A ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring

Abstract

The present invention relates to the technical field of electrocardiogram baseline drift, and specifically relates to a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring. The system includes a data acquisition module for acquiring waveform segments; a baseline analysis module for determining the baseline position by combining the amplitude changes based on the different continuous contribution degrees of different trend segments on the waveform segments; an adjustment analysis module for determining the drift adjustment segment according to the continuous baseline position changes, and determining the adjustment degree through the amplitude fluctuations and duration deviations between the waveform segments in the drift adjustment segment; and an identification module for adjusting the waveform segments based on the adjustment degree and the baseline position situation of the drift adjustment segment to obtain an adjusted electrocardiogram signal for ventricular fibrillation identification. The present invention determines the adjustment of the baseline position by superimposing the contributions of different trend changes in the waveform, and adjusts the drift correction degree by the deviations between the waveform segments in the drift adjustment segment, improving the accuracy of electrocardiogram signal drift adjustment, retaining the signal characteristics, and making the subsequent ventricular fibrillation identification more accurate and efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of electrocardiogram baseline drift, and particularly relates to a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring. Background Art

[0002] Critical care electrocardiogram monitoring is an extremely important part of clinical medicine. By real-time monitoring of the patient's electrocardiogram, various arrhythmias can be detected and processed in a timely manner, thereby effectively preventing the occurrence of serious events such as cardiac arrest. As a lethal arrhythmia, ventricular fibrillation often occurs with great suddenness and unpredictability. Accurately and quickly identifying ventricular fibrillation signals is of great significance for clinical monitoring and ensuring the patient's life safety.

[0003] During the process of identifying ventricular fibrillation signals, it will be affected by various factors, such as the patient's breathing, poor contact between the electrode and the skin, etc., resulting in baseline drift problems in the collected electrocardiogram signals. Especially in the electrocardiogram signals of critically ill patients, it may make the characteristic waveforms of ventricular fibrillation become blurred or difficult to identify, thus increasing the difficulty of identifying ventricular fibrillation signals. Since the drift problem accumulates over time, there is generally a problem that the timeliness and accuracy of baseline filtering adjustment do not meet the requirements in critical care electrocardiogram monitoring, affecting the identification of the patient's ventricular fibrillation signals. Summary of the Invention

[0004] In order to solve the technical problem that in the prior art, due to the drift problem accumulating over time, there is a lack of timeliness in general baseline filtering adjustment in critical care electrocardiogram monitoring, which affects the identification of the patient's ventricular fibrillation signals, the purpose of the present invention is to provide a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring. The specific technical solutions adopted are as follows:

[0005] The present invention provides a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring, and the system includes:

[0006] A data acquisition module, configured to acquire the electrocardiogram signal curve of the patient; determine waveform segments according to the change trend of the amplitude in the electrocardiogram signal curve; there is the same trend segment between every two adjacent waveform segments;

[0007] A baseline analysis module, configured to obtain the baseline position contribution degree of each trend segment according to the time and waveform duration on each trend segment in each waveform segment; determine the baseline position of each waveform segment based on the baseline position contribution degree of the trend segment in each waveform segment and combining the amplitude change of the trend segment;

[0008] An adjustment analysis module, configured to determine the drift adjustment segment in the electrocardiogram signal curve according to the continuous change of the baseline positions of the waveform segments in time sequence; obtain the adjustment degree of each waveform segment according to the deviation degree of the amplitude fluctuation and duration between each waveform segment in the drift adjustment segment and the subsequent waveform segment;

[0009] An identification module, which is used to adjust the amplitude in the waveform segment in the electrocardiogram signal curve based on the adjustment degree of each waveform segment in the drift adjustment segment and the baseline position of the waveform segment before the drift adjustment segment, so as to obtain an adjusted electrocardiogram signal; and perform ventricular fibrillation identification based on the adjusted electrocardiogram signal.

[0010] Further, the method for obtaining the baseline position contribution degree includes:

[0011] For any trend segment, the length of the electrocardiogram signal curve on this trend segment is used as the waveform duration index of this trend segment;

[0012] Combining the time period length and the waveform duration index of this trend segment, the baseline position contribution degree of this trend segment is obtained.

[0013] Further, the method for determining the baseline position includes:

[0014] For any waveform segment, the proportion of the baseline position contribution degree of each trend segment in this waveform segment in all baseline position contribution degrees is used as the contribution degree of each trend segment;

[0015] The average value of the maximum amplitude and the minimum amplitude of each trend segment in this waveform segment is used as the central amplitude of each trend segment;

[0016] Based on the contribution degree of each trend segment in this waveform segment, the central amplitude is weighted and summed to obtain the baseline position of this waveform segment.

[0017] Further, the method for determining the drift adjustment segment includes:

[0018] Arrange the baseline positions of the waveform segments in chronological order to obtain a baseline sequence; the difference between each baseline position and the subsequent baseline position in the baseline sequence is used as the position deviation degree of each baseline position; in the order of the baseline sequence, the position deviation degrees of the baseline positions are accumulated to obtain a cumulative offset degree;

[0019] When the cumulative offset degree is greater than a preset offset threshold, stop accumulating, and use the waveform segments of all baseline positions except the last baseline position that participate in the accumulation as stable waveform segments;

[0020] According to the baseline position deviation situation between the baseline position after the stable wave band and the baseline position of the stable wave band in the baseline sequence, determine a new accumulation position; in the baseline sequence, continue iterative accumulation backward from the new accumulation position until all stable waveform segments are determined;

[0021] The time periods corresponding to all unstable waveform segments are used as the drift adjustment segments.

[0022] Further, determining a new accumulation position according to the deviation between the baseline position after the stable band and the baseline position of the stable band in the baseline sequence includes:

[0023] After the stable waveform segment of the baseline sequence, calculate the difference between each baseline position and the first baseline position in the baseline sequence to obtain an offset judgment index;

[0024] When the offset judgment index is less than or equal to a preset offset threshold, use the corresponding baseline position as the new accumulation position.

[0025] Further, the method for obtaining the adjustment degree includes:

[0026] Take the amplitude range in each waveform segment as the fluctuation index of each waveform segment; for any waveform segment in the drift adjustment segment, take the difference between the waveform index of this waveform segment and the waveform segment in the previous time sequence as the fluctuation deviation degree of this waveform segment;

[0027] Take the difference in the time period length between this waveform segment and the waveform segment in the previous time sequence as the time deviation degree of this waveform segment;

[0028] Combine the fluctuation deviation degree and the time deviation degree of this waveform segment to obtain the adjustment degree of this waveform segment.

[0029] Further, the method for obtaining the adjusted electrocardiogram signal includes:

[0030] For any waveform segment in the drift adjustment segment, determine a reference baseline position in the non-drift adjustment segment before this waveform segment;

[0031] Take the difference between the reference baseline position and the baseline position of this waveform segment as the correction degree of this waveform segment; take the product of the correction degree and the adjustment degree of this waveform segment as the adjustment value of this waveform segment; add each amplitude on this waveform segment to the adjustment value to obtain the adjusted waveform segment of this waveform segment;

[0032] Arrange the waveform segments of the non-drift adjustment segment and the adjusted waveform segments in time sequence to form an adjusted electrocardiogram signal.

[0033] Further, the method for obtaining the reference baseline position includes:

[0034] Among all the non-drift adjustment segments before this waveform segment, take the waveform segment with the smallest time distance as the reference waveform segment of this waveform segment; take the baseline position of the reference waveform segment as the reference baseline position of this waveform segment.

[0035] Further, the method for obtaining the waveform segment includes:

[0036] Obtain all the extreme points in the electrocardiogram signal curve; take the time period between two adjacent extreme points as a trend segment; take every two adjacent trend segments as a waveform segment.

[0037] Further, the ventricular fibrillation recognition based on the adjusted electrocardiogram signal includes:

[0038] Input the adjusted electrocardiogram signal into the trained ventricular fibrillation signal recognition model and output the recognition result.

[0039] The present invention has the following beneficial effects:

[0040] When obtaining the waveform segment, the present invention improves the accuracy of the baseline change between subsequent different waveforms through the superposition time period analysis. By the different continuous contribution degrees of different trend segments on the waveform segment, the determination of the baseline position is adjusted to improve the accuracy of the baseline position, providing a more accurate data basis for the determination of subsequent drift conditions. After determining the drift segment according to the continuous baseline position change, considering the possible heartbeat fluctuation of the patient itself needs to be retained, the adjustment degree is determined through the amplitude fluctuation and duration deviation between the waveform segment and the subsequent waveform segment, so that the subsequent baseline drift adjustment can retain more floating information of critically ill patients and improve the signal recognition accuracy. Finally, based on the adjustment degree and the baseline position situation before the drift segment, the waveform segment is adjusted to obtain an adjusted electrocardiogram signal for more accurate ventricular fibrillation recognition. The present invention adjusts the determination of the baseline position by superimposing the contributions of different trend changes in the waveform, and adjusts the drift correction degree with the deviation between waveform segments in the drift adjustment segment, improving the accuracy of electrocardiogram signal drift adjustment, retaining signal characteristics and making subsequent ventricular fibrillation recognition more accurate and efficient. Description of the Drawings

[0041] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0042] Figure 1 It is a structural block diagram of a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring provided by an embodiment of the present invention;

[0043] Figure 2 It is a schematic diagram of an electrocardiogram signal curve provided by an embodiment of the present invention;

[0044] Figure 3 It is a schematic diagram of a waveform segment provided by an embodiment of the present invention;

[0045] Figure 4 It is a schematic diagram of a trend segment provided by an embodiment of the present invention;

[0046] Figure 5 A schematic diagram of the baseline position distribution provided by an embodiment of the present invention. Detailed implementation manners

[0047] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring according to the present invention, including its specific implementation manners, structures, features, and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0049] The following specifically describes the specific solution of a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring provided by the present invention with reference to the accompanying drawings.

[0050] Please refer to Figure 1 , which shows a block diagram of the structure of a ventricular fibrillation signal recognition system for critical care electrocardiogram monitoring provided by an embodiment of the present invention. The system includes: a data acquisition module 101, a baseline analysis module 102, an adjustment analysis module 103, and a recognition module 104.

[0051] The data acquisition module 101 is used to obtain the electrocardiogram signal curve of the patient; determine the waveform segments according to the change trend of the amplitude in the electrocardiogram signal curve; there is the same trend segment between every two adjacent waveform segments.

[0052] In the embodiment of the present invention, an electrocardiograph is used to collect the electrocardiogram signal of the patient, and perform processing such as filtering, amplification, and analog-to-digital conversion on it, converting the electrical activity of the heart into a visual waveform diagram as the electrocardiogram, that is, the electrocardiogram signal curve of the patient. Please refer to Figure 2 , which shows a schematic diagram of an electrocardiogram signal curve provided by an embodiment of the present invention.

[0053] When monitoring the electrocardiogram signal of a critically ill patient, it may be affected by various other factors, resulting in baseline drift problems in the collected electrocardiogram signal, making it impossible to identify ventricular fibrillation signals. Therefore, based on the analysis of the collected electrocardiogram signal of the patient, the signal segments with baseline drift problems in the entire electrocardiogram signal are judged, and these fluctuations are further adjusted according to the trend of the signal segments, so as to help complete the identification of ventricular fibrillation signals. In a normal electrocardiogram, all heartbeat waveforms basically fluctuate up and down around a baseline. Therefore, the waveforms are first divided.

[0054] In an embodiment of the present invention, all extreme points in the electrocardiogram signal curve are obtained. The extreme points include maximum points and minimum points, and the maximum points and minimum points are alternately distributed in time sequence in the fluctuating signal. Therefore, the time period between two adjacent extreme points is used as a trend segment. The trend segment from the maximum point to the minimum point is a downward trend, and the trend segment from the minimum point to the maximum point is an upward trend.

[0055] Taking every two adjacent trend segments as a waveform segment, the waveform from the maximum point to the minimum point and then to the maximum point shows a concave waveform segment, and the waveform from the minimum point to the maximum point and then to the minimum point shows a convex waveform segment. Therefore, when continuously dividing the waveform, due to the alternating distribution of convex and concave shapes, there is the same trend segment between every two adjacent waveform segments. Through the overlapping analysis of the waveforms, the accuracy of the subsequent change analysis is higher after the baseline position is determined. Please refer to Figure 3 , which shows a schematic diagram of a waveform segment provided by an embodiment of the present invention. The waveform segment on the left is convex, and the waveform segment on the right is concave.

[0056] The baseline analysis module 102 is configured to obtain the baseline position contribution degree of each trend segment according to the time and waveform duration on each trend segment in each waveform segment; and determine the baseline position of each waveform segment based on the baseline position contribution degree of the trend segments in each waveform segment and in combination with the amplitude change of the trend segments.

[0057] When the patient's heartbeat is unstable, that is, at this time, problems such as poor device contact and patient movement occur, resulting in different durations of each heartbeat waveform of the patient collected. Therefore, there may be a deviation in determining the baseline position only according to the amplitude change. Therefore, different trend segments in each waveform segment are analyzed, and the contribution degree of the trend segments to the position determination is analyzed according to the duration of the trend segments in terms of time and waveform.

[0058] Preferably, in an embodiment of the present invention, the method for obtaining the baseline position contribution degree includes:

[0059] For any one trend segment, the length of the electrocardiogram signal curve on this trend segment is used as the waveform duration index of this trend segment. The size of the length of the electrocardiogram signal curve reflects the continuous trend of the heartbeat waveform. When the waveform duration index is larger, it indicates that the proportion participating in the baseline determination is higher.

[0060] Further combine the time period length of the trend segment and the waveform duration index to obtain the baseline position contribution degree of the trend segment. In the embodiments of the present invention, the product of the time period length of the trend segment and the waveform duration index is normalized to obtain the baseline position contribution degree of the trend segment. When the waveform duration on one side of the waveform is shorter, it indicates that the patient's heartbeat is unstable at this point and the contribution degree to the baseline is small. When the waveform length on one side is small, it indicates that the patient's heartbeat intensity is weak at this point and the contribution degree to the baseline is small. Please refer to Figure 4 , which shows a schematic diagram of a trend segment provided by an embodiment of the present invention.

[0061] It should be noted that the normalization process is a well-known technical means to those skilled in the art. The normalization selection can be standard normalization or linear normalization, etc., which will not be limited and elaborated here.

[0062] After analyzing the contribution degrees of different trend segments participating in the baseline determination, through contribution degree weighting, based on the change of the signal amplitude, the baseline amplitude of the up and down changes of each waveform is determined, that is, the baseline position. The baseline position is also a position of an amplitude size in the signal.

[0063] Preferably, in the embodiments of the present invention, the method for determining the baseline position includes:

[0064] For any waveform segment, the proportion of the baseline position contribution degree of each trend segment in the waveform segment in all baseline position contribution degrees is used as the contribution degree of each trend segment, and the participation size of the trend segment in all trend segments is reflected by the proportion size.

[0065] Further, the mean value of the maximum amplitude and the minimum amplitude of each trend segment in the waveform segment is used as the central amplitude of each trend segment to reflect the fluctuation center situation in each trend segment. Finally, the central amplitude is weighted based on the contribution degree of each trend segment in the waveform segment, and the sum value is obtained to obtain the baseline position of the waveform segment. Combining the performance of each trend segment on the waveform, the distribution baseline situation of the waveform is determined. As an example, the method for obtaining the baseline position includes:

[0066] ;

[0067] In the formula, represents the baseline position of the th waveform segment, represents the total number of trend segments in the waveform segment, represents the central amplitude of the th trend segment in the th waveform segment, represents the baseline position contribution degree of the th trend segment in the th waveform segment, Denoted as the contribution of the th trend segment in the

[0068] Thus, the analysis of the baseline position of each waveform is completed.

[0069] The adjustment analysis module 103 is used to determine the drift adjustment segment in the electrocardiogram signal curve according to the continuous change of the baseline position of the waveform segments in time series; and obtain the adjustment degree of each waveform segment according to the deviation degree of the amplitude fluctuation and duration between each waveform segment and the subsequent waveform segment in the drift adjustment segment.

[0070] When the patient is in a normal and stable state, the heartbeat is also relatively stable, which is reflected in the electrocardiogram signal that the baseline signals of adjacent fluctuations are in the same horizontal direction. However, when the electrocardiograph is not in good contact with the patient's skin, or the patient moves, the electrocardiogram signal collected at this time is unstable, and the electrocardiogram shows vertical drift, and the baseline signal also changes continuously. Please refer to Figure 5 , which shows a schematic diagram of the baseline position distribution provided by an embodiment of the present invention. Therefore, when the distribution instability of the baseline position is relatively high, the baseline drift problem may occur. In order to adjust the drift part more accurately, the time period of drift adjustment is first determined.

[0071] Preferably, in the embodiment of the present invention, the method for obtaining the drift adjustment segment includes:

[0072] First, arrange the baseline positions of the waveform segments in chronological order to obtain a baseline sequence, and use the difference between each baseline position in the baseline sequence and the subsequent baseline position as the position deviation degree of each baseline position, which reflects the possibility of the baseline position shifting. For the last baseline position without a subsequent baseline position, the position deviation degree of the previous baseline position can be used as the position deviation degree.

[0073] Accumulate the position deviation degrees of the baseline positions in the order of the baseline sequence to obtain the cumulative deviation degree. Judge the stability of the baseline deviation by the degree of continuous deviation. When the cumulative deviation degree is greater than the preset deviation threshold, stop accumulating, and use the waveform segments of all baseline positions except the last baseline position participating in the accumulation as stable waveform segments. Considering the possible interference of the patient's electrocardiogram signal, set threshold analysis. When the cumulative situation breaks through the threshold, it means that the drift is more obvious, and the part that did not break through the threshold during the previous accumulation can be used as the time period of the stable situation for analysis. In the embodiment of the present invention, the preset deviation threshold is set to 5, and the specific value can be adjusted by the implementer according to the specific implementation scenario, which is not limited here.

[0074] Further, according to the deviation between the baseline position after the stable wave band and the baseline position of the stable wave band in the baseline sequence, a new accumulation position is determined. Since the drift behavior is a slow process, there may be a normal backward shift after the drift behavior persists for a period of time. Therefore, subsequently, whether there is a continuous situation of a new stable waveform segment is determined through the baseline position deviation.

[0075] In the embodiment of the present invention, the method for obtaining the new accumulation position includes:

[0076] After the stable waveform segment of the baseline sequence, calculate the difference between each baseline position and the first baseline position in the baseline sequence to obtain an offset judgment index. Since the drift is a continuous process, after the drift that exceeds the threshold occurs, it will be manifested as a situation of continuous high drift, that is, a situation where the offset judgment index is high.

[0077] Therefore, when the offset judgment index is less than or equal to the preset offset threshold, it indicates that the drift situation has shifted back, and drift analysis can be restarted. Therefore, the corresponding baseline position is used as the new accumulation position. In the baseline sequence, continue to perform iterative accumulation backward from the new accumulation position until all stable waveform segments are determined. Perform accumulation analysis and threshold judgment on each subsequent unanalyzed baseline position in the order of the baseline sequence until the baseline sequence is traversed. At this time, all stable waveforms with insignificant drift are determined.

[0078] Then, the time periods corresponding to all non-stable waveform segments are used as drift adjustment segments. Correct the signal waveform with significant drift, and consider that when correcting and adjusting, since there may be more characteristic information in the heartbeat fluctuations of critically ill patients, the loss of important information needs to be reduced.

[0079] When the patient is affected by other factors and the collected electrocardiogram signal shows baseline drift, some characteristics of the patient's heart beating will not change, such as: the heart rate and the change in the heart beating intensity over a period of time. Therefore, through the deviation between the waveforms in the drift adjustment segment, the correction degree of each waveform is adjusted to retain more original characteristic information.

[0080] Preferably, in the embodiment of the present invention, the method for obtaining the adjustment degree includes:

[0081] First, take the amplitude range in each waveform segment as the fluctuation index of each waveform segment to reflect the fluctuation situation of the waveform. For any waveform segment in the drift adjustment segment, take the difference between the waveform index of this waveform segment and the waveform segment in the previous time sequence as the fluctuation deviation degree of this waveform segment. When the amplitude difference between adjacent fluctuations is small, it indicates that the patient's heart beating is relatively stable after excluding the influence of other factors at these two moments. On the contrary, this indicates that the patient's heart beating has indeed fluctuated at this time, and this performance should be appropriately retained when eliminating the influence of baseline drift on it subsequently.

[0082] Further, the time period length difference between this waveform segment and the previous waveform segment in terms of time sequence is used as the time deviation degree of this waveform segment. When the patient's heartbeat performance is stable, after being interfered by external factors, even if baseline drift occurs, the floating performance of adjacent waveforms still remains relatively similar, that is, the durations are also relatively similar.

[0083] Finally, by combining the fluctuation deviation degree and the time deviation degree of this waveform segment, the adjustment degree of this waveform segment is obtained. In the embodiment of the present invention, the product of the fluctuation deviation degree and the time deviation degree of this waveform segment is subjected to negative correlation mapping and normalization processing to obtain the adjustment degree of this waveform segment. When the heartbeat of the patient within adjacent times shows small amplitude fluctuation differences and small duration differences, it indicates that the patient's heartbeat performance is relatively stable at this time. Only due to external factor interference, there is a baseline drift problem. When correcting the amplitude on this waveform, the greater the adjustment degree. On the contrary, when the differences are greater, it indicates that the waveform drift contains the characteristic information of the patient itself. When correcting, the adjustment degree is smaller.

[0084] It should be noted that negative correlation mapping is a well-known technical means in the art, and can adopt forms such as inverse proportion form or negative exponential power form, etc., and will not be limited and elaborated here.

[0085] So far, the analysis of the waveform performance on the drift segment is completed, and the acquisition of the correction adjustment degree is carried out.

[0086] The recognition module 104 is used to, in the electrocardiogram signal curve, based on the adjustment degree of each waveform segment in the drift adjustment segment and the baseline position of the waveform segment before the drift adjustment segment, adjust the amplitude in the waveform segment to obtain an adjusted electrocardiogram signal; and perform ventricular fibrillation recognition based on the adjusted electrocardiogram signal.

[0087] Based on the adjustment degree, adjustments are made in the correction of baseline drift, and the characteristic information possessed by the patient itself is retained as much as possible to obtain an adjusted electrocardiogram signal. Preferably, in the embodiment of the present invention, the method for obtaining the adjusted electrocardiogram signal includes:

[0088] First, for any waveform segment in the drift adjustment segment, a reference baseline position is determined in the non-drift adjustment segment before this waveform segment, that is, the baseline position for reference during drift adjustment. In the embodiment of the present invention, the method for obtaining the reference baseline includes:

[0089] Among all the non-drift adjustment segments before this waveform segment, the waveform segment with the smallest time distance is used as the reference waveform segment of this waveform segment. Through the waveform segment with the closest distance, that is, the smallest time interval, in the stable wave band in the previous time sequence, as the waveform segment that can be used for reference and correction, and the baseline position of the reference waveform segment is used as the reference baseline position of this waveform segment.

[0090] Further, the difference between the reference baseline position and the baseline position of this waveform segment is used as the correction degree of this waveform segment, reflecting the magnitude of the correction degree required for drift. And the product of the correction degree and the adjustment degree of this waveform segment is used as the adjustment value of this waveform segment. The smaller the adjustment degree, the higher the requirement to retain the result, and the lower the correction and adjustment requirement. Therefore, the adjustment value will be lower.

[0091] Further, each amplitude on this waveform segment is added to the adjustment value to obtain the adjusted waveform segment of this waveform segment, completing the correction of baseline drift. Finally, the waveform segments of the non-drift adjustment segment and the adjusted waveform segment are arranged in time sequence to form an adjusted electrocardiogram signal. The adjusted electrocardiogram signal completes the adjustment of the baseline drift problem, retains as much information as possible, and ventricular fibrillation recognition can be performed based on the adjusted electrocardiogram signal.

[0092] In the embodiment of the present invention, the adjusted electrocardiogram signal is input into a trained ventricular fibrillation signal recognition model, and the recognition result is output. When the output result is that there is a ventricular fibrillation signal, the system can immediately call a doctor and a nurse and issue an alarm.

[0093] In summary, when the present invention obtains waveform segments, it improves the accuracy of baseline changes between subsequent different waveforms through superposition period analysis. By the different continuous contribution degrees of different trend segments on the waveform segment, the determination of the baseline position is adjusted to improve the accuracy of the baseline position, providing a more accurate data basis for the determination of subsequent drift conditions. After determining the drift segment according to the continuous baseline position changes, considering the possible heartbeat fluctuations of the patient himself that need to be retained, the adjustment degree is determined through the amplitude fluctuation and the deviation in the duration between the waveform segment and the subsequent waveform segment, so that the subsequent baseline drift adjustment can retain more floating information of critically ill patients and improve the signal recognition accuracy. Finally, based on the adjustment degree and the baseline position situation before the drift segment, the waveform segment is adjusted to obtain an adjusted electrocardiogram signal for more accurate ventricular fibrillation recognition. The present invention adjusts the determination of the baseline position by the contributions of different trend changes in the superimposed waveform, and adjusts the drift correction degree by the deviation between waveform segments in the drift adjustment segment, improving the accuracy of electrocardiogram signal drift adjustment and retaining signal characteristics to make subsequent ventricular fibrillation recognition more accurate and efficient.

[0094] It should be noted that the above sequence of embodiments of the present invention is only for description and does not represent the advantages or disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments.

Claims

1. A ventricular fibrillation signal recognition system for critical care ECG monitoring, characterized in that: The system comprises: The data acquisition module is used to obtain the patient's ECG signal curve; determine the waveform segment according to the change trend of the amplitude in the ECG signal curve; there is a same trend segment between every two adjacent waveform segments; A baseline analysis module is used to obtain the baseline position contribution of each trend segment according to the time and waveform duration of each trend segment in each waveform segment; based on the baseline position contribution of the trend segment in each waveform segment, the baseline position of each waveform segment is determined in combination with the amplitude change of the trend segment; The adjustment analysis module is used to determine the drift adjustment segment in the ECG signal curve according to the continuous change of the baseline position of the waveform segment in the time sequence; and obtain the adjustment degree of each waveform segment according to the degree of deviation between each waveform segment in the drift adjustment segment and the subsequent waveform segment in amplitude fluctuation and duration; The recognition module is used to adjust the amplitude of the waveform segment in the ECG signal curve based on the adjustment degree of each waveform segment in the drift adjustment segment and the baseline position of the waveform segment before the drift adjustment segment to obtain an adjusted ECG signal; and perform ventricular fibrillation recognition based on the adjusted ECG signal; The method for determining the baseline position includes: For any waveform segment, the proportion of the baseline position contribution of each trend segment in the waveform segment to all baseline position contributions is taken as the contribution of each trend segment; The average of the maximum amplitude and the minimum amplitude of each trend segment in the waveform segment is taken as the central amplitude of each trend segment; The central amplitude is weighted based on the contribution of each trend segment in the waveform segment, and the sum is calculated to obtain the baseline position of the waveform segment; The method for determining the drift adjustment section includes: The baseline positions of the waveform segments are arranged in time sequence to obtain a baseline sequence; the difference between each baseline position and the next baseline position in the baseline sequence is used as the position deviation of each baseline position; the position deviations of the baseline positions are accumulated in the order of the baseline sequence to obtain a cumulative offset; When the accumulated deviation is greater than a preset deviation threshold, the accumulation is stopped, and the waveform segments of all baseline positions except the last baseline position involved in the accumulation are taken as stable waveform segments; According to the deviation between the baseline position of the baseline sequence after the stable band and the baseline position of the stable band, a new accumulation position is determined; in the baseline sequence, iterative accumulation is continued from the new accumulation position backward until all stable waveform segments are determined; The time periods corresponding to all unstable waveform segments are used as drift adjustment segments.

2. A ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The method for obtaining the baseline position contribution includes: For any trend segment, the length of the ECG signal curve on the trend segment is used as the waveform continuity index of the trend segment; The contribution of the baseline position of the trend segment is obtained by combining the time period length and waveform continuity index of the trend segment.

3. A ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The step of determining a new accumulation position according to the deviation between the baseline position of the baseline sequence after the stable band and the baseline position of the stable band comprises: After the stable waveform segment of the baseline sequence, the difference between each baseline position and the first baseline position in the baseline sequence is calculated to obtain the offset judgment index; When the deviation judgment index is less than or equal to the preset deviation threshold, the corresponding baseline position is used as the new accumulated position.

4. The ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The method for obtaining the adjustment degree includes: The amplitude extreme difference in each waveform segment is used as the fluctuation index of each waveform segment; for any waveform segment in the drift adjustment segment, the difference in waveform index between the waveform segment and the previous waveform segment in time sequence is used as the fluctuation deviation degree of the waveform segment; The difference in time period length between the waveform segment and the previous waveform segment in time sequence is taken as the time deviation degree of the waveform segment; The adjustment degree of the waveform segment is obtained by combining the fluctuation deviation degree and the time deviation degree of the waveform segment.

5. The ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The method for obtaining the adjusted electrocardiogram signal comprises: For any waveform segment in the drift adjustment segment, a reference baseline position is determined in a non-drift adjustment segment preceding the waveform segment; The difference between the reference baseline position and the baseline position of the waveform segment is used as the correction degree of the waveform segment; the product of the correction degree and the adjustment degree of the waveform segment is used as the adjustment value of the waveform segment; each amplitude on the waveform segment is added to the adjustment value to obtain the adjusted waveform segment of the waveform segment; The waveform segment of the non-drift adjustment segment and the adjustment waveform segment are arranged in time sequence to form an adjustment electrocardiogram signal.

6. A ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 5, characterized in that: The method for obtaining the reference baseline position includes: Among all non-drift adjustment segments before the waveform segment, the waveform segment with the smallest time distance is used as the reference waveform segment of the waveform segment; and the baseline position of the reference waveform segment is used as the reference baseline position of the waveform segment.

7. The ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The method for obtaining the waveform segment includes: Obtain all extreme value points in the ECG signal curve; regard the period between two adjacent extreme value points as a trend segment; and regard every two adjacent trend segments as a waveform segment.

8. The ventricular fibrillation signal recognition system for critical care ECG monitoring according to claim 1, characterized in that: The method of identifying ventricular fibrillation based on adjusting the electrocardiogram signal includes: The adjusted ECG signal is input into the trained ventricular fibrillation signal recognition model and the recognition result is output.

Images