4G / 5G / LTE mobile communication network, a second wireless method based on Bluetooth communication, a third wireless method based on wireless local area network, and a fourth wireless method based on near field communication NFC protocol.

[0013] Preferably, the ECG data cache queue includes a plurality of the first ECG data cache records; the first ECG data cache record includes a first short frame number and a plurality of first short frame data groups; the first short frame data group includes a first lead identifier and a first lead short frame signal;

[0014] The warning data cache queue includes multiple first warning data cache records; the first warning data cache record includes a first long frame number and a first warning data sequence; the first warning event sequence includes multiple first time period warning data groups; the first time period warning data group includes the first time period start and end time, the first time period warning event and the first time period event level.

[0015] Preferably, the forwarding module is specifically used to, when performing warning data deduplication processing on the current warning data cache record according to the previous warning data cache record, record the first warning data sequences of the previous and current warning data cache records as the corresponding previous sequence and current sequence respectively; and record the two first time period warning data groups with the same start and end time of the first time period in the previous and current sequences as the corresponding first matching pairs; and traverse each of the first matching pairs; during traversal, if the two first time period warning events of the first matching pair currently traversed are the same and the two first time period event levels are the same, then the first time period warning data group belonging to the current sequence in the first matching pair currently traversed is recorded as a duplicate warning data group; at the end of the traversal, delete all the duplicate warning data groups in the current warning data cache record, and use the deleted current warning data cache record as the corresponding first downlink data packet.

[0016] Preferably, the analysis module is specifically used to perform sequential signal splicing processing on multiple first lead short frame signals with the same first lead identifier in the first recording sequence in ascending order of the corresponding first short frame numbers to obtain corresponding first lead long frame signals when performing multi-lead long signal splicing processing according to the first recording sequence; and each first lead identifier and the corresponding first lead long frame signal constitute a corresponding first long frame data group; and the first short frame number with the largest numerical value in the first recording sequence is used as the corresponding first long frame number; and the first long frame number and multiple first long frame data groups constitute the corresponding multi-lead long frame signal record.

[0017] Preferably, the analysis module is specifically used for, when performing electrocardiographic signal analysis and processing according to the multi-lead long frame signal record,

[0018] Identify square wave signal segments and dropout signal segments of each first lead long frame signal in the multi-lead long frame signal record, and perform signal splicing processing on the identified square wave signal segments and dropout signal segments based on the signal duration of the current first lead long frame signal to generate corresponding first lead square wave signals and first lead dropout signals;

[0019] and performing QRS wave group detection on each of the first lead long frame signals to generate a corresponding first lead QRS wave group sequence, and taking the peak point of each QRS wave group in the first lead QRS wave group sequence as the corresponding first lead R point, and sorting all the obtained first lead R points in chronological order to generate a corresponding first lead R point sequence;

[0020] And according to the first lead square wave signal and the first lead drop-off signal corresponding to each of the first lead long frame signals, the corresponding first lead R point sequence is subjected to noise filtering to generate a corresponding second lead R point sequence; and the second lead R point sequence of all leads is subjected to R point fusion processing to generate a corresponding first R point time sequence; wherein the first R point time sequence includes multiple first R point times;

[0021] And in each of the first lead long frame signals, taking each of the first R point times as the center, a segment of the electrocardiogram signal with a preset first duration is intercepted forward and backward to form a corresponding first heartbeat signal segment, and a segment of the electrocardiogram signal with a preset second duration is intercepted forward and backward to form a corresponding second heartbeat signal segment; and a heartbeat quality assessment process is performed on each of the first heartbeat signal segments to generate a corresponding first quality grade; and a conventional heartbeat morphology classification process is performed on each of the second heartbeat signal segments to generate a corresponding first heartbeat type; and the corresponding first heartbeat is classified according to each of the first quality grades. The first heartbeat type is corrected, and when the first quality level is a weak interference heartbeat level or a strong interference heartbeat level, the corresponding first heartbeat type is changed to an interference type; and all the first heartbeat types after correction are sorted in chronological order to generate a corresponding first lead heartbeat type sequence; wherein, the first duration is less than the second duration; the first quality level includes a normal heartbeat level, a weak interference heartbeat level and a strong interference heartbeat level; the first heartbeat type includes a sinus heartbeat type, an atrial fibrillation heartbeat type, an atrial heartbeat type, a ventricular heartbeat type and a junctional heartbeat type; the first heartbeat type corresponds to the first R point time;

[0022] and performing QRS complex, P wave and T wave feature recognition processing on each of the first lead long frame signals according to the first R point time sequence and the first lead heart beat type sequence to generate a corresponding first QRS complex feature sequence, a first P wave feature sequence and a first T wave feature sequence; wherein the first QRS complex feature sequence includes a plurality of first QRS complex features; the first P wave feature sequence includes a plurality of first P wave features; the first T wave feature sequence includes a plurality of first T wave features; each of the first QRS complex feature, the first P wave feature and the first T wave feature corresponds to the first R point time;

[0023] and performing ventricular flutter and ventricular fibrillation feature recognition processing on each of the first lead long frame signals to generate a corresponding first lead unconventional feature sequence; wherein the first lead unconventional feature sequence includes a plurality of first ventricular flutter features and a plurality of first ventricular fibrillation features; the first ventricular flutter feature includes a first ventricular flutter starting point and a first ventricular flutter ending point; the first ventricular fibrillation feature includes a first ventricular fibrillation starting point and a first ventricular fibrillation ending point;

[0024] and performing low voltage type setting processing on each of the first lead heart beat type sequences according to the first R point time sequence;

[0025] and performing atrial fibrillation heartbeat type correction processing on each of the first lead heartbeat type sequences;

[0026] and performing refinement processing on each of the first lead heart beat type sequences into atrial, ventricular and junctional heart beats according to the first R point time sequence;

[0027] The second lead R point sequence, the first QRS wave group characteristic sequence, the first P wave characteristic sequence, the first T wave characteristic sequence, the first lead unconventional characteristic sequence and the first lead heart beat type sequence corresponding to each of the first lead long frame signals constitute a corresponding first lead feature set; and all of the first lead feature sets constitute the corresponding multi-lead feature set.

[0028] Furthermore, the analysis module is specifically used to merge the second lead R points of the second lead R point sequences of all leads together to form a corresponding R point set when performing R point fusion processing on the second lead R point sequences of all leads; and cluster multiple R points in the R point set whose time interval is less than a preset minimum time threshold into the same R point group; and calculate the mean of the R point time of each of the R point groups and use the calculation result as the corresponding first R point time; and generate a corresponding first R point time series by sorting all the obtained first R point times in chronological order.

[0029] Further, the analysis module is specifically used for performing QRS complex, P wave and T wave feature recognition processing on each of the first lead long frame signals according to the first R point time sequence and the first lead heart beat type sequence,

[0030] In each of the first lead long frame signals, a signal point corresponding to the first R point time corresponding to the first heart beat type that is not an interference type is recorded as a first-category point, and a signal point corresponding to the first R point time corresponding to the first heart beat type that is specifically an interference type is recorded as a second-category point;

[0031] The starting and ending positions of the QRS complex where each of the first-class points is located are identified to generate a corresponding first QRS complex starting point and a first QRS complex ending point, and the time difference between the first ending point and the first starting point is calculated to generate a corresponding first QRS complex duration, and the first QRS complex starting point, the first QRS complex ending point and the first QRS complex duration form a corresponding first QRS complex feature; and the first QRS complex feature corresponding to each of the second-class points is set as a preset invalid feature; and all the obtained first QRS complex features are sorted in chronological order to generate a corresponding first QRS complex feature sequence;

[0032] And the peak point, starting point and ending point of the previous P wave of the QRS wave group where each of the first-class points is located are identified to generate the corresponding first P wave peak point, first P wave starting point and first P wave ending point to form the corresponding first P wave feature; and the first P wave feature corresponding to each of the second-class points is set as a preset invalid feature; and all the obtained first P wave features are sorted in chronological order to generate the corresponding first P wave feature sequence;

[0033] And the peak point, starting point and ending point of the next T of the QRS wave group where each of the first-class points is located are identified to generate the corresponding first T wave peak point, first T wave starting point and first T wave ending point to form the corresponding first T wave feature; and the first T wave feature corresponding to each of the second-class points is set as a preset invalid feature; and all the obtained first T wave features are sorted in chronological order to generate the corresponding first T wave feature sequence.

[0034] Furthermore, the analysis module is specifically used to record the first lead long frame signal corresponding to the current first lead heart beat type sequence as the current lead long frame signal when the low voltage type setting processing is performed on each of the first lead heart beat type sequences according to the first R point time sequence; and record the signal point corresponding to each of the first R point times in the current lead long frame signal as the corresponding first signal point; and record the first signal point whose signal point amplitude is lower than the preset low voltage amplitude as the corresponding second signal point; and record the first R point time corresponding to each of the second signal points as the corresponding second R point time; and set the first heart beat type corresponding to each of the second R point times in the current first lead heart beat type sequence to the corresponding low voltage type.

[0035] Furthermore, the analysis module is specifically used to traverse any three consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial fibrillation heart beat type correction processing on each of the first lead heart beat type sequences; when traversing, the three first heart beat types currently traversed are recorded as corresponding front heart beat type, center heart beat type and back heart beat type in the order of front, middle and back; and confirm whether the center heart beat type is an atrial fibrillation heart beat type; if it is confirmed to be an atrial fibrillation heart beat type, confirm whether both the front heart beat type and the back heart beat type are not atrial fibrillation heart beat types; if it is confirmed that neither is an atrial fibrillation heart beat type, change the center heart beat type to a sinus heart beat type.

[0036] Further, the analysis module is specifically used to traverse any two consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial, ventricular and junctional heart beat refinement processing on each first lead heart beat type sequence according to the first R point time sequence; when traversing, the two first heart beat types currently traversed are recorded as corresponding front heart beat types and back heart beat types in a front and back order; and when the front and back heart beat types are both atrial heart beat types, the absolute difference between the two first R point times corresponding to the front and back heart beat types is calculated to generate the corresponding first RR interval, and the heart rate is estimated according to the first RR interval to generate the corresponding first heart rate, and when the first heart rate is lower than a preset atrial escape heart rate threshold, the front and back heart beat types are changed to corresponding atrial escape heart beat types, and when the first heart rate is higher than a preset atrial tachycardia heart rate threshold, the front and back heart beat types are changed to corresponding atrial tachycardia heart beat types; and when the front and back heart beat types are both ventricular heart beat types, An absolute difference calculation is performed on the two first R point times corresponding to the front and back heartbeat types to generate a corresponding second RR interval, and a heart rate estimation is performed based on the second RR interval to generate a corresponding second heart rate, and when the second heart rate is lower than a preset ventricular escape heart rate threshold, the front and back heartbeat types are changed to corresponding ventricular escape heart beat types, and when the second heart rate is higher than a preset ventricular tachycardia heart rate threshold, the front and back heartbeat types are changed to corresponding ventricular tachycardia heart beat types; and when the front and back heartbeat types are both junctional heartbeat types, an absolute difference calculation is performed on the two first R point times corresponding to the front and back heartbeat types to generate a corresponding third RR interval, and a heart rate estimation is performed based on the third RR interval to generate a corresponding third heart rate, and when the third heart rate is lower than a preset junctional escape heart rate threshold, the front and back heartbeat types are changed to corresponding junctional escape heart beat types, and when the third heart rate is higher than a preset junctional tachycardia heart rate threshold, the front and back heartbeat types are changed to corresponding junctional tachycardia heart beat types.

[0037] Preferably, the analysis module is specifically used for, when performing the early warning data analysis and processing according to the multi-lead feature set and the early warning parameter set,

[0038] Performing warning event and warning event level identification processing according to the multi-lead feature set and the warning parameter set to generate a corresponding first identification report; wherein the first identification report includes a plurality of first identification records; the first identification record includes a first event name, a first event level and a first event start time;

[0039] and equally divide the time interval of the first record sequence according to the specified number to obtain the specified number of first time periods; and assign a corresponding first time period warning data group to each first time period; and set the first time period start and end time of the first time period warning data group according to the start and end time of each first time period, and initialize the first time period warning event and the first time period event level of each first time period warning data group to null;

[0040] and traverse each of the first time periods; during the traversal, record the first time period currently traversed as the current time period; and select the first identification record with the highest first event level from all the first identification records whose first event start time meets the current time period as the current identification record; and set the first time period warning event and the first time period event level of the first time period warning data group corresponding to the current time period according to the first event name and the first event level of the current identification record;

[0041] At the end of the traversal, the specified number of the first time period warning data groups are sorted in the order of the corresponding time periods to generate the corresponding first warning data sequence; and the first long frame number and the first warning data sequence form the corresponding warning data cache record.

[0042] The embodiment of the present invention provides an electrocardiogram signal analysis and processing system, which can be connected to any electrocardiogram acquisition component or device. The system includes: a forwarding module, a first buffer module, a second buffer module and an analysis module. The above-mentioned ECG acquisition components or devices are collectively referred to as acquisition devices, and the forwarding module is responsible for receiving the upload data packets sent by the acquisition device and forwarding them to the cache queue of the first cache module for storage; the first cache module is responsible for regularly forming a first record sequence with a nearly specified number of cache records and pushing it to the analysis module; in order to improve the analysis accuracy, the analysis module performs long frame signal splicing based on the first record sequence, and performs ECG signal feature analysis and heart beat morphology feature analysis based on multi-lead long frame signals, and performs warning event and warning level identification based on the multi-lead feature set obtained by analysis to generate corresponding warning data cache records, and sends the warning data cache records to the cache queue of the second cache module for storage; the second cache module queues and stores each new warning data cache record received, and extracts the last warning data cache record from the queue and pushes it to the forwarding module together with the current warning data cache record in real time; after receiving the last and current warning data cache records, the forwarding module filters out the overlapping warning information in the current record that overlaps with the last record in the overlapping period, and sends the deduplicated warning data cache record as a downlink data packet to the acquisition device. By combining the system of the present invention with existing ECG acquisition components or devices on the market, the problem that bedside monitors cannot be widely deployed is solved, and the problem that existing ECG acquisition components or devices do not have the ability to analyze complex ECG signals is also solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of the system structure of an electrocardiogram signal analysis and processing system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] Figure 1 A schematic diagram of the system structure of an electrocardiogram signal analysis and processing system provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the ECG signal analysis and processing system 1 comprises: a forwarding module 11 , a first buffer module 12 , a second buffer module 13 and an analysis module 14 ; the ECG signal analysis and processing system 1 is connected to the acquisition device 2 .

[0046] Here, the acquisition device 2 can be a terminal connected to various ECG acquisition components, such as a mobile phone, mobile terminal, PAD or computer connected to an ECG acquisition patch; it can also be an ECG acquisition device with storage capabilities, such as a dynamic ECG acquisition device Holter; it can also be a special device or instrument with ECG acquisition capabilities, such as an ECG machine, a bedside monitor, etc.

[0047] (I) Forwarding module 11

[0048] The forwarding module 11 is connected to the collection device 2 , the first buffer module 12 , the second buffer module 13 and the analysis module 14 respectively.

[0049] Among them, the forwarding module 11 is connected to the acquisition device 2 via a wired or wireless method; the wired method at least includes a first wired method based on a network cable, a second wired method based on a coaxial data cable, a third wired method based on a serial communication data cable, and a fourth wired method based on a parallel communication data cable; the wireless method includes a first wireless method based on a 3G / 4G / 5G / LTE mobile communication network, a second wireless method based on Bluetooth communication, a third wireless method based on a wireless local area network, and a fourth wireless method based on a near-field communication NFC protocol.

[0050] Here, in order to improve the system's processing capability for high concurrency, reduce the communication congestion pressure of the analysis module, and balance the load capacity of the entire system, the embodiment of the present invention adopts an asynchronous communication mechanism similar to the message queue mechanism to process the data transmission between the forwarding module 11 and the analysis module 14. The first cache module 12 in the system is the input cache management end from the forwarding module 11 to the analysis module 14, and the second cache module 13 in the system is the output cache management end from the analysis module 14 to the forwarding module 11.

[0051] The forwarding module 11 is used to receive a first uplink data packet sent by the acquisition device 2; extract a first data type and a first data body from the first uplink data packet; and send the first data body to the analysis module 14 when the first data type is the first type; and send the first data body to the first cache module 12 when the first data type is the second type;

[0052] The first data type includes a first type and a second type.

[0053] Here, on the side of the analysis module 14, each collection device 2 has a set of default warning parameter sets for early warning analysis and processing. The user can also modify this set of warning parameter sets through the collection device 2. When the user modifies it, the collection device 2 collects the warning parameter set newly set by the user as the first data body, sets the first data type to the first type, and sends the first uplink data packet of the first data type + the first data body to the forwarding module 11. After receiving the first uplink data packet of the first data type as the first type, the forwarding module 11 forwards it to the analysis module 14 to complete the parameter set update;

[0054] In addition, when the acquisition device 2 enters the signal acquisition processing process, it will perform real-time ECG acquisition and upload for the user; each time the acquisition signal is uploaded, the acquisition device 2 will assign a number that is sequentially increased by 1, that is, the first short frame number, to this upload; and will generate multiple first short frame data groups based on the signals of each lead identifier and the latest specified segment length (the default is 6 seconds) of the synchronous acquisition of each lead, the first lead identifier of each first short frame data group is the lead identifier information of the corresponding lead, the first lead short frame signal is the synchronous acquisition signal of the corresponding lead, and the duration of the first lead short frame signal is the specified segment length; and the first short frame number + multiple first short frame data groups constitute the first data body; then the first uplink data packet consisting of the first data body and the first data type specifically set to the second type is sent to the forwarding module 11; after receiving the first uplink data packet whose first data type is the second type, the forwarding module 11 will forward the first data body (first short frame number + multiple first short frame data groups) therein to the first cache module 12 for caching.

[0055] The forwarding module 11 is also used to receive the current warning data cache record and the previous warning data cache record pushed by the second cache module 13; and perform warning data deduplication processing on the current warning data cache record according to the previous warning data cache record to generate a corresponding first downlink data packet; and send the first downlink data packet to the collection device 2;

[0056] Among them, the previous and current warning data cache records include a first long frame number and a first warning data sequence; the first warning event sequence includes multiple first-time period warning data groups; the first-time period warning data group includes the first-time period start and end time, the first-time period warning event and the first-time period event level.

[0057] In a specific implementation of an embodiment of the present invention, the forwarding module 11 is specifically used to, when deduplicating warning data for the current warning data cache record based on the previous warning data cache record, record the first warning data sequences of the previous and current warning data cache records as the corresponding previous sequence and current sequence respectively; and record the two first time period warning data groups with the same start and end time of the first time period in the previous and current sequences as the corresponding first matching pairs; and traverse each first matching pair; during traversal, if the two first time period warning events of the first matching pair currently traversed are the same and the levels of the two first time period events are the same, then the first time period warning data group belonging to the current sequence in the first matching pair currently traversed is recorded as a duplicate warning data group; at the end of the traversal, delete all duplicate warning data groups in the current warning data cache record, and use the deleted current warning data cache record as the corresponding first downlink data packet.

[0058] Here, after receiving the current warning data cache record output by the analysis module 14, the second cache module 13 will extract the previous warning data cache record from the local cache queue and push it together with the current warning data cache record to the second cache module 13; the forwarding module 11 will delete the warning data of the same time period in the current warning data cache record according to the previous warning data cache record, and then send the deleted record content as the first downlink data packet to the collection device 2.

[0059] (II) First cache module 12

[0060] The first cache module 12 is connected to the analysis module 14 .

[0061] The first cache module 12 is used to store the ECG data cache queue;

[0062] Among them, the ECG data cache queue includes multiple first ECG data cache records; the first ECG data cache record includes a first short frame number and multiple first short frame data groups; the first short frame data group includes a first lead identifier and a first lead short frame signal.

[0063] The first buffer module 12 is also used to receive the first data body sent by the forwarding module 11; store the first data body in the ECG data buffer queue as the latest first ECG data buffer record; and obtain the latest specified number of first ECG data buffer records from the ECG data buffer queue to form a corresponding first record sequence and push it to the analysis module 14;

[0064] The default value of the specified number is 12.

[0065] Here, the first cache module 12 performs storage management on the ECG data cache queue in a first-in-first-out manner, that is, before the storage space of the ECG data cache queue is used up, the newly added first ECG data cache record is recorded and added in a sequential storage manner, and once the storage space is used up, the newly added first ECG data cache record will record over the earliest added first ECG data cache record;

[0066] The ECG data cache queue of the embodiment of the present invention is a message queue (Message Queue, MQ) for storing ECG data. Between the forwarding module 11, the analysis module 14 and the first cache module 12, the forwarding module 11 is the data producer (Producer) of the message queue, the first cache module 12 is the queue manager or broker of the message queue, and the analysis module 14 is the data consumer (Consumer) of the message queue; the forwarding module 11 sends the first data body (first short frame number + multiple first short frame data groups) to the first cache module 12 based on the message publishing format agreed in advance with the first cache module 12, and the first cache module 12 adds the first data body (first short frame number + multiple first short frame data groups) as a newly added first ECG data cache record to the ECG data cache queue, thereby realizing the message publishing operation of the message queue;

[0067] The embodiment of the present invention will not analyze the short-frame ECG of the specified segment length (the default is 6 seconds) uploaded by the acquisition device each time, because the features of such short-frame signals are not rich enough and analysis deviations are prone to occur; in order to take into account both the real-time and accuracy of the analysis, the embodiment of the present invention stipulates that each analysis is based on the most recent long-frame signal with a duration of the specified number * the specified segment length (the default is 12*6=72 seconds) for analysis; and to ensure that the analysis module 14 can always obtain the latest multi-lead long-frame signal in a timely manner, it is necessary for the first cache module 12 to push an instant message to the analysis module 14 through the push mechanism of the message queue; the embodiment of the present invention uses the record addition operation of the ECG data cache queue as the message queue. The activation point of information push is that once the first cache module 12 stores the new first data body into the ECG data cache queue, the message push process is activated. The message push process is to extract the most recently specified number (the default is 12) of first ECG data cache records from the ECG data cache queue to form a corresponding first record sequence and push it to the analysis module 14; it should be noted that in the initial stage, when the number of first ECG data cache records in the ECG data cache queue is less than the specified number (the default is 12), the first cache module 12 will extract all the first ECG data cache records in the ECG data cache queue to form a corresponding first record sequence and push it to the analysis module 14 when processing the long frame signal message push process.

[0068] (III) Second cache module 13

[0069] The second cache module 13 is connected to the analysis module 14 .

[0070] The second cache module 13 is used to store the warning data cache queue;

[0071] Among them, the warning data cache queue includes multiple first warning data cache records; the first warning data cache record includes a first long frame number and a first warning data sequence; the first warning event sequence includes multiple first time period warning data groups; the first time period warning data group includes the first time period start and end time, the first time period warning event and the first time period event level.

[0072] Here, each first warning data cache record corresponds to a first record sequence pushed by the first cache module 12 to the analysis module 14, because the first record sequence includes a specified number (default is 12) of first ECG data cache records and each first ECG data cache record corresponds to a first short frame number, that is, each first long frame number corresponds to a specified number (default is 12) of first short frame numbers; it can be seen from the previous text that the first short frame number is encoded in a sequential addition 1 manner, and it can be seen from the following text that the first long frame number is the maximum value of a corresponding set of specified number (default is 12) of first short frame numbers; similarly, the number of first time period warning data groups of the first warning data sequence is also consistent with the specified number (default is 12), and the time period of each first time period warning data group corresponds to the time period of a first ECG data cache record in the first record sequence; the first time period warning event and the first time period event level of the first time period warning data group can be empty. If the first time period warning event and the first time period event level are empty, it means that there is no warning event in the current time period.

[0073] The second cache module 13 is also used to receive the current warning data cache record sent by the analysis module 14; and store the current warning data cache record into the warning data cache queue as the latest first warning data cache record; and extract the previous record of the current warning data cache record from the warning data cache queue as the corresponding previous warning data cache record; and push the current warning data cache record and the previous warning data cache record to the forwarding module 11.

[0074] Here, the second cache module 13 performs storage management on the warning data cache queue in a first-in-first-out manner, that is, before the storage space of the warning data cache queue is used up, the newly added first warning data cache record is recorded and added in a sequential storage manner. Once the storage space is used up, the newly added first warning data cache record will record overwrite the earliest added first warning data cache record;

[0075] The warning data cache queue of the embodiment of the present invention is a message queue for storing warning data. Between the forwarding module 11, the analysis module 14 and the second cache module 13, the analysis module 14 is the data producer of the message queue, the second cache module 13 is the queue manager of the message queue, and the forwarding module 11 is the data consumer of the message queue; the analysis module 14 sends the warning data cache record (first long frame number + first warning data sequence) to the second cache module 13 based on the message publishing format agreed in advance with the second cache module 13, and the second cache module 13 adds the warning data cache record (first long frame number + first warning data sequence) as the newly added first warning data cache record to the warning data cache queue, thereby realizing the message publishing operation of the message queue;

[0076] In order to ensure that the forwarding module 11 can always obtain the latest warning data cache record in time, the second cache module 13 needs to push instant messages to the forwarding module 11 through the push mechanism of the message queue in the embodiment of the present invention; the embodiment of the present invention uses the record addition operation of the warning data cache queue as the activation point of the message push, and once the second cache module 13 stores the current warning data cache record into the warning data cache queue, the message push process is activated, and the message push process is to extract the previous warning data cache record from the warning data cache queue and push it together with the current warning data cache record to the forwarding module 11; it should be noted that the reason for extracting the previous warning data cache record is only because there are multiple first warning data groups of repeated time periods in adjacent warning data cache records, and the first warning data groups of these repeated time periods are very likely to be the same. In order to avoid sending duplicate warning data to the collection device 2, the forwarding module 11 is required to deduplicate the duplicate warning data in the current warning data cache record based on the previous warning data cache record before pushing down the downlink data packet each time.

[0077] (IV) Analysis module 14

[0078] The analysis module 14 is used to receive the first data body sent by the forwarding module 11; and extract the warning parameter set from the first data body and store it locally; the analysis module 14 is also used to receive the first record sequence pushed by the first cache module 12; and perform multi-lead long signal splicing processing according to the first record sequence to generate a corresponding multi-lead long frame signal record; and perform electrocardiogram signal analysis processing according to the multi-lead long frame signal record to generate a corresponding multi-lead feature set; and perform warning data analysis processing according to the multi-lead feature set and the warning parameter set to generate a corresponding warning data cache record for the current time; and send the warning data cache record for the current time to the second cache module 13.

[0079] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to perform sequential signal splicing processing on multiple first lead short frame signals with the same first lead identifier in the first recording sequence in ascending order of the corresponding first short frame numbers to obtain corresponding first lead long frame signals when performing multi-lead long signal splicing processing according to the first recording sequence; and each first lead identifier and the corresponding first lead long frame signal constitute a corresponding first long frame data group; and the first short frame number with the largest value in the first recording sequence is used as the corresponding first long frame number; and the first long frame number and multiple first long frame data groups constitute a corresponding multi-lead long frame signal record.

[0080] For example, the acquisition device acquires 18-lead data, the specified number is 12, the specified segment length is 6 seconds, and the first record sequence consists of the first ECG data cache records 1-12; wherein, the first ECG data cache record 1: the first short frame number = 101, the 18 first short frame data groups are respectively the first lead identifiers (1-18) of the 1st to 18th leads and the 6-second long first lead short frame signal s 1,1 -s 1,18 ; First ECG data cache record 2: first short frame number = 102, 18 first short frame data groups are first lead identifiers (1-18) of leads 1-18 and 6 seconds long first lead short frame signals s 2,1 -s 2,18 ; Similarly, the first ECG data cache record 12: the first short frame number = 112, 18 first short frame data groups are the first lead identifiers (1-18) of the 1st to 18th leads and the 6-second first lead short frame signal s 12,1 -s 12,18 ;

[0081] Then, by performing sequential signal splicing processing on the multiple first lead short frame signals with the same first lead identifier in the first recording sequence in ascending order of the corresponding first short frame numbers, 18 first lead long frame signals can be obtained: the first lead long frame signal 1 is (the first lead short frame signal s 1,1 ...short frame signal of the first lead s 12,1 ), the first lead long frame signal 2 is (the first lead short frame signal s 1,2 ...short frame signal of the first lead s 12,2 ), and so on, the first lead long frame signal 18 is (the first lead short frame signal s 1,18 ...short frame signal of the first lead s 12,18); then 18 first long frame data groups can be obtained: the first long frame data group 1 is (first lead identifier 1, first lead long frame signal 1), the first long frame data group 2 is (first lead identifier 2, first lead long frame signal 2), and so on, the first long frame data group 18 is (first lead identifier 18, first lead long frame signal 18); and the first long frame number is max(101, 102…112)=112; the final multi-lead long frame signal record is: the first long frame number=112 and the first long frame data group 1-18.

[0082] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used for performing electrocardiographic signal analysis and processing according to multi-lead long-frame signal records.

[0083] Step A1, identifying square wave signal segments and dropout signal segments of each first lead long frame signal in the multi-lead long frame signal record, and performing signal splicing processing on the identified square wave signal segments and dropout signal segments based on the signal duration of the current first lead long frame signal to generate corresponding first lead square wave signals and first lead dropout signals;

[0084] Here, because the regular ECG signals are all sinusoidal wave signals, the square wave signal in the ECG signal should be regarded as an interference signal for identification, and the corresponding signal segment is called a square wave signal segment. When the acquisition device 2 performs signal acquisition, the electrode may fall off. In this case, the generated acquisition signal segment is called a falling signal segment; the analysis module 14 first needs to identify the square wave signal segment and the falling signal segment before analysis. The duration of the first lead square wave signal and the first lead falling signal is the same as the duration of the first lead long frame signal. If the duration of the first lead long frame signal is 72 seconds, the first lead square wave signal and the first lead falling signal are also 72 seconds long, but the amplitudes of the signal points without real square wave signal segments and falling signal segments on these two signals are set to the preset baseline amplitude;

[0085] Step A2, performing QRS wave group detection on each first lead long frame signal to generate a corresponding first lead QRS wave group sequence, taking the peak point of each QRS wave group in the first lead QRS wave group sequence as the corresponding first lead R point, and sorting all the obtained first lead R points in chronological order to generate a corresponding first lead R point sequence;

[0086] Here, each first lead long frame signal includes multiple heartbeat ECG signals. Under normal circumstances, each heartbeat ECG signal should be composed of P wave, QRS wave group and T wave, and the peak point of QRS wave group is R point. If the QRS wave group feature, P wave feature and T wave feature cannot be identified, the position of R point of each lead needs to be confirmed first, that is, the R point sequence of the first lead.

[0087] Step A3, performing noise filtering on the corresponding first lead R point sequence according to the first lead square wave signal and the first lead dropout signal corresponding to each first lead long frame signal to generate a corresponding second lead R point sequence; and performing R point fusion processing on the second lead R point sequences of all leads to generate a corresponding first R point time sequence; wherein the first R point time sequence includes multiple first R point times;

[0088] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to merge the second lead R points of the second lead R point sequences of all leads together to form a corresponding R point set when performing R point fusion processing on the second lead R point sequences of all leads; and cluster multiple R points in the R point set whose time interval is less than a preset minimum time threshold into the same R point group; and calculate the mean of the R point time of each R point group and use the calculation result as the corresponding first R point time; and generate a corresponding first R point time series by sorting all the obtained first R point times in chronological order;

[0089] Here, after obtaining the first lead R point sequence of each lead, it is necessary to use the first lead square wave signal and the first lead dropout signal as mask signals to perform noise elimination on each first lead R point sequence. Specifically, if a first lead R point in the first lead R point sequence is a non-baseline signal point on the first lead square wave signal or the first lead dropout signal, the first lead R point is deleted as a noise point; after completing the noise point elimination, the second lead R point sequence of all leads is subjected to R point fusion. In principle, the time of the signal point corresponding to the same real R point in each lead signal should be the same, but in reality, there is always a small deviation in the time of the signal point corresponding to the same real R point in each lead signal. The embodiment of the present invention is to take the average of the time of the signal point corresponding to the same real R point in each lead signal to achieve the purpose of reducing the error; the minimum time threshold used is a pre-set empirical value;

[0090] Step A4, and in each first lead long frame signal, taking each first R point time as the center, intercepting a segment of the ECG signal with a preset first duration forward and backward to form a corresponding first heartbeat signal segment, and intercepting a segment of the ECG signal with a preset second duration forward and backward to form a corresponding second heartbeat signal segment; and performing heartbeat quality assessment processing on each first heartbeat signal segment to generate a corresponding first quality level; and performing conventional heartbeat morphology classification processing on each second heartbeat signal segment to generate a corresponding first heartbeat type; and correcting the corresponding first heartbeat type according to each first quality level, and changing the corresponding first heartbeat type to an interference type when the first quality level is a weak interference heartbeat level or a strong interference heartbeat level; and sorting all the corrected first heartbeat types in chronological order to generate a corresponding first lead heartbeat type sequence;

[0091] The first duration is less than the second duration; the first quality level includes a normal heartbeat level, a weak interference heartbeat level and a strong interference heartbeat level; the first heartbeat type includes a sinus heartbeat type, an atrial fibrillation heartbeat type, an atrial heartbeat type, a ventricular heartbeat type and a junctional heartbeat type; the first heartbeat type corresponds to the first R point time;

[0092] Here, the first duration defaults to 1 second, and the second duration defaults to 2 seconds; the analysis module 14 takes the ECG signals of 1 second before and after each R point and splices them into a first heartbeat signal segment with a total length of 2 seconds, and performs a heartbeat quality assessment based on the first heartbeat signal segment to obtain one of three quality levels (normal heartbeat level, weak interference heartbeat level, and strong interference heartbeat level); takes the ECG signals of 2 seconds before and after each R point and splices them into a second heartbeat signal segment with a total length of 4 seconds, and performs a heartbeat morphology assessment based on the second heartbeat signal segment to obtain one of five conventional heartbeat morphologies (sinus heartbeat type, atrial fibrillation heartbeat type, atrial heartbeat type, etc.) type, ventricular beat type and junctional beat type); and the first beat type is corrected based on the first quality level, that is, if the first quality level corresponding to a certain R point is a weak interference beat level or a strong interference beat level, the first beat type corresponding to the point is corrected to an interference type; after step A4, a preliminary beat morphology type positioning can be performed on the single beat signal at each first R point time on each first lead long frame signal, and at this time, the first beat type of each beat signal is increased to 6 categories: sinus beat type, atrial fibrillation beat type, atrial beat type, ventricular beat type, junctional beat type and interference type;

[0093] It should be noted that a variety of evaluation methods can be used to implement the heartbeat quality assessment. One of them is to calculate the high-frequency interference energy ratio and the baseline interference energy ratio of the first heartbeat signal segment based on wavelet decomposition, and to calculate the weighted sum of the two, and to evaluate the quality level of the weighted sum based on the pre-set three-level interference energy ratio threshold range. Another is to use a classification model based on support vector machine (SVM) to perform three-level classification recognition and output the three-level classification probability and select the level corresponding to the maximum probability as the recognition result. Other implementation methods are not described here one by one. A variety of evaluation methods can also be used to implement the heartbeat morphology assessment. One of them is to pre-set the corresponding morphological feature parameter thresholds for sinus, atrial fibrillation, atrial, and ventricular junctional heartbeats in each lead and classify the second heartbeat signal segment of each lead based on the morphological feature parameter thresholds. Another is to use a convolutional neural network (Convolutional Neural Network)-based classification model. Networks, CNN) classification model is used for classification, and other implementation methods are not described one by one here; because there are cost differences between different implementation methods, the system of the embodiment of the present invention does not limit the specific implementation methods of heartbeat quality assessment and heartbeat morphology assessment, and only sets a standardized calling interface in advance for the analysis module 14 to perform heartbeat quality assessment and heartbeat morphology assessment. During specific implementation, the heartbeat quality assessment module and the heartbeat morphology assessment module are developed according to the optimal assessment method selected by the customer, and the development module is encapsulated by a standardized interface according to a preset standardized calling interface, and the encapsulated heartbeat quality assessment module and heartbeat morphology assessment module are loaded into the system of the embodiment of the present invention to realize that the analysis module 14 can call it;

[0094] Step A5, performing QRS complex, P wave and T wave feature recognition processing on each first lead long frame signal according to the first R point time sequence and the first lead heart beat type sequence to generate a corresponding first QRS complex feature sequence, first P wave feature sequence and first T wave feature sequence;

[0095] The first QRS wave group characteristic sequence includes a plurality of first QRS wave group characteristics; the first P wave characteristic sequence includes a plurality of first P wave characteristics; the first T wave characteristic sequence includes a plurality of first T wave characteristics; each of the first QRS wave group characteristics, the first P wave characteristics and the first T wave characteristics corresponds to the first R point time;

[0096] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used for, when performing QRS wave group, P wave and T wave feature recognition processing on each first lead long frame signal according to the first R point time sequence and the first lead heart beat type sequence, in each first lead long frame signal, recording the signal point corresponding to the first R point time corresponding to the first heart beat type that is not an interference type as a first-class point, and recording the signal point corresponding to the first R point time corresponding to the first heart beat type that is specifically an interference type as a second-class point; and identifying the starting and ending positions of the QRS wave group where each first-class point is located to generate the corresponding first QRS wave group starting point and the first QRS wave group ending point, and calculating the time difference between the first ending point and the first starting point to generate the corresponding first QRS wave group duration, and the first QRS wave group starting point, the first QRS wave group ending point and the first QRS wave group duration form the corresponding first QRS wave group feature; and setting the first QRS wave group feature corresponding to each second-class point to be Preset invalid features; and all the first QRS wave group features obtained are sorted in chronological order to generate the corresponding first QRS wave group feature sequence; and the peak point, starting point and ending point of the previous P wave of the QRS wave group where each first-class point is located are identified to generate the corresponding first P wave peak point, the first P wave starting point and the first P wave ending point to form the corresponding first P wave feature; and the first P wave feature corresponding to each second-class point is set as the preset invalid feature; and all the first P wave features obtained are sorted in chronological order to generate the corresponding first P wave feature sequence; and the peak point, starting point and ending point of the next T wave of the QRS wave group where each first-class point is located are identified to generate the corresponding first T wave peak point, the first T wave starting point and the first T wave ending point to form the corresponding first T wave feature; and the first T wave feature corresponding to each second-class point is set as the preset invalid feature; and all the first T wave features obtained are sorted in chronological order to generate the corresponding first T wave feature sequence;

[0097] Here, the start, end and width of the QRS complex, the start and end position of the P wave, and the start and end position of the T wave, which are not interference signals, are calculated based on the known ECG signal, R point time and heart beat type, so as to obtain the first QRS complex, P wave and T wave characteristic sequence;

[0098] Step A6, performing ventricular flutter and ventricular fibrillation feature recognition processing on each first lead long frame signal to generate a corresponding first lead unconventional feature sequence;

[0099] The first lead unconventional feature sequence includes a plurality of first ventricular flutter features and a plurality of first ventricular fibrillation features; the first ventricular flutter feature includes a first ventricular flutter starting point and a first ventricular flutter ending point; the first ventricular fibrillation feature includes a first ventricular fibrillation starting point and a first ventricular fibrillation ending point;

[0100] Here, the electrocardiogram of ventricular flutter, i.e. ventricular flutter, presents a fast and regular sine-like waveform, and the electrocardiogram of ventricular fibrillation, i.e. ventricular fibrillation, presents a baseline swing with irregular amplitude and shape, and the common feature of the two is the disappearance of the R wave; that is to say, the method of extracting the second heartbeat signal segment for heartbeat feature detection at each first R point time in the aforementioned step A4 cannot detect the characteristic position of ventricular flutter or ventricular fibrillation, so it is necessary to perform additional detection on these two types of unconventional heartbeat features; in the implementation process, there are a variety of detection methods based on the above-mentioned morphological characteristics of ventricular flutter and ventricular fibrillation to locate the characteristic position of ventricular flutter and ventricular fibrillation appearing in a segment of electrocardiogram, such as determining the starting and ending positions of ventricular flutter based on sinusoidal wave detection, and determining the starting and ending positions of ventricular flutter based on baseline swing amplitude difference. Or the square threshold of amplitude difference is used to determine the starting and ending positions of ventricular fibrillation, which will not be described one by one here; because there are various ways to identify ventricular flutter and ventricular fibrillation features, and their accuracy is also related to the software and hardware resources of the system, the system of the embodiment of the present invention does not limit the specific implementation method of ventricular flutter and ventricular fibrillation feature identification, but only sets a standardized calling interface in advance for the analysis module 14 to perform ventricular flutter and ventricular fibrillation feature identification. During specific implementation, the ventricular flutter and ventricular fibrillation feature identification module is developed according to the optimal identification method selected by the customer, and the development module is packaged with a standardized interface according to the pre-set standardized calling interface, and the packaged ventricular flutter and ventricular fibrillation feature identification module is loaded into the system of the embodiment of the present invention to realize that the analysis module 14 can call it;

[0101] Step A7, performing low voltage type setting processing on each first lead heartbeat type sequence according to the first R point time sequence;

[0102] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to record the first lead long frame signal corresponding to the current first lead heart beat type sequence as the current lead long frame signal when performing low voltage type setting processing on each first lead heart beat type sequence according to the first R point time sequence; and record the signal point corresponding to each first R point time in the current lead long frame signal as the corresponding first signal point; and record the first signal point whose signal point amplitude is lower than the preset low voltage amplitude as the corresponding second signal point; and record the first R point time corresponding to each second signal point as the corresponding second R point time; and set the first heart beat type corresponding to each second R point time in the current first lead heart beat type sequence to the corresponding low voltage type;

[0103] Here, the low voltage amplitude is a preset empirical value; the first heartbeat type after processing in step A7 is increased to 7 types: sinus heartbeat type, atrial fibrillation heartbeat type, atrial heartbeat type, ventricular heartbeat type, junctional heartbeat type, interference type and low voltage type;

[0104] Step A8, performing atrial fibrillation heartbeat type correction processing on each first lead heartbeat type sequence;

[0105] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to traverse any three consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial fibrillation heart beat type correction processing on each first lead heart beat type sequence; when traversing, the three first heart beat types currently traversed are recorded as corresponding front heart beat types, center heart beat types and back heart beat types in the order of front, middle and back; and confirm whether the center heart beat type is an atrial fibrillation heart beat type; if it is confirmed to be an atrial fibrillation heart beat type, confirm whether the front heart beat type and the back heart beat type are not an atrial fibrillation heart beat type; if it is confirmed that neither is an atrial fibrillation heart beat type, change the center heart beat type to a sinus heart beat type;

[0106] Here, the correction rule preset in the embodiment of the present invention is: if the heartbeat types of the two heartbeat signals before and after a single heartbeat signal identified as atrial fibrillation are not atrial fibrillation heartbeat types, then the type of the current heartbeat signal is corrected to the sinus heartbeat type;

[0107] Step A9, performing refinement processing on each first lead heart beat type sequence into atrial, ventricular and junctional heart beats according to the first R point time sequence;

[0108] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to traverse any two consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial, ventricular and junctional heart beat refinement processing on each first lead heart beat type sequence according to the first R point time sequence;

[0109] During traversal, the two first heartbeat types currently traversed are recorded as the corresponding front heartbeat type and back heartbeat type in the order of front and back respectively;

[0110] When both the preceding and succeeding heartbeat types are atrial heartbeat types, the absolute difference between the two first R point times corresponding to the preceding and succeeding heartbeat types is calculated to generate the corresponding first RR interval, and the heart rate is estimated according to the first RR interval to generate the corresponding first heart rate, and when the first heart rate is lower than a preset atrial escape heart rate threshold, the preceding and succeeding heartbeat types are changed to the corresponding atrial escape heartbeat types, and when the first heart rate is higher than a preset atrial tachycardia heart rate threshold, the preceding and succeeding heartbeat types are changed to the corresponding atrial tachycardia heartbeat types;

[0111] When both the preceding and succeeding heartbeat types are ventricular heartbeat types, the absolute difference between the two first R point times corresponding to the preceding and succeeding heartbeat types is calculated to generate a corresponding second RR interval, and the heart rate is estimated according to the second RR interval to generate a corresponding second heart rate, and when the second heart rate is lower than a preset ventricular escape heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding ventricular escape heartbeat types, and when the second heart rate is higher than a preset ventricular tachycardia heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding ventricular tachycardia heartbeat types;

[0112] When both the preceding and following heartbeat types are of the junctional heartbeat type, the absolute difference between the two first R point times corresponding to the preceding and following heartbeat types is calculated to generate a corresponding third RR interval, and the heart rate is estimated according to the third RR interval to generate a corresponding third heart rate, and when the third heart rate is lower than a preset junctional escape heart rate threshold, the preceding and following heartbeat types are changed to the corresponding junctional escape heartbeat types, and when the third heart rate is higher than a preset junctional tachycardia heart rate threshold, the preceding and following heartbeat types are changed to the corresponding junctional tachycardia types;

[0113] Here, the embodiment of the present invention stipulates that if two consecutive heartbeat signals are both identified as atrial (ventricular, junctional) heartbeat types, the heartbeat types of the two heartbeat signals need to be refined; during refinement, the corresponding RR interval is calculated based on the first R point time corresponding to the two heartbeats, and the corresponding heart rate is calculated based on the RR interval, and then the corresponding atrial (ventricular, junctional) escape heartbeat type subdivision or atrial (ventricular, junctional) tachycardia heartbeat type subdivision is performed based on a pair of atrial (ventricular, junctional) escape heartbeat rate thresholds and atrial tachycardia (ventricular tachycardia, junctional tachycardia) heart rate thresholds; the first heartbeat type after processing in step A8 increases to 10 categories: sinus heartbeat type, atrial fibrillation heartbeat type, atrial escape heartbeat type, atrial tachycardia heartbeat type, ventricular escape heartbeat type, ventricular tachycardia heartbeat type, junctional escape heartbeat type, junctional tachycardia heartbeat type, interference type and low voltage type; the three pairs of heart rate thresholds used here are all pre-set empirical values;

[0114] Step A10: The second lead R point sequence, the first QRS wave group characteristic sequence, the first P wave characteristic sequence, the first T wave characteristic sequence, the first lead unconventional characteristic sequence and the first lead heart beat type sequence corresponding to each first lead long frame signal constitute a corresponding first lead feature set; and all first lead feature sets constitute a corresponding multi-lead feature set.

[0115] In another specific implementation of the embodiment of the present invention, the analysis module 14 is specifically used to analyze and process the warning data according to the multi-lead feature set and the warning parameter set.

[0116] Step B1, performing a warning event and a warning event level recognition process according to a multi-lead feature set and a warning parameter set to generate a corresponding first recognition report;

[0117] The first identification report includes a plurality of first identification records; the first identification record includes a first event name, a first event level and a first event start time;

[0118] Here, because we have found in actual applications that different medical institutions have different definitions of warning events and corresponding event levels when processing ECG signal warnings, the system of the embodiment of the present invention does not limit the specific implementation methods of warning events and warning event level identification. It only sets a standardized calling interface in advance for the analysis module 14 to identify warning events and warning event levels. During specific implementation, the warning event and warning event level identification module are developed according to the identification method selected by the customer, and the warning event and warning event level identification module are packaged according to the pre-set standardized calling interface. The packaged warning event and warning event level identification module is loaded into the system of the embodiment of the present invention to realize that the analysis module 14 can call it; common warning event names such as atrial fibrillation events, ventricular flutter events, ventricular fibrillation events, ventricular tachycardia events, ventricular bradycardia events, extreme tachycardia events, extreme bradycardia events, etc. Different medical structures can also customize events and levels, which are not listed one by one here;

[0119] Step B2, equally divide the time interval of the first record sequence into a specified number to obtain a specified number of first time periods; and assign a corresponding first time period warning data group to each first time period; and set the first time period start and end time of the first time period warning data group according to the start and end time of each first time period, and initialize the first time period warning event and the first time period event level of each first time period warning data group to empty;

[0120] Here, each first identification report corresponds to a first record sequence pushed by the first cache module 12 to the analysis module 14; the first record sequence is composed of a specified number (default is 12) of first ECG data cache records, and each first ECG data cache record corresponds to a time period with a specified segment length (default is 6 seconds); each first time period obtained by equally dividing the time interval of the first record sequence by the system of the embodiment of the present invention naturally corresponds to the time period of each first ECG data cache record;

[0121] Step B3, and traverse each first time period; when traversing, record the first time period currently traversed as the current time period; and select the first identification record with the highest first event level from all first identification records whose first event start time meets the current time period as the current identification record; and set the first time period warning event and the first time period event level of the first time period warning data group corresponding to the current time period according to the first event name and the first event level of the current identification record;

[0122] Here, the traversal of each first time period is to perform the highest level screening on one or more first identification records that match within the current time period; during the traversal, if there is no first event start time that satisfies the first identification record of the current time period, it means that there is no matching first identification record within the current time period, and at this time, the first time period warning event and the first time period event level of the first time period warning data group corresponding to the current time period are both assumed to be empty; during the traversal, if there is a first event start time that satisfies the first identification record of the current time period, it means that there is a matching first identification record within the current time period, and at this time, the highest level is selected from the one or more matching first identification records as the current identification record, and the first time period warning event of the first time period warning data group corresponding to the current time period is set based on the first event name of the current identification record, and the first time period event level of the first time period warning data group corresponding to the current time period is set based on the first event level of the current identification record;

[0123] Step B4, when the traversal is finished, the obtained specified number of first time period warning data groups are sorted in the order of the corresponding time periods to generate the corresponding first warning data sequence; and the first long frame number and the first warning data sequence form the corresponding warning data cache record.

[0124] Here, after the traversal is completed, the warning events and event levels corresponding to each time period are set. By sorting them in order, a specified number (the default is 12) of time period warning data sequences, namely the first warning data sequence, can be obtained; because each first record sequence corresponds to a first long frame number and a first warning data sequence, the current warning data cache record used to display the analysis results of the first record sequence is composed of the first long frame number + the first warning data sequence.

[0125] It should be noted that it should be understood that the electrocardiogram signal analysis and processing system provided in the embodiment of the present invention is an electronic device, a server, a server system, a cloud platform or a microservice platform.

[0126] When the ECG signal analysis and processing system provided by the embodiment of the present invention is an electronic device or a server, it should be understood that the various modules of the ECG signal analysis and processing system are functional modules of the electronic device or server, and these functional modules can be fully or partially integrated into a physical entity, or physically separated. And these functional modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; they can also be partially implemented in the form of software called by processing elements, and partially implemented in the form of hardware. For example, the analysis module can be a separately established processing element, or it can be integrated in a chip of the above-mentioned electronic device or server for implementation. In addition, it can also be stored in the memory of the above-mentioned electronic device or server in the form of program code and called and executed by a processing element of the above-mentioned electronic device or server. The implementation of other functional modules is similar. In addition, these functional modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities.

[0127] When the ECG signal analysis and processing system provided by the embodiment of the present invention is a server system, it should be understood that the various modules of the ECG signal analysis and processing system are functional service ends of the server system, and these functional service ends can be a terminal device, a database or a server, or a terminal device cluster, a database cluster or a server cluster. When the ECG signal analysis and processing system provided by the embodiment of the present invention is a cloud platform or a microservice platform, it should be understood that the various modules of the ECG signal analysis and processing system are cloud service interfaces of the cloud platform or microservice interfaces of the microservice platform.

[0128] The embodiment of the present invention provides an electrocardiogram signal analysis and processing system, which can be connected to any electrocardiogram acquisition component or device. The system includes: a forwarding module, a first buffer module, a second buffer module and an analysis module. The above-mentioned ECG acquisition components or devices are collectively referred to as acquisition devices, and the forwarding module is responsible for receiving the upload data packets sent by the acquisition device and forwarding them to the cache queue of the first cache module for storage; the first cache module is responsible for regularly forming a first record sequence with a nearly specified number of cache records and pushing it to the analysis module; in order to improve the analysis accuracy, the analysis module performs long frame signal splicing based on the first record sequence, and performs ECG signal feature analysis and heart beat morphology feature analysis based on multi-lead long frame signals, and performs warning event and warning level identification based on the multi-lead feature set obtained by analysis to generate corresponding warning data cache records, and sends the warning data cache records to the cache queue of the second cache module for storage; the second cache module queues and stores each new warning data cache record received, and extracts the last warning data cache record from the queue and pushes it to the forwarding module together with the current warning data cache record in real time; after receiving the last and current warning data cache records, the forwarding module filters out the overlapping warning information in the current record that overlaps with the last record in the overlapping period, and sends the deduplicated warning data cache record as a downlink data packet to the acquisition device. By combining the system of the present invention with existing ECG acquisition components or devices on the market, the problem that bedside monitors cannot be widely deployed is solved, and the problem that existing ECG acquisition components or devices do not have the ability to analyze complex ECG signals is also solved.

[0129] Professionals should also be further aware that the systems, modules, units and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0130] The steps of the system, module, unit or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0131] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An electrocardiogram signal analysis and processing system, characterized in that: The system comprises: a forwarding module, a first buffer module, a second buffer module and an analysis module; the system is connected to a collection device; The forwarding module is connected to the acquisition device, the first buffer module, the second buffer module and the analysis module respectively; the forwarding module is used to receive the first uplink data packet sent by the acquisition device; and extract the first data type and the first data body from the first uplink data packet; and send the first data body to the analysis module when the first data type is the first type; and send the first data body to the first buffer module when the first data type is the second type; the forwarding module is also used to receive the current warning data cache record and the previous warning data cache record pushed by the second buffer module; and perform warning data deduplication processing on the current warning data cache record according to the previous warning data cache record to generate the corresponding first downlink data packet; and send the first downlink data packet to the acquisition device; the first data type includes the first type and the second type, when the first data type is the first type, the first data body is a warning parameter set, and when the first data type is the second type, the first data body consists of a first short frame number and a plurality of first short frame data groups; The first cache module is connected to the analysis module; the first cache module is used to store the ECG data cache queue; the first cache module is also used to receive the first data body sent by the forwarding module; and store the first data body in the ECG data cache queue as the latest first ECG data cache record; and obtain the latest specified number of the first ECG data cache records from the ECG data cache queue to form a corresponding first record sequence and push it to the analysis module; The second cache module is connected to the analysis module; the second cache module is used to store the warning data cache queue; the second cache module is also used to receive the current warning data cache record sent by the analysis module; and store the current warning data cache record into the warning data cache queue as the latest first warning data cache record; and extract the previous record of the current warning data cache record from the warning data cache queue as the corresponding previous warning data cache record; and push the current warning data cache record and the previous warning data cache record to the forwarding module; The analysis module is used to receive the first data body sent by the forwarding module; and extract the warning parameter set from the first data body and store it locally; the analysis module is also used to receive the first record sequence pushed by the first cache module; and perform multi-lead long signal splicing processing according to the first record sequence to generate a corresponding multi-lead long frame signal record; and perform electrocardiogram signal analysis processing according to the multi-lead long frame signal record to generate a corresponding multi-lead feature set; and perform warning data analysis processing according to the multi-lead feature set and the warning parameter set to generate the corresponding warning data cache record of the time; and send the warning data cache record of the time to the second cache module.

2. The electrocardiogram signal analysis and processing system according to claim 1, characterized in that: The forwarding module is connected to the acquisition device via a wired or wireless method; The wired mode at least includes a first wired mode based on a network cable, a second wired mode based on a coaxial data cable, a third wired mode based on a serial communication data cable, and a fourth wired mode based on a parallel communication data cable; The wireless methods include a first wireless method based on 3G / 4G / 5G / LTE mobile communication network, a second wireless method based on Bluetooth communication, a third wireless method based on wireless local area network, and a fourth wireless method based on near field communication NFC protocol.

3. The electrocardiogram signal analysis and processing system according to claim 1, characterized in that: The ECG data cache queue includes a plurality of the first ECG data cache records; the first ECG data cache record includes a first short frame number and a plurality of first short frame data groups; the first short frame data group includes a first lead identifier and a first lead short frame signal; The warning data cache queue includes a plurality of the first warning data cache records; the first warning data cache record includes a first long frame number and a first warning data sequence; the first warning data sequence includes a plurality of first time period warning data groups; The first period warning data group includes the start and end time of the first period, the first period warning event and the first period event level.

4. The electrocardiogram signal analysis and processing system according to claim 3, characterized in that: The forwarding module is specifically used to, when deduplicating warning data for the current warning data cache record according to the previous warning data cache record, record the first warning data sequences of the previous and current warning data cache records as the corresponding previous sequence and current sequence respectively; and record the two first time period warning data groups with the same start and end time of the first time period in the previous and current sequences as the corresponding first matching pairs; and traverse each of the first matching pairs; during traversal, if the two first time period warning events of the first matching pair currently traversed are the same and the two first time period event levels are the same, then the first time period warning data group belonging to the current sequence in the first matching pair currently traversed is recorded as a duplicate warning data group; at the end of the traversal, delete all the duplicate warning data groups in the current warning data cache record, and use the deleted current warning data cache record as the corresponding first downlink data packet.

5. The electrocardiogram signal analysis and processing system according to claim 3, characterized in that: The analysis module is specifically used for performing sequential signal splicing processing on a plurality of first lead short frame signals with the same first lead identifier in the first recording sequence in ascending order of the corresponding first short frame numbers to obtain corresponding first lead long frame signals when performing multi-lead long signal splicing processing according to the first recording sequence; and forming a corresponding first long frame data group from each of the first lead identifiers and the corresponding first lead long frame signal; and taking the first short frame number with the largest value in the first recording sequence as the corresponding first long frame number; and forming the corresponding multi-lead long frame signal record from the first long frame number and a plurality of the first long frame data groups.

6. The electrocardiogram signal analysis and processing system according to claim 5, characterized in that: The analysis module is specifically used for performing electrocardiographic signal analysis and processing according to the multi-lead long frame signal record. Identify square wave signal segments and dropout signal segments of each first lead long frame signal in the multi-lead long frame signal record, and perform signal splicing processing on the identified square wave signal segments and dropout signal segments based on the signal duration of the current first lead long frame signal to generate corresponding first lead square wave signals and first lead dropout signals; and performing QRS wave group detection on each of the first lead long frame signals to generate a corresponding first lead QRS wave group sequence, and taking the peak point of each QRS wave group in the first lead QRS wave group sequence as the corresponding first lead R point, and sorting all the obtained first lead R points in chronological order to generate a corresponding first lead R point sequence; And according to the first lead square wave signal and the first lead drop-off signal corresponding to each of the first lead long frame signals, the corresponding first lead R point sequence is subjected to noise filtering to generate a corresponding second lead R point sequence; and the second lead R point sequence of all leads is subjected to R point fusion processing to generate a corresponding first R point time sequence; wherein the first R point time sequence includes multiple first R point times; And in each of the first lead long frame signals, taking each of the first R point times as the center, a segment of the electrocardiogram signal with a preset first duration is intercepted forward and backward to form a corresponding first heartbeat signal segment, and a segment of the electrocardiogram signal with a preset second duration is intercepted forward and backward to form a corresponding second heartbeat signal segment; and a heartbeat quality assessment process is performed on each of the first heartbeat signal segments to generate a corresponding first quality grade; and a conventional heartbeat morphology classification process is performed on each of the second heartbeat signal segments to generate a corresponding first heartbeat type; and the corresponding first heartbeat is classified according to each of the first quality grades. The first heartbeat type is corrected, and when the first quality level is a weak interference heartbeat level or a strong interference heartbeat level, the corresponding first heartbeat type is changed to an interference type; and all the first heartbeat types after correction are sorted in chronological order to generate a corresponding first lead heartbeat type sequence; wherein, the first duration is less than the second duration; the first quality level includes a normal heartbeat level, a weak interference heartbeat level and a strong interference heartbeat level; the first heartbeat type includes a sinus heartbeat type, an atrial fibrillation heartbeat type, an atrial heartbeat type, a ventricular heartbeat type and a junctional heartbeat type; the first heartbeat type corresponds to the first R point time; and performing QRS complex, P wave and T wave feature recognition processing on each of the first lead long frame signals according to the first R point time sequence and the first lead heart beat type sequence to generate a corresponding first QRS complex feature sequence, a first P wave feature sequence and a first T wave feature sequence; wherein the first QRS complex feature sequence includes a plurality of first QRS complex features; the first P wave feature sequence includes a plurality of first P wave features; the first T wave feature sequence includes a plurality of first T wave features; each of the first QRS complex feature, the first P wave feature and the first T wave feature corresponds to the first R point time; and performing ventricular flutter and ventricular fibrillation feature recognition processing on each of the first lead long frame signals to generate a corresponding first lead unconventional feature sequence; wherein the first lead unconventional feature sequence includes a plurality of first ventricular flutter features and a plurality of first ventricular fibrillation features; the first ventricular flutter feature includes a first ventricular flutter starting point and a first ventricular flutter ending point; the first ventricular fibrillation feature includes a first ventricular fibrillation starting point and a first ventricular fibrillation ending point; and performing low voltage type setting processing on each of the first lead heart beat type sequences according to the first R point time sequence; and performing atrial fibrillation heart beat type correction processing on each of the first lead heart beat type sequences; and performing atrial, ventricular and junctional heart beat refinement processing on each of the first lead heart beat type sequences according to the first R point time sequence; The second lead R point sequence, the first QRS wave group characteristic sequence, the first P wave characteristic sequence, the first T wave characteristic sequence, the first lead unconventional characteristic sequence and the first lead heart beat type sequence corresponding to each of the first lead long frame signals constitute a corresponding first lead feature set; and all of the first lead feature sets constitute the corresponding multi-lead feature set.

7. The electrocardiogram signal analysis and processing system according to claim 6, characterized in that: The analysis module is specifically configured to merge the second lead R points of the second lead R point sequences of all leads together to form a corresponding R point set when performing R point fusion processing on the second lead R point sequences of all leads; and cluster multiple R points in the R point set whose time interval is less than a preset minimum time threshold into the same R point group; and calculating the mean of the R point times of each of the R point groups and using the calculated result as the corresponding first R point time; And all the first R point times obtained are sorted in chronological order to generate the corresponding first R point time series.

8. The electrocardiogram signal analysis and processing system according to claim 6, characterized in that: The analysis module is specifically used for performing QRS complex, P wave and T wave feature recognition processing on each of the first lead long frame signals according to the first R point time sequence and the first lead heart beat type sequence, In each of the first lead long frame signals, a signal point corresponding to the first R point time corresponding to the first heart beat type that is not an interference type is recorded as a first-category point, and a signal point corresponding to the first R point time corresponding to the first heart beat type that is specifically an interference type is recorded as a second-category point; and identifying the starting and ending positions of the QRS complex where each of the first-class points is located to generate a corresponding first QRS complex starting point and a first QRS complex ending point, and calculating a time difference between the first QRS complex ending point and the first QRS complex starting point to generate a corresponding first QRS complex duration, and forming a corresponding first QRS complex feature with the first QRS complex starting point, the first QRS complex ending point and the first QRS complex duration; and setting the first QRS wave group feature corresponding to each of the second type points as a preset invalid feature; and generating a corresponding first QRS wave group feature sequence by sorting all the obtained first QRS wave group features in chronological order; And the peak point, starting point and ending point of the previous P wave of the QRS wave group where each of the first-class points is located are identified to generate the corresponding first P wave peak point, first P wave starting point and first P wave ending point to form the corresponding first P wave feature; and the first P wave feature corresponding to each of the second-class points is set as a preset invalid feature; and all the obtained first P wave features are sorted in chronological order to generate the corresponding first P wave feature sequence; And the peak point, starting point and ending point of the next T of the QRS wave group where each of the first-class points is located are identified to generate the corresponding first T wave peak point, first T wave starting point and first T wave ending point to form the corresponding first T wave feature; and the first T wave feature corresponding to each of the second-class points is set as a preset invalid feature; and all the obtained first T wave features are sorted in chronological order to generate the corresponding first T wave feature sequence.

9. The electrocardiogram signal analysis and processing system according to claim 6, characterized in that: The analysis module is specifically used for recording the first lead long frame signal corresponding to the current first lead heart beat type sequence as the current lead long frame signal when the low voltage type setting processing is performed on each of the first lead heart beat type sequences according to the first R point time sequence; recording the signal point corresponding to each of the first R point times in the current lead long frame signal as the corresponding first signal point; recording the first signal point whose signal point amplitude is lower than the preset low voltage amplitude as the corresponding second signal point; and recording the first R point time corresponding to each of the second signal points as the corresponding second R point time; And the first heart beat type corresponding to each second R point time in the current first lead heart beat type sequence is set to the corresponding low voltage type.

10. The electrocardiogram signal analysis and processing system according to claim 6, characterized in that: The analysis module is specifically used to traverse any three consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial fibrillation heart beat type correction processing on each of the first lead heart beat type sequences; when traversing, the three first heart beat types currently traversed are recorded as corresponding front heart beat type, center heart beat type and back heart beat type in the order of front, middle and back; and confirm whether the center heart beat type is an atrial fibrillation heart beat type; if it is confirmed to be an atrial fibrillation heart beat type, confirm whether the front heart beat type and the back heart beat type are not an atrial fibrillation heart beat type; if it is confirmed that neither of them is an atrial fibrillation heart beat type, change the center heart beat type to a sinus heart beat type.

11. The electrocardiogram signal analysis and processing system according to claim 6, characterized in that: The analysis module is specifically used for traversing any two consecutive first heart beat types in the current first lead heart beat type sequence when performing atrial, ventricular and junctional heart beat refinement processing on each first lead heart beat type sequence according to the first R point time sequence; when traversing, the two first heart beat types currently traversed are recorded as corresponding front heart beat type and back heart beat type in front and back order respectively; and when the front and back heart beat types are both atrial heart beat types, the absolute difference calculation is performed on the two first R point times corresponding to the front and back heart beat types to generate the corresponding first RR interval, and the heart rate is estimated according to the first RR interval to generate the corresponding first heart rate, and when the first heart rate is lower than a preset atrial escape heart rate threshold, the front and back heart beat types are changed to the corresponding atrial escape heart beat type, and when the first heart rate is higher than a preset atrial tachycardia heart rate threshold, the front and back heart beat types are changed to the corresponding atrial tachycardia heart beat type; and when the front and back heart beat types are both ventricular heart beat types, An absolute difference calculation is performed on the two first R point times corresponding to the preceding and succeeding heartbeat types to generate a corresponding second RR interval, and a heart rate estimation is performed based on the second RR interval to generate a corresponding second heart rate, and when the second heart rate is lower than a preset ventricular escape heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding ventricular escape heartbeat types, and when the second heart rate is higher than a preset ventricular tachycardia heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding ventricular tachycardia heartbeat types; and when both the preceding and succeeding heartbeat types are junctional heartbeat types, an absolute difference calculation is performed on the two first R point times corresponding to the preceding and succeeding heartbeat types to generate a corresponding third RR interval, and a heart rate estimation is performed based on the third RR interval to generate a corresponding third heart rate, and when the third heart rate is lower than a preset junctional escape heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding junctional escape heartbeat types, and when the third heart rate is higher than a preset junctional tachycardia heart rate threshold, the preceding and succeeding heartbeat types are changed to corresponding junctional tachycardia heartbeat types.

12. The electrocardiogram signal analysis and processing system according to claim 5, characterized in that: The analysis module is specifically used for performing early warning data analysis and processing according to the multi-lead feature set and the early warning parameter set. Performing warning event and warning event level identification processing according to the multi-lead feature set and the warning parameter set to generate a corresponding first identification report; wherein the first identification report includes a plurality of first identification records; the first identification record includes a first event name, a first event level and a first event start time; and equally divide the time interval of the first record sequence according to the specified number to obtain the specified number of first time periods; and assign a corresponding first time period warning data group to each first time period; and set the first time period start and end time of the first time period warning data group according to the start and end time of each first time period, and initialize the first time period warning event and the first time period event level of each first time period warning data group to null; and traverse each of the first time periods; during the traversal, record the first time period currently traversed as the current time period; and select the first identification record with the highest first event level from all the first identification records whose first event start time meets the current time period as the current identification record; and set the first time period warning event and the first time period event level of the first time period warning data group corresponding to the current time period according to the first event name and the first event level of the current identification record; At the end of the traversal, the specified number of the first time period warning data groups are sorted in the order of the corresponding time periods to generate the corresponding first warning data sequence; and the first long frame number and the first warning data sequence form the corresponding warning data cache record.

Images