An efficient management system for patients' electrocardiogram data

Abstract

The present invention relates to the technical field of data compression, and in particular to an efficient management system for electrocardiogram data of a patient; vertex confidences are obtained according to data change characteristics of electrocardiogram data; vertex confidences are clustered and wave group vertices are obtained according to data characteristics of data clusters and difference characteristics of vertex confidences; data segmentation points are obtained according to data change characteristics before any wave group vertex and segmented to obtain different electrocardiogram data segments; difference degrees, reference data segments and marked data segments are obtained according to data difference characteristics between any electrocardiogram data segments; the present invention replaces the marked data segments to obtain replacement numbers; grouping and clustering are performed according to the difference degrees of reference data segments to obtain different group compression clusters; group compression clusters are compressed and stored to obtain electrocardiogram compression sets; and storage space utilization of electrocardiogram data is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data compression, and in particular to an efficient management system for patient electrocardiogram data. Background Art

[0002] Electrocardiogram is one of the important information for evaluating heart health. By recording changes in the heart's electrical activity, it can detect a variety of heart diseases. In order to improve the efficiency of ECG data access and storage, the recording, management and storage of ECG data have gradually turned to digitalization. The storage of ECG data mostly relies on local servers or traditional database systems. As the number of patients gradually increases, the storage space occupied by ECG data increases significantly, and the data storage efficiency decreases. At the same time, some storage systems have insufficient space and are unable to meet the increasing demand for ECG data storage, resulting in an increase in the storage cost of ECG data. Summary of the invention

[0003] In order to solve the above technical problems that as the amount of ECG data increases, the space occupied and cost of data storage increases significantly, and the storage efficiency decreases, the purpose of the present invention is to provide an efficient management system for patient ECG data. The technical solutions adopted are as follows:

[0004] A data acquisition module, used for acquiring electrocardiogram data of the patient;

[0005] A data analysis module, used to obtain vertex confidences according to data change characteristics of the neighborhood of data points in the electrocardiogram data; cluster the vertex confidences of the data points to obtain two data clusters; and obtain wave group vertices according to data characteristics of the data clusters and difference characteristics of vertex confidences of adjacent data points;

[0006] A data classification and sorting module is used to obtain data segmentation points according to data change characteristics before any wave group vertex; segment the electrocardiogram data according to all data segmentation points to obtain different electrocardiogram data segments; obtain difference and similar electrocardiogram data segments according to data difference characteristics between any electrocardiogram data segments of all patients; obtain reference data segments and marked data segments according to data length characteristics of similar electrocardiogram data segments; replace the marked data segments according to the reference data segments to obtain replacement numbers; group and cluster the reference data segments according to the difference of the reference data segments to obtain different grouping compression clusters;

[0007] The data compression management module is used to compress and store the group compression cluster to obtain an ECG compression set; and search and view the ECG compression set according to the preset patient label of the patient's baseline data segment and the alternative number of the marked data segment.

[0008] Furthermore, the step of obtaining vertex confidence according to the data change characteristics of the neighborhood of the data point in the electrocardiogram data includes:

[0009] ,

[0010] In the formula, represents the vertex confidence of the i-th data point in the electrocardiogram data, represents the normalization function, represents the value of the i-th data point, It represents the average rate of change of n other data points connected before the i-th data point. It represents the average rate of change of n other data points connected after the i-th data point. Indicates the opposite degree of change trend. Indicates the degree of neighborhood change.

[0011] Furthermore, the step of obtaining the wave group vertex according to the data characteristics of the data cluster and the difference characteristics of vertex confidences of adjacent data points includes:

[0012] The numerical average of the data points in the data cluster is calculated to obtain the overall characterization value; for any data point in the data cluster with the largest overall characterization value, if the vertex confidence of the arbitrary data point is greater than the vertex confidence of other adjacent data points in the electrocardiogram data, the arbitrary data point is the vertex of the wave group.

[0013] Furthermore, the step of obtaining the data segmentation point according to the data change characteristics before the arbitrary wave group vertex includes:

[0014] The ECG data are baseline calibrated and denoised by minimum mean square error to obtain ECG filtered data; the ECG filtered data are differentiated, and the data with a differential result of a constant 0 in the ECG filtered data are used as smoothed data, and the connected smoothed data are used as smoothed data segments; starting from an arbitrary wave group vertex in the ECG filtered data, the moment in the second smoothed data segment that is closest to the arbitrary wave group vertex along the direction of historical moments is used as a data segmentation point.

[0015] Furthermore, the step of obtaining the difference and similar ECG data segments according to the data difference characteristics between any ECG data segments of all patients includes:

[0016] Align any two ECG data segments from the starting position, calculate the average value of all data differences in the same position after the alignment of any two ECG data segments, and obtain the difference between the any two ECG data segments; for any ECG data segment, take other ECG data segments whose difference with the any ECG data segment is a constant 0 as similar ECG data segments of the any ECG data segment.

[0017] Furthermore, the step of obtaining the reference data segment and the marked data segment according to the data length characteristics of the similar ECG data segment includes:

[0018] Among the arbitrary ECG data segments and the corresponding similar ECG data segments, the ECG data segment with the largest amount of data is used as a reference data segment, and the non-reference data segment is used as a marked data segment.

[0019] Furthermore, the step of replacing the marked data segment according to the reference data segment to obtain a replacement number includes:

[0020] The preset patient label corresponding to the reference data segment, the preset patient label corresponding to the marked data segment and the data length value of the marked data segment are concatenated and combined to obtain a replacement number for the marked data segment.

[0021] Furthermore, the step of grouping and clustering the reference data segments according to the differences of the reference data segments to obtain different group compression clusters includes:

[0022] Calculate the average value of the difference between any benchmark data segment and all other benchmark data segments to obtain the comprehensive difference of the arbitrary benchmark data segment; cluster the benchmark data segments whose comprehensive difference does not exceed the preset difference threshold to obtain different group compression clusters, and treat all benchmark data segments whose comprehensive difference exceeds the preset difference threshold as separate group compression clusters.

[0023] Furthermore, the step of compressing and storing the grouped compressed clusters to obtain an ECG compression set includes:

[0024] The reference data segment in the grouped compression cluster is compressed by Huffman coding to obtain an ECG compression set corresponding to the grouped compression cluster.

[0025] The present invention has the following beneficial effects:

[0026] In the present invention, obtaining vertex confidence can obtain the position of each ECG cycle in the ECG data according to the changing characteristics of the ECG cycle; obtaining two data clusters can distinguish data points with obvious differences in vertex confidence, and preliminarily improve the position of the wave group vertex; obtaining the wave group vertex can accurately determine the position of each ECG cycle in the ECG data. Obtaining data segmentation points can determine the segmentation points of the ECG cycle according to the changing law of the ECG cycle, thereby improving the accuracy of the analysis of the difference characteristics of different subsequent ECG data segments; obtaining ECG data segments can facilitate the replacement of ECG data segments with the same data, reducing the storage space occupancy rate of the data; obtaining the difference degree can determine whether there is the same data between different ECG data segments, thereby determining similar ECG data segments. Obtaining the reference data segment and the marked data segment can determine the ECG data segments used for replacement and replaced, and obtaining the replacement number can reduce the occupancy rate of the marked data segment in the storage space, so that the marked data segment can be replaced with only a shorter string. Obtaining the group compression cluster can cluster similar reference data segments, thereby further improving the compression degree of the reference data segment in the compression process and improving the storage space utilization rate. Acquiring an ECG compression set of a reference data segment and marking an alternative number of a data segment can significantly reduce the space occupancy of ECG data during storage and reduce storage costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 A block diagram of an efficient management system for patient electrocardiogram data provided by one embodiment of the present invention;

[0029] Figure 2 This is a model diagram of a complete ECG cycle of the present invention. DETAILED DESCRIPTION

[0030] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of an efficient management system for patient electrocardiogram data proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0031] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0032] The specific scheme of an efficient management system for patient electrocardiogram data provided by the present invention is described in detail below with reference to the accompanying drawings.

[0033] See also Figure 1 , which shows a block diagram of an efficient management system for patient electrocardiogram data provided by an embodiment of the present invention, the system includes the following modules:

[0034] The data acquisition module S1 is used to acquire the electrocardiogram data of the patient.

[0035] In an embodiment of the present invention, the implementation scenario is to compress and manage the ECG data to improve the utilization rate of data storage space and reduce storage costs. First, the ECG data of the patient is obtained. The ECG data of each patient corresponds to a preset patient label. For example, the ECG data corresponding to a certain patient is a. Subsequently, the preset patient label a of the patient can be entered in the storage space to retrieve the ECG data, thereby quickly obtaining the ECG data of the patient; the implementer can determine the preset patient labels for different patients according to the implementation scenario. The ECG data obtained by the medical device is stored in the cache area. When the ECG data in the cache area exceeds the preset length, the ECG data in the cache area is compressed and analyzed. In the embodiment of the present invention, the preset length is 1 hour of ECG data, which can be determined by the implementer according to the implementation scenario.

[0036] The data analysis module S2 is used to obtain vertex confidence according to the data change characteristics of the neighborhood of the data point in the electrocardiogram data; cluster the vertex confidence of the data point to obtain two data clusters; and obtain the wave group vertex according to the data characteristics of the data cluster and the difference characteristics of the vertex confidence of adjacent data points.

[0037] In the electrocardiogram (ECG), a complete cardiac cycle mainly includes: P wave, which indicates the depolarization process of the atrium; PR interval, which is the time from the start of the P wave to the start of the QRS complex, indicating the conduction time from the atrium to the ventricle; QRS complex, which indicates the depolarization process of the ventricle; ST segment, which is the time period from the end of the QRS complex to the start of the T wave; T wave, which indicates the repolarization process of the ventricle. It should be noted that the naming and meaning of the cardiac cycle are well-known contents and will not be described in detail. Figure 2The model diagram of a complete ECG cycle is shown. ECG data is composed of multiple similar ECG cycles, so the ECG data of patients has periodic changes. The ECG cycle data of each cycle segment is similar. In order to initially reduce the storage space occupancy rate of data, the data of some repeated ECG cycles can be replaced with one ECG cycle. Therefore, it is necessary to divide different ECG cycles to determine the repeated ECG cycles.

[0038] Furthermore, the values ​​and changes at the apex of the QRS complex in the electrocardiogram data are most obvious, and determining the location of the apex of the complex can help to divide different electrocardiogram cycles. Because the patient's electrocardiogram data will be affected by the external environment, etc., the values ​​of different electrocardiogram cycles are different. It is possible that the values ​​of non-complex apex in a certain electrocardiogram cycle are the same as the values ​​of complex apex in other electrocardiogram cycles, resulting in low accuracy in determining the apex of the complex based only on the original electrocardiogram data; therefore, in order to improve the accuracy of obtaining the apex of the complex, it is necessary to analyze the data change characteristics at the QRS complex within the electrocardiogram cycle, so the vertex confidence can be obtained based on the data change characteristics of the data point neighborhood in the electrocardiogram data; preferably, in an embodiment of the present invention, the step of obtaining the vertex confidence includes:

[0039]

[0040] In the formula, represents the vertex confidence of the i-th data point in the electrocardiogram data, represents a normalization function, and in the embodiment of the present invention, the maximum and minimum value method is used for normalization; represents the value of the i-th data point. The value of the QRS wave group apex in one ECG cycle is the largest, so The larger it is, the more likely the data point is the vertex of the wave group; It represents the average rate of change of n other data points connected before the i-th data point. It represents the average change rate of n other data points connected after the ith data point. In the embodiment of the present invention, n is 10, that is, the average change rate of the ten data points before and after the data point. The implementer can determine it according to the implementation scenario. The difference in the change rate of the data before and after the apex of the QRS complex is the most obvious in the entire ECG cycle. Indicates the opposite degree of the change trend. The greater the opposite degree of the change trend, the more likely the data point is to be the vertex of the wave group. It represents the degree of neighborhood change. The absolute values ​​of the change rates of the data on both sides of the wave group vertex are large. Therefore, the larger the degree of neighborhood change, the more likely the data point is to be the vertex of the wave group.

[0041] The greater the vertex confidence, the more likely the data point is to be the vertex of the QRS complex in the ECG cycle, while the vertex confidence of other data points that are not the vertex of the wave group is relatively small; furthermore, there will be obvious differences in the vertex confidence of the wave group vertex and the non-wave group vertex in the electrocardiogram data, so the vertex confidence of the data points can be clustered to obtain two data clusters; in the embodiment of the present invention, clustering is performed using the existing K-means++ clustering algorithm. It should be noted that this clustering algorithm is an improved algorithm of the K-means clustering algorithm, and the clustering accuracy is higher. The specific clustering steps will not be repeated. Among the two data clusters obtained by clustering, the vertex confidence of the data points in one of the data clusters is relatively large, and the data points in the data cluster are more likely to be wave group vertices. In order to further screen out accurate wave group vertices, the wave group vertices can be obtained based on the data characteristics of the data cluster and the difference characteristics of the vertex confidence of adjacent data points.

[0042] Preferably, in an embodiment of the present invention, the step of obtaining the wave group vertex includes: calculating the numerical average of the data points in the data cluster to obtain the overall characterization value; the larger the overall characterization value, the more likely the data point in the data cluster is to be a wave group vertex. For any data point in the data cluster with the largest overall characterization value, if the vertex confidence of any data point is greater than the vertex confidence of other adjacent data points in the electrocardiogram data, then the arbitrary data point is a wave group vertex; since the data points adjacent to the wave group vertex in some electrocardiogram cycles will also be in the data cluster with the largest overall characterization value, it is necessary to compare the vertex confidence of each data point in the data cluster with the adjacent data points in the electrocardiogram data to avoid classifying non-wave group vertices as wave group vertices.

[0043] The data classification and sorting module S3 is used to obtain data segmentation points according to the data change characteristics before any wave group vertex; segment the ECG data according to all data segmentation points to obtain different ECG data segments; obtain the difference and similar ECG data segments according to the data difference characteristics between any ECG data segments of all patients; obtain the benchmark data segments and the marked data segments according to the data length characteristics of the similar ECG data segments; replace the marked data segments according to the benchmark data segments to obtain the replacement numbers; group and cluster the benchmark data segments according to the difference of the benchmark data segments to obtain different grouping compression clusters.

[0044] After obtaining all the wave group vertices in the electrocardiogram data, the position of each electrocardiogram cycle can be determined, so the data segmentation point can be obtained according to the data change characteristics before any wave group vertex; preferably, in the embodiment of the present invention, the step of obtaining the data segmentation point includes: the real electrocardiogram data will show subtle sawtooth changes, which is not conducive to determining the segmentation point, so the electrocardiogram data is baseline calibrated and denoised by the minimum mean square error to obtain the electrocardiogram filtered data; it should be noted that the minimum mean square error algorithm belongs to the prior art and can smooth the original data, and the specific steps will not be repeated; the smoothed electrocardiogram data is closer to Figure 2 The model diagram of a complete ECG cycle. The ECG filter data is differentiated, and the data with a differential result of constant 0 in the ECG filter data is used as smoothed data, and the connected smoothed data is used as smoothed data segments; according to Figure 2 It can be seen that there are two smooth data segments before the PR segment and before the P wave. In the ECG filtering data, starting from the arbitrary wave group vertex, along the historical time direction, the moment in the second smooth data segment closest to the arbitrary wave group vertex is used as the data segmentation point. Since a complete ECG cycle starts from the P wave, the second smooth data segment is the data segment before the P wave, and the moment in the second smooth data segment closest to the arbitrary wave group vertex is the moment before the P wave starting point, so this moment is used as the data segmentation point. The interval from each data segmentation point to the next data segmentation point contains a complete ECG cycle. Then, the ECG data is segmented according to all data segmentation points to obtain different ECG data segments, and each ECG data segment contains an ECG cycle.

[0045] Furthermore, traditional data storage saves all ECG data, but there are a large number of identical ECG cycles in the ECG data, and the same data will be stored repeatedly during storage, resulting in a waste of space; therefore, in order to improve space utilization, the same ECG data segments can be replaced. Since the same ECG data segments may exist in different patients, the difference and similar ECG data segments are obtained according to the data difference characteristics between any ECG data segments of all patients; preferably, in the embodiment of the present invention, the step of obtaining the difference and similar ECG data segments includes: aligning any two ECG data segments from the starting position, calculating the average value of the data difference of all the same positions after the alignment of any two ECG data segments, and obtaining the difference of any two ECG data segments; since the data length of the ECG data segments may be different, but the starting position of the ECG data segments is at the beginning of the P wave, any two ECG data segments are aligned. If the value difference of the two ECG data segments at all the same positions after the alignment is 0, the difference is 0, which means that the ECG values ​​of the two ECG data segments at the same position are exactly the same; after the alignment of the two ECG data segments, the redundant part of a certain ECG data segment does not participate in the calculation of the difference. For any ECG data segment, other ECG data segments with a constant difference of 0 from the arbitrary ECG data segment are used as similar ECG data segments of the arbitrary ECG data segment.

[0046] For any ECG data segment and the corresponding similar ECG data segment, there will be a segment of completely identical changed data; therefore, the baseline data segment and the mark data segment can be obtained according to the data length characteristics of the similar ECG data segment; preferably, in an embodiment of the present invention, the step of obtaining the baseline data segment and the mark data segment includes: in any ECG data segment and the corresponding similar ECG data segment, the ECG data segment with the largest amount of data is used as the baseline data segment, and the non-baseline data segment is used as the mark data segment; and then the mark data segment can be represented by part of the data segment in the baseline data segment, and only the baseline data segment can be compressed during compression storage, and the mark data segment corresponding to the baseline data segment can be replaced by a shorter number, thereby ultimately improving the utilization rate of storage space.

[0047] Furthermore, since part of the data in the marked data segment and the reference data segment are the same, the marked data segment can be replaced according to the reference data segment to obtain a replacement number; preferably, in an embodiment of the present invention, the step of obtaining the replacement number includes: splicing and combining the preset patient label corresponding to the reference data segment, the preset patient label corresponding to the marked data segment and the data length value of the marked data segment to obtain the replacement number of the marked data segment; for example, the numerical sequence of the reference data segment is 2354173, the corresponding sequence of the marked data segment is 2354, the data length of the marked data segment is 4, which is the same as the first 4 bits of the reference data segment, so the data length value of the marked data segment is 4; if the preset patient label of the patient corresponding to the reference data segment is a, and the reference data segment is the third segment after the electrocardiogram data of the patient is segmented, then the preset patient label of the reference data segment is a3; the preset patient label corresponding to the marked data segment is b, and the marked data segment is the eighth segment after the electrocardiogram data of the patient is segmented, and similarly the preset patient label of the marked data segment is b8. After the splicing is completed, the replacement number of the marked data segment is , where the symbol Indicates concatenation, the substitution code of the marked data segment means that the 8th ECG data segment in the ECG data with the preset patient label b is the same as the first 4 data of the 3rd ECG data segment in the ECG data with the preset patient label a. Similarly, other marked data segments corresponding to the reference data segment will obtain the same form of substitution code. When storing data, only the reference data segment occupies more space, while the marked data segment only needs a short substitution code to complete the storage, thereby greatly improving space utilization.

[0048] After processing the marked data segments, the benchmark data segments need to be compressed and stored to further improve space utilization and reduce storage costs; before compression, the benchmark data segments can be clustered to cluster similar benchmark data segments into a cluster. Compressing the benchmark data segments within the cluster can further reduce redundancy and improve compression compared to compressing all benchmark data segments together. However, when clustering the benchmark data segments, due to the abnormal ECG data of some patients, the benchmark data segments will be significantly different from most other benchmark data segments. If all benchmark data segments are clustered, such obviously abnormal benchmark data segments will form isolated points in the clustering process, affecting the clustering accuracy, and then affecting the compression efficiency of the data within the cluster; therefore, the benchmark data segments need to be screened before clustering to improve clustering accuracy and compression. Therefore, the benchmark data segments are grouped and clustered according to the difference of the benchmark data segments to obtain different group compression clusters.

[0049] Preferably, in an embodiment of the present invention, the step of obtaining a group compression cluster includes: calculating the average value of the difference between any benchmark data segment and all other benchmark data segments to obtain the comprehensive difference of any benchmark data segment; when the comprehensive difference is larger, it means that the difference between the ECG data of the arbitrary benchmark data segment and other benchmark data segments is larger, and the arbitrary benchmark data segment is rarer, and the arbitrary benchmark data segment can be distinguished before clustering. The benchmark data segments whose comprehensive difference does not exceed the preset difference threshold are clustered to obtain different group compression clusters; in an embodiment of the present invention, the preset difference threshold is 0.3, which can be determined by the implementer according to the implementation scenario. The comprehensive difference does not exceed the preset difference threshold, which means that the benchmark data segment is relatively similar to other benchmark data segments and can participate in clustering; clustering uses the K-means++ clustering algorithm in module S2. All benchmark data segments whose comprehensive difference exceeds the preset difference threshold are regarded as separate group compression clusters. The benchmark data segments in the separate group compression cluster are relatively rare. In order not to affect the overall clustering accuracy, they are separated into a cluster.

[0050] The data compression management module S4 is used to compress and store the grouped compression clusters to obtain an ECG compression set; and search and view the ECG compression set according to the preset patient label of the patient's baseline data segment and the alternative number of the marked data segment.

[0051] After obtaining different group compression clusters of reference data segments, the group compression clusters can be compressed and stored to obtain an ECG compression set; in an embodiment of the present invention, the reference data segments in the group compression cluster are compressed by Huffman coding to obtain an ECG compression set corresponding to the group compression cluster. It should be noted that Huffman coding belongs to the prior art, and the specific compression steps are not repeated here. When the coding algorithm compresses data with greater similarity, the compression degree is higher, which can reduce the space occupied when the data is stored; therefore, the reference data segments in different group compression clusters are compressed separately, which can further improve the compression degree and reduce the storage space occupancy rate compared to compressing all reference data segments at the same time. After the compression is completed, the ECG compression set and the replacement number of the marked data segment are stored. When checking the ECG data later, the preset patient label of the patient's baseline data segment and the replacement number of the marked data segment can be used to search and check in the ECG compression set; for example, if the preset patient label corresponding to the patient is a, the baseline data segment contained in the ECG data corresponding to a is determined in all ECG compression sets; at the same time, according to the replacement number of the marked data segment corresponding to a, the baseline data segment to be replaced is searched in the ECG compression set for decoding to obtain the original marked data segment; finally, all the baseline data segments and marked data segments corresponding to a are spliced ​​in the order in the ECG data to obtain the ECG data of a. So far, by replacing and grouping ECG data segments containing the same data, the space utilization rate of ECG data during storage is greatly improved, and the storage cost is reduced.

[0052] In summary, the embodiments of the present invention provide an efficient management system for patient electrocardiogram data; vertex confidence is obtained according to the data change characteristics of the electrocardiogram data; the vertex confidence is clustered and the wave group vertices are obtained according to the data characteristics of the data cluster and the difference characteristics of the vertex confidence; data segmentation points are obtained according to the data change characteristics before any wave group vertex and segmented to obtain different electrocardiogram data segments; the difference degree, the benchmark data segment and the marked data segment are obtained according to the data difference characteristics between any electrocardiogram data segments; the present invention replaces the marked data segment to obtain a replacement number; grouping and clustering are performed according to the difference degree of the benchmark data segment to obtain different group compression clusters; the group compression clusters are compressed and stored to obtain an electrocardiogram compression set; and the storage space utilization rate of the electrocardiogram data is improved.

[0053] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying 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.

[0054] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An efficient management system for patient electrocardiogram data, characterized in that: The system includes the following modules: A data acquisition module, used for acquiring electrocardiogram data of the patient; A data analysis module, used for obtaining vertex confidence according to data change characteristics of a neighborhood of a data point in the electrocardiogram data; Clustering the vertex confidences of the data points to obtain two data clusters; obtaining wave group vertices according to data features of the data clusters and difference features of vertex confidences of adjacent data points; A data classification and sorting module is used to obtain data segmentation points according to data change characteristics before any wave group vertex; segment the electrocardiogram data according to all data segmentation points to obtain different electrocardiogram data segments; and obtain difference and similar electrocardiogram data segments according to data difference characteristics between any electrocardiogram data segments of all patients; Obtain a reference data segment and a marked data segment according to the data length characteristics of the similar ECG data segment; replace the marked data segment according to the reference data segment to obtain a replacement number; group and cluster the reference data segments according to the difference of the reference data segments to obtain different grouping compression clusters; A data compression management module is used to compress and store the grouped compression clusters to obtain an ECG compression set; and to search and view the ECG compression set according to a preset patient label of the patient's baseline data segment and an alternative number of the marked data segment; The step of obtaining the difference and similar ECG data segments according to the data difference characteristics between any ECG data segments of all patients comprises: Align any two ECG data segments from the starting position, calculate the average value of the data difference of all the same positions after the alignment of any two ECG data segments, and obtain the difference between the two ECG data segments; if the value difference of the two ECG data segments after the alignment is 0 at all the same positions, the difference is 0; the redundant part of an ECG data segment after the alignment of two ECG data segments does not participate in the calculation of the difference; for any ECG data segment, other ECG data segments with a constant difference of 0 from the arbitrary ECG data segment are regarded as similar ECG data segments to the arbitrary ECG data segment; The step of obtaining a reference data segment and a marked data segment according to the data length characteristics of the similar ECG data segment comprises: Among the arbitrary ECG data segments and the corresponding similar ECG data segments, the ECG data segment with the largest amount of data is used as a reference data segment, and the non-reference data segment is used as a marked data segment; The step of replacing the marked data segment according to the reference data segment to obtain a replacement number comprises: Concatenate and combine the preset patient label corresponding to the reference data segment, the preset patient label corresponding to the marked data segment, and the data length value of the marked data segment to obtain a replacement number for the marked data segment; The step of grouping and clustering the reference data segments according to the differences of the reference data segments to obtain different group compression clusters comprises: Calculate the average value of the difference between any benchmark data segment and all other benchmark data segments to obtain the comprehensive difference of the arbitrary benchmark data segment; cluster the benchmark data segments whose comprehensive difference does not exceed the preset difference threshold to obtain different group compression clusters, and treat all benchmark data segments whose comprehensive difference exceeds the preset difference threshold as separate group compression clusters.

2. The efficient management system for patient electrocardiogram data according to claim 1, characterized in that: The step of obtaining vertex confidence according to the data change characteristics of the data point neighborhood in the electrocardiogram data comprises: ， In the formula, represents the vertex confidence of the i-th data point in the electrocardiogram data, represents the normalization function, represents the value of the i-th data point, It represents the average rate of change of n other data points connected before the i-th data point. It represents the average rate of change of n other data points connected after the i-th data point. Indicates the opposite degree of change trend. Indicates the degree of neighborhood change.

3. The efficient management system for patient electrocardiogram data according to claim 1, characterized in that: The step of obtaining the wave group vertex according to the data characteristics of the data cluster and the difference characteristics of vertex confidences of adjacent data points comprises: The numerical average of the data points in the data cluster is calculated to obtain the overall characterization value; for any data point in the data cluster with the largest overall characterization value, if the vertex confidence of the arbitrary data point is greater than the vertex confidence of other adjacent data points in the electrocardiogram data, the arbitrary data point is the vertex of the wave group.

4. The efficient management system for patient electrocardiogram data according to claim 1, characterized in that: The step of obtaining the data segmentation point according to the data change characteristics before the arbitrary wave group vertex includes: The ECG data are baseline calibrated and denoised by minimum mean square error to obtain ECG filtered data; the ECG filtered data are differentiated, and the data with a differential result of a constant 0 in the ECG filtered data are used as smoothed data, and the connected smoothed data are used as smoothed data segments; starting from an arbitrary wave group vertex in the ECG filtered data, the moment in the second smoothed data segment that is closest to the arbitrary wave group vertex along the direction of historical moments is used as a data segmentation point.

5. The efficient management system for patient electrocardiogram data according to claim 1, characterized in that: The step of compressing and storing the grouped compressed clusters to obtain an ECG compression set comprises: The reference data segment in the grouped compression cluster is compressed by Huffman coding to obtain an ECG compression set corresponding to the grouped compression cluster.

Images