Electrocardiogram report digital processing method and device, equipment, medium and product

Abstract

The invention relates to an electrocardiogram report digital processing method which comprises the following steps: preprocessing a paper electrocardiogram report to obtain report basic information and a lead waveform image to be processed; performing connected region detection on the lead waveform to-be-processed image, and dividing the lead waveform to-be-processed image into a plurality of to-be-processed image regions; when the adjacent lead waveform images in the to-be-processed image area have overlapped waveforms, determining accurate positions of wave crests in the overlapped waveforms; based on the position of each to-be-processed image area in the lead waveform to-be-processed image, determining a lead category corresponding to the lead waveform image in each to-be-processed image area; extracting the electrocardiogram data of the corresponding lead category from the lead waveform image in each to-be-processed image area; and generating a digital report corresponding to the paper electrocardiogram report. The invention further relates to a device applying the method, and computer equipment, a computer readable storage medium and a computer program product which are used for implementing the method.

Description

Technical Field

[0001] This application belongs to the field of image processing technology, and particularly relates to a method for digitizing electrocardiogram (ECG) reports and an apparatus for digitizing ECG reports using the method, as well as computer equipment, computer-readable storage media, and computer program products for implementing the various steps of the method. Background Technology

[0002] Electrocardiogram (ECG) is an important non-invasive tool for diagnosing heart and cardiovascular diseases. Traditional paper-based ECG reports face increasing limitations in storing, retrieving, and sharing patient data, particularly in the establishment of electronic medical records and integration into IoT-based smart healthcare systems. For example, paper-based ECG reports are stored as scanned images, thus lacking the numerical data required for these applications, severely limiting the ability of these systems to utilize the vast amounts of valuable historical ECG data. Existing ECG digitization methods lack effective and accurate data extraction techniques when there is overlap between ECG waveforms, resulting in a high discard rate when processing overlapping ECG leads; or they introduce significant computational overhead, making them difficult to deploy in large-scale or resource-constrained environments; or they lack full automation capabilities, with some methods requiring manual definition of lead areas, lacking the ability to digitize ECG waveforms in batches, and some algorithms are incompatible with multiple ECG report types.

[0003] Therefore, a fully automated solution for digital processing of electrocardiogram (ECG) reports is desired. Summary of the Invention

[0004] This section presents the concepts of this application in a simplified form, which will be further illustrated in the detailed description below. This section is not intended to identify key or essential features of the claimed subject matter, nor is it intended to assist in determining the scope of the claimed subject matter.

[0005] According to a first aspect of this application, a method for digitizing an electrocardiogram (ECG) report is provided, comprising: preprocessing a paper ECG report to obtain basic report information and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed; performing connected component detection on the lead waveform images to be processed to divide the lead waveform images to be processed into multiple image regions to be processed; when adjacent lead waveform images in the image regions to be processed have overlapping waveforms, moving a portion of the waveform including the overlapping peak relative to the overlapping waveform, so as to determine the accurate position of the peak in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform; determining the lead category corresponding to the lead waveform images in each image region to be processed based on the position of each image region to be processed in the lead waveform images to be processed; extracting ECG data corresponding to the lead category from the lead waveform images in each image region to be processed based on a reference pulse waveform in the lead waveform images to be processed; and generating a digital report corresponding to the paper ECG report based on the basic report information and the ECG data.

[0006] According to some exemplary embodiments, the preprocessing of the paper electrocardiogram report to obtain basic report information and lead waveform images to be processed includes: scanning and recognizing the paper electrocardiogram report to obtain the basic report information and original lead waveform images; converting the original lead waveform images from the RGB color space to the HSV color space; adjusting the hue value, saturation, and brightness of red in the original lead waveform images to determine the position of the grid lines; performing binarization processing on the original lead waveform images to generate a binarized image of the original lead waveform images; using mask information to set the value of the pixel at the position of the grid lines to zero in the binarized image, thereby removing the grid lines and generating the lead waveform images to be processed.

[0007] According to some exemplary embodiments, the step of performing connected component detection on the lead waveform image to be processed and dividing the lead waveform image to be processed into multiple image regions to be processed includes: dividing the lead waveform images that are continuous in the lead waveform image to be processed into separate image regions to be processed; and dividing adjacent lead waveform images that have overlapping waveforms in the lead waveform image to be processed into the same image region to be processed.

[0008] According to some exemplary embodiments, when adjacent lead waveforms in the image region to be processed have overlapping waveforms, moving a portion of the waveform including the overlapping peak relative to the overlapping waveform to determine the accurate position of the peak in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform includes: determining an overlapping lead connected region in the image region to be processed, wherein the overlapping lead connected region surrounds at least one pair of overlapping troughs and peaks; determining a non-overlapping region in the overlapping lead connected region, wherein the non-overlapping region is formed by being surrounded by the boundary of the lead waveform including the peak and the overlapping lead connected region and includes the preliminary position of the peak; shifting the non-overlapping region along the direction indicated by the peak to form an offset non-overlapping region; weighting and summing the Manhattan distance, cross-union ratio, Hamming distance, and correlation coefficient between each offset non-overlapping region and the overlapping lead connected region to obtain the composite similarity between each offset non-overlapping region and the overlapping lead connected region; and determining the accurate position of the peak in the overlapping waveform by adding the minimum value of the composite similarity to the preliminary position of the peak.

[0009] According to some exemplary embodiments, determining the lead category corresponding to the lead waveform image in each image region to be processed based on the position of each image region to be processed in the lead waveform image to be processed includes: summing the pixel values ​​of each row of pixels in the lead waveform image to be processed to generate the pixel intensity of each row of pixels; determining the row corresponding to the peak pixel intensity among all pixel intensities as the row index in the lead waveform image to be processed; calculating the centroid of the pixel distribution in each image region to be processed; determining the column corresponding to each centroid as the column index in the lead waveform image to be processed; and determining the lead category corresponding to the lead waveform image in each image region to be processed based on the row index and the column index, as well as prior knowledge about the distribution position of different categories of lead waveforms in the paper electrocardiogram report.

[0010] According to some exemplary embodiments, the step of extracting ECG data of the corresponding lead category from the lead waveform images in each image region to be processed based on the reference pulse waveform in the lead waveform image to be processed includes: determining the ratio between pixels and voltage values ​​in the lead waveform image to be processed based on the reference pulse waveform; and converting the lead waveform images in each image region to be processed into corresponding voltage value data based on the determined ratio between pixels and voltage values.

[0011] According to a second aspect of this disclosure, an electrocardiogram (ECG) report digitization processing apparatus is provided, which applies the ECG report digitization processing method according to the first aspect of this disclosure and its exemplary embodiments. The apparatus includes: an image preprocessing module, a to-be-processed image region generation module, an overlapping waveform processing module, a lead category determination module, a data extraction module, and a digitized report generation module. The image preprocessing module is configured to preprocess a paper ECG report to obtain basic report information and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed. The to-be-processed image region generation module is configured to perform connected component detection on the lead waveform images to be processed, dividing the lead waveform images to be processed into multiple to-be-processed image regions. The overlapping waveform processing module is configured to: when adjacent lead waveform images in the to-be-processed image regions have overlapping waveforms, move a portion of the waveform including the overlapping peak relative to the overlapping waveform, so as to determine the accurate position of the peak in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform. The lead category determination module is configured to determine the lead category corresponding to the lead waveform image in each image region to be processed, based on the position of each image region to be processed within the lead waveform image to be processed. The data extraction module is configured to extract the corresponding lead category ECG data from the lead waveform images in each image region to be processed, based on a reference pulse waveform in the lead waveform image to be processed. The digital report generation module is configured to generate a digital report corresponding to the paper ECG report based on the basic report information and the ECG data.

[0012] According to a third aspect of this disclosure, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program, implements the steps of the electrocardiogram report digitization processing method according to the first aspect of this disclosure and its exemplary embodiments.

[0013] According to a fourth aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the electrocardiogram report digitization processing method according to the first aspect of this disclosure and its exemplary embodiments.

[0014] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the steps of the electrocardiogram report digitization processing method according to a first aspect of this disclosure and its exemplary embodiments. Attached Figure Description

[0015] Other or additional features, advantages, and details are presented by way of example only in the following detailed description of the embodiments. In the accompanying drawings:

[0016] Figure 1 A flowchart schematically illustrates an exemplary embodiment of an electrocardiogram (ECG) report digitization processing method according to this disclosure;

[0017] Figure 2 According to exemplary embodiments of this disclosure, further examples are shown Figure 1 Details of the electrocardiogram report digitization process shown;

[0018] Figure 3 The original image of the identified lead waveform is shown schematically;

[0019] Figure 4 According to exemplary embodiments of this disclosure, further examples are shown Figure 1 Details of the electrocardiogram report digitization process shown;

[0020] Figure 5 The diagram schematically illustrates multiple image regions to be processed formed by dividing a local region of the image to be processed in the lead waveform;

[0021] Figure 6 According to exemplary embodiments of this disclosure, further examples are shown Figure 1 Details of the electrocardiogram report digitization process shown;

[0022] Figure 7 An overlapping lead connectivity region is schematically shown in the area of ​​the image to be processed;

[0023] Figure 8 The diagram schematically illustrates the overlapping peaks and troughs in an overlapping waveform;

[0024] Figure 9 The process of determining the precise location of wave crests in overlapping waveforms based on composite similarity is illustrated schematically.

[0025] Figure 10 The iterative calculation of similarity weight coefficients and the relationship between composite similarity are illustrated schematically.

[0026] Figure 11 According to exemplary embodiments of this disclosure, further examples are shown Figure 1 Details of the electrocardiogram report digitization process shown;

[0027] Figure 12 This schematically illustrates the process of determining row and column indices in a lead waveform image to determine the lead category.

[0028] Figure 13 According to exemplary embodiments of this disclosure, further examples are shown Figure 1 Details of the electrocardiogram report digitization process shown;

[0029] Figure 14 This schematically illustrates the process of extracting data from a lead waveform image based on a reference pulse waveform;

[0030] Figure 15 The diagram illustrates a comparison between the waveform plotted from the digitized result obtained using the ECG report digitization method of this disclosure and the waveform in the original paper ECG report.

[0031] Figure 16 The diagram illustrates a comparison between the waveform plotted from the digitized result obtained using the ECG report digitization method of this disclosure and the waveform in the original paper ECG report.

[0032] Figure 17 An electrocardiogram report digitization processing apparatus according to an exemplary embodiment of the present disclosure is schematically illustrated in block diagram form;

[0033] Figure 18 The structure of a computer device according to an exemplary embodiment of the present disclosure is illustrated schematically.

[0034] It should be understood that throughout the accompanying drawings, corresponding reference numerals identify similar or corresponding parts or features. It should also be understood that the drawings are not necessarily drawn to scale and are merely schematic illustrations of exemplary embodiments of the present application, and are not intended to limit the scope of the application. Detailed Implementation

[0035] Exemplary embodiments of this application will now be described with reference to the accompanying drawings. It should be understood that the following description is merely exemplary in nature and is not intended to limit the application, its use, or its application.

[0036] See Figure 1 The flowchart schematically illustrates an exemplary embodiment of an electrocardiogram (ECG) report digitization processing method according to this disclosure. Figure 1 As shown, the electrocardiogram report digitization processing method 100 may include steps 110, 120, 130, 140, 150, and 160:

[0037] In step 110, the paper electrocardiogram report is preprocessed to obtain basic information of the report and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed.

[0038] In step 120, connected component detection is performed on the lead waveform image to be processed, and the lead waveform image to be processed is divided into multiple image regions to be processed.

[0039] In step 130, when adjacent lead waveform images in the image region to be processed have overlapping waveforms, a portion of the waveform including the overlapping peaks is moved relative to the overlapping waveform so as to determine the accurate position of the peaks in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform.

[0040] In step 140, based on the position of each image region to be processed in the image of the lead waveform to be processed, the lead category corresponding to the lead waveform image in each image region to be processed is determined;

[0041] In step 150, based on the reference pulse waveform in the image to be processed of the lead waveform, the electrocardiogram data of the corresponding lead category is extracted from the lead waveform image in each image region to be processed;

[0042] In step 160, based on the basic information of the report and the electrocardiogram data, a digital report corresponding to the paper electrocardiogram report is generated.

[0043] Therefore, the ECG report digitization method 100 of this disclosure can automatically extract multi-lead ECG data from routine ECG reports in batches, avoiding the need for manual selection of the corresponding region for each lead in subsequent reports. When the ECG report has an unconventional layout, the ECG report digitization method 100 only needs to adjust the preset layout to achieve multi-lead ECG digitization. In addition, the ECG report digitization method 100 can automatically and effectively handle overlapping areas between lead waveforms, further improving the extraction efficiency and utilization rate of waveform data, and overcoming the problems of insufficient overlapping area extraction capability and high computational cost.

[0044] See Figure 2 According to an exemplary embodiment of this disclosure, it further illustrates Figure 1 Details of step 110 in the illustrated ECG report digitization method 100. (See attached image.) Figure 2 As shown, step 110 of the ECG report digitization processing method 100 may include steps 111, 112, 113, 114 and 115.

[0045] In step 111, the paper electrocardiogram report is scanned and identified to obtain the basic information of the report and the original images of the lead waveforms.

[0046] A standard paper ECG report typically includes basic information such as the patient's demographics (name, age, gender, medical record number), the date and time of recording, and the physician's identification information. In addition, a standard paper ECG report may include waveforms from the standard 12 leads (I, II, III, aVR, aVL, aVF, V1-V6) and reference pulse waveforms. These waveforms can be scanned to generate raw images of the leads. See also... Figure 3 It schematically illustrates the original image of the identified lead waveform. For example... Figure 3 As shown, the upper half of the original lead waveform image 200 displays the waveforms of the standard 12 leads I, II, III, aVR, aVL, aVF, V1-V6, and the reference pulse waveform RF. The lower half of the original lead waveform image 200 displays the waveforms of three leads selected from the 12 leads, V1, II, and V5, over a longer time span. The waveforms of adjacent leads are generally spaced apart, but in some adjacent leads, the peaks and troughs may overlap, such as... Figure 3 The waveforms of leads V4, V5, and V6 are shown in the diagram. Furthermore, standard paper ECG reports typically have grid lines to aid in reading the voltage values ​​of the ECG lead signals; therefore, the original image of the scanned lead waveforms also contains a grid-line background.

[0047] See also Figure 2 See also Figure 3 In step 112, the original image of the lead waveform is converted from the RGB color space to the HSV color space; in step 113, the position of the grid lines in the original image of the lead waveform is determined by adjusting the hue, saturation and brightness of the original image of the lead waveform.

[0048] like Figure 3 As shown, the waveforms of each lead in the original lead waveform image 200 differ from the grid lines serving as the background in both hue and brightness. Therefore, with appropriate image processing, the waveforms of each lead can be distinguished from the grid lines in the original lead waveform image 200. As an example, the color space of the original lead waveform image 200 is converted from RGB to HSV. This allows the position of individual grid lines to be determined by adjusting only the hue, saturation, and brightness of the original lead waveform image 200 to match the parameters of the grid lines in the original image. It should be understood that the parameters of the HSV color space used can be customized as needed, thus ensuring compatibility with more ECG reports in different formats (here, grid lines of different colors / hues).

[0049] See also Figure 2In step 114, the original lead waveform image is binarized to generate a binarized image of the original lead waveform image. In this step, the entire original lead waveform image 200 (including the waveforms and grid lines of each lead) is binarized to generate a binarized image. In step 115, the values ​​of the pixels at the positions of the grid lines in the binarized image are set to zero, thereby removing the grid lines and generating the lead waveform image to be processed. That is, by using the grid lines separated in the original lead waveform image 200 as a mask and setting the values ​​of the pixels at the positions of the grid lines to zero in the binarized image, the grid lines can be removed from the binarized image (i.e., noise reduction), and this binarized image with the grid lines removed can be used as the lead waveform image to be processed.

[0050] See Figure 4 According to an exemplary embodiment of this disclosure, it further illustrates Figure 1 Details of step 120 in the illustrated ECG report digitization method 100. (See attached image.) Figure 4 As shown, step 120 of the ECG report digitization processing method 100 may include steps 121 and 122.

[0051] In step 121, the continuous lead waveform images in the image to be processed are divided into separate regions of the image to be processed. (As mentioned above...) Figure 3 As described, waveforms in adjacent leads are generally spaced apart. Therefore, as a non-limiting example, a connected component detection method can be used to filter out irrelevant details by analyzing the proportion of connected regions of waveforms in the binarized image within the entire image, and then delineating the waveforms in the image to be processed using bounding boxes. Between some adjacent leads, the peaks and troughs of the waveforms may overlap, such as... Figure 3 The waveforms of leads V4, V5, and V6 are shown in the figure. Therefore, in step 122, adjacent lead waveform images with overlapping waveforms in the image to be processed are divided into the same image region.

[0052] See Figure 5 This schematically illustrates multiple image regions to be processed, formed by dividing a local area of ​​the lead waveform image to be processed. For example... Figure 5 As shown, in the lead waveform image 300 to be processed, the continuous lead waveform image is divided into separate regions of the image to be processed, for example... Figure 5 The image region 310 to be processed is shown, and adjacent lead waveforms with overlapping waveforms are also classified as belonging to the same image region to be processed, for example... Figure 5 The image region 320 to be processed is shown. Therefore, in Figure 5In the local area of ​​the lead waveform image 300 to be processed shown, 8 image regions similar to image region 310 to be processed and 2 image regions similar to image region 320 to be processed are divided.

[0053] See Figure 6 According to an exemplary embodiment of this disclosure, it further illustrates Figure 1 Details of step 130 in the illustrated ECG report digitization processing method 100. (See example...) Figure 6 As shown, step 130 of the ECG report digitization processing method 100 may include steps 131, 132, 133, 134 and 135.

[0054] In step 131, overlapping lead connectivity regions are determined in the image region to be processed, wherein the overlapping lead connectivity regions surround at least one pair of overlapping troughs and peaks. See also Figure 7 This schematically illustrates an overlapping lead connectivity region identified in the area of ​​the image to be processed. For example... Figure 7 As shown, the image region to be processed includes the upper lead waveform Wave1 and the lower lead waveform Wave2, with three overlapping waveforms between them. Each overlapping waveform includes a trough in lead waveform Wave1 and a peak in lead waveform Wave2. Figure 7 The leftmost overlapping waveform shown indicates the peak pixel P of the lead waveform Wave1. top And determine the trough pixel P of the lead waveform Wave2. bottom Therefore, based on the peak pixel P top And the trough pixel P bottom This allows us to determine the upper and lower boundaries of the overlapping lead connectivity region C1 corresponding to the overlapping waveform at that location. Then, we can use the peak pixel P... top Based on the baseline, the distance from the peak pixel P on both its left and right sides is... top The left and right boundaries of the overlapping lead connectivity region C1 are determined at a preset distance ΔP, where the preset distance ΔP can be a pre-set number of pixels. It should be understood that... Figure 7 The image shows the peak pixel P. top The left and right boundaries of the overlapping lead connectivity region C1 are determined based on the baseline; however, the valley pixel P is used as the reference. bottom Using this as a benchmark to determine the left and right boundaries of the connected region C1 of the overlapping leads is also acceptable. Furthermore, a similar approach can be used to... Figure 7The other two overlapping waveforms shown in the image respectively determine the corresponding overlapping lead connectivity regions. It should be understood that, in this disclosure, the phrase "the overlapping lead connectivity region surrounds at least one pair of overlapping troughs and peaks" mentioned in step 131 should be interpreted broadly, that is, it includes both the case where the overlapping lead region surrounds one overlapping trough and one overlapping peak (e.g., ...). Figure 7 (As shown in the diagram), this also includes situations where, when there is significant overlap, the overlapping lead connection region surrounds the overlapping multiple ECG lead waveforms.

[0055] See also Figure 6 In step 132, a non-overlapping region is determined within the overlapping lead connectivity region, wherein the non-overlapping region is formed by being surrounded by a preliminarily determined lead waveform including the peak and the boundary of the overlapping lead connectivity region, and includes the preliminary position of the peak. As a non-limiting example, edge information can be used to preliminarily determine the non-overlapping region within the overlapping lead connectivity region. See also Figure 8 See also Figure 6 This schematically illustrates the overlapping peaks and troughs in an overlapping waveform. For example... Figure 8 As shown in the magnified view, the troughs of lead waveform Wave1 and the peaks of lead waveform Wave2 are slightly staggered in time. Therefore, the preliminary positions of the peaks and / or troughs can be determined in the connected region of the overlapping leads using edge information.

[0056] Continue to see Figure 6 In step 133, the non-overlapping region is shifted along the direction indicated by the wave peak to form an offset non-overlapping region; in step 134, the Manhattan distance, cross-union ratio, Hamming distance, and correlation coefficient between each offset non-overlapping region and the overlapping lead connected region are weighted and summed to obtain the composite similarity between each offset non-overlapping region and the overlapping lead connected region; in step 135, the offset corresponding to the minimum value of the composite similarity is added to the initial position of the wave peak to determine the accurate position of the wave peak in the overlapping waveform.

[0057] See Figure 9 And in conjunction with see Figure 6 This schematically illustrates the process of determining the precise location of wave crests in overlapping waveforms based on composite similarity (i.e., steps 133, 134, and 135 above). Figure 9View (a) shows the identified overlapping lead connectivity region (CRO), which has a rectangular shape and includes the waveforms of the two leads where the waveforms overlap. As mentioned above, edge information can be used to initially determine the non-overlapping region (CNR) (i.e., the gray area in this view) within the overlapping lead connectivity region. Because the pixels in the overlapping region are too dense, the edge information may not accurately reflect the location of the true peaks. Figure 9 View (a) shows the preliminary position F-Peak of the peak identified in the non-overlapping region CNR, but it is still some distance from the accurate position A-Peak. Therefore, the preliminary position F-Peak represents a false peak position, and the accurate position A-Peak needs to be determined. In view (a), the bottom edge of the non-overlapping region CNR coincides with the bottom edge of the overlapping lead connection region CRO, so the offset between them is 0. To determine the accurate position A-Peak of the peak, the non-overlapping region CNR is aligned along the direction indicated by the peak (i.e., Figure 9 Move in the direction indicated by arrow A, that is, make the offset between the two non-overlapping regions (CNR) not equal to 0. Calculate the composite similarity (WS) between the non-overlapping region (CNR) and the overlapping lead connected region (CRO) for every 1 pixel offset. Figure 9 View (b) shows the case where the bottom edge of the non-overlapping region CNR is offset by 22 pixels relative to the bottom edge of the overlapping lead connected region CRO. In this case, the initial position F-Peak roughly coincides with the accurate position A-Peak. Figure 9 View (c) shows the case where the bottom edge of the non-overlapping region CNR is offset by 63 pixels relative to the bottom edge of the overlapping lead connected region CRO. In this case, the initial position F-Peak has already passed the accurate position A-Peak and is directly above it. Figure 9 View (d) shows the curve of the composite similarity WS between the non-overlapping region CNR and the overlapping lead connected region CRO, relative to the offset. As the curve shows, the composite similarity WS gradually decreases as the offset increases from 0 pixels to 22 pixels, and then gradually increases as the offset continues to increase. Therefore, the offset corresponding to the composite similarity WS is the distance between the initial position F-Peak of the peak and the accurate position A-Peak of the peak. Thus, as... Figure 9As shown, using the non-overlapping region CNR as a template, the composite similarity WS between the non-overlapping region CNR and the overlapping lead connected region CRO is calculated by sliding along direction A in the overlapping lead connected region CRO. The value of offset Offset is determined based on the minimum value of composite similarity WS. Then, combined with the preliminary position F-Peak of the peak in the non-overlapping region CNR, the accurate position of the peak in the overlapping waveform can be determined.

[0058] The composite similarity WS between the non-overlapping region CNR and the overlapping lead connected region CRO is related to their Manhattan distance, crossover-union ratio, Hamming distance, and correlation coefficient. Therefore, it can be calculated using the following formula:

[0059]

[0060] Among them, D M The Manhattan distance (IoU) represents the distance between each pixel in the non-overlapping region CNR and each pixel in the overlapping lead connected region CRO. The cross-union ratio (IoU) represents the degree of image overlap between the non-overlapping region CNR and the overlapping lead connected region CRO. H R is the Hamming distance, which represents the number of pixels with different values ​​between the non-overlapping region CNR and the overlapping lead connected region CRO. R is the correlation coefficient, which is calculated using the variance of the pixel values ​​in the non-overlapping region CNR and the overlapping lead connected region CRO, as well as the covariance of their pixel values. Specifically, the Manhattan distance D... M Intersection over Union (IoU) and Hamming distance (D) H The correlation coefficient R can be calculated using the following formulas:

[0061]

[0062] Among them, CNR offset The non-overlapping region after offset is represented by x and y, where x and y represent the positions of each pixel in the overlapping lead connected region CRO, and n and m represent the number of pixels in the length and width of the overlapping lead connected region CRO, respectively. In Equation 5, Cov represents the calculation of covariance, and D represents the calculation of variance. It should be understood that, for ease of calculation, the non-overlapping regions CNR after offset in the above formulas are... offset The non-overlapping region CNR has the same number of pixels as the non-overlapping region CNR, i.e., n×m pixels, except that the non-overlapping region CNR is offset. offset In addition to image regions (e.g., Figure 9 In views (a), (b), and (c), all pixels except those in the grayscale image region are set to 0. Refer to Equation 1, where coefficients w1, w2, w3, and w4 are similarity weight coefficients.

[0063] The similarity weight coefficients w1, w2, w3, and w4 can be calculated based on existing data using existing optimization algorithms. For example, the Nelder-Mead simplex algorithm can be used to predetermine the values ​​of the similarity weight coefficients w1, w2, w3, and w4. As an example, Table 1 below shows the similarity weight coefficients w1, w2, w3, and w4 obtained by iterative calculation using the Nelder-Mead simplex algorithm based on existing ECG image data.

[0064] Table 1

[0065] w1 w2 w3 w4 initial value 0.8 1 1 1 The Lower World 0.0000001 0.1 0.1 1 Upper Realm 0.8 10 10 10 Optimization result 0.7998 0.1000 0.1000 1.0000

[0066] As shown in Table 1, for each of the similarity weight coefficients w1, w2, w3 and w4, corresponding upper and lower bounds are set, and then iterative calculations are performed based on the preset initial values ​​to obtain the optimal results of each coefficient. These optimal results can be used as the values ​​of the similarity weight coefficients w1, w2, w3 and w4.

[0067] See Figure 10 This schematically illustrates the relationship between the iterative calculation of similarity weight coefficients and composite similarity. For example... Figure 10 As shown, with the increase of the number of iterations, the composite similarity WS calculated based on the similarity weight coefficients w1, w2, w3, and w4 for the existing ECG image data gradually decreases and tends to stabilize at a minimum value. Therefore, the iterative calculation results of the similarity weight coefficients w1, w2, w3, and w4 corresponding to the minimum value of this composite similarity WS can be determined as the optimal results in Table 1, and thus can be determined as the values ​​of the similarity weight coefficients w1, w2, w3, and w4.

[0068] See Figure 11 According to an exemplary embodiment of this disclosure, it further illustrates Figure 1 Details of step 140 in the illustrated ECG report digitization processing method 100. (See attached image.) Figure 11 As shown, step 140 of the ECG report digitization processing method 100 may include steps 141, 142, 143, 144, and 145:

[0069] In step 141, the pixel values ​​of each row of pixels in the lead waveform image to be processed are summed to generate the pixel intensity of each row of pixels.

[0070] In step 142, the row corresponding to each peak pixel intensity in all pixel intensities is determined as the row index in the image of the lead waveform to be processed;

[0071] In step 143, the centroid of the pixel distribution in each image region to be processed is calculated;

[0072] In step 144, the column corresponding to each centroid is determined as the column index in the image of the lead waveform to be processed;

[0073] In step 145, based on the row index and the column index, as well as prior knowledge about the distribution of different types of lead waveforms in the paper electrocardiogram report, the lead category corresponding to the lead waveform image in each image region to be processed is determined.

[0074] See Figure 12 See also Figure 11 This schematically illustrates the process of determining row and column indices in the image of the lead waveform to be processed, thereby determining the lead category. For example... Figure 12 As shown, the pixel values ​​of each row of pixels in the lead waveform image 400 are summed to generate the pixel intensity of each row, thus obtaining... Figure 12 The pixel intensity distribution curve 410 is shown in the figure. Therefore, the row corresponding to the peak value in the pixel intensity distribution curve 410 can be determined as the row index in the lead waveform image to be processed. In addition, the centroid of each region of the image to be processed is calculated in the lead waveform image 400, i.e. Figure 12 The dots in each image region to be processed are displayed. Then, the columns corresponding to these centroids can be determined as column indices in the lead waveform image to be processed. As mentioned above... Figure 3 As described, in existing paper ECG reports, the waveforms of each lead category have fixed positions. Therefore, by combining prior knowledge of the distribution positions of different lead categories in the paper ECG report, and based on the determined row and column indices, it is possible to determine the lead category corresponding to the lead waveform image in each image region to be processed.

[0075] See Figure 13 According to an exemplary embodiment of this disclosure, it further illustrates Figure 1 Details of step 150 in the illustrated ECG report digitization processing method 100. (See attached image.) Figure 13 As shown, step 150 of the electrocardiogram report digitization processing method 100 may include steps 151 and 152:

[0076] In step 151, based on the reference pulse waveform, the ratio between pixels and voltage values ​​in the image to be processed of the lead waveform is determined;

[0077] In step 152, based on the determined ratio between pixels and voltage values, the lead waveform images in each image region to be processed are converted into corresponding voltage value data.

[0078] See Figure 14 See also Figure 13 , Figure 14 This schematically illustrates the extraction of data from lead waveform images based on reference pulse waveforms. For example... Figure 14 As shown, the ECG report contains a reference pulse waveform used as a reference benchmark, where the reference pulse is typically a calibration pulse with a known amplitude (e.g., 1mV). Figure 14 The reference pulse waveform is indicated by a black dashed box. By determining the number of pixels in the reference pulse waveform, the ratio between pixels and voltage values ​​can be determined. Then, using the determined ratio, the waveform images of each lead can be converted into corresponding voltage values. In this way, the ECG report digitization processing method 100 can convert two-dimensional image data in the waveform images of each lead into a one-dimensional signal that accurately reflects the true signal amplitude (e.g., in mV).

[0079] According to the ECG report digitization method 100 and its exemplary embodiments disclosed herein, not only can fully automated batch digitization processing of paper ECG reports be provided, but also the problems of insufficient overlapping region extraction capability and high computational cost are simultaneously solved by using image processing algorithms with low computational overhead and adaptive composite correlation algorithms. Furthermore, customizable parameters allow for compatibility with multiple ECG report formats. In addition, it should be understood that the above description pertains to... Figure 1 The various details further described in the steps of the illustrated method are exemplary and not limiting; therefore, there are various ways to implement them. Figure 1 Alternative schemes for each step of the method shown are provided, and all such alternative schemes are considered to fall within the scope of this disclosure.

[0080] Table 2 below shows the evaluation results of comparing the ECG report processed using the ECG report digitization method 100 taught in this disclosure with the paper ECG report provided by the hospital (i.e., comparing the digitized results with the true values ​​provided by the paper ECG report). The PCC (Pearson correlation coefficient) and RMSE (Root Mean Square Error) values ​​for each ECG lead indicate that the digitization method according to this disclosure has high accuracy and reliability. For example, the PCC values ​​for all leads are above 0.971, with the highest PCC value in lead V1 at 0.984; the RMSE values ​​also show a low error rate, with average values ​​ranging from 0.026 mV to 0.047 mV across different leads. Furthermore, the low standard deviation of PCC and RMSE indicates good consistency in the digitization results of this method.

[0081] Table 2

[0082]

[0083] See Figure 15This illustration schematically shows a comparison between the waveform plotted from the digitized result obtained using the ECG report digitization method of this disclosure and the waveform in the original paper ECG report. Figure 15 As shown, the waveform plotted based on the digitization results matches the waveform in the original paper ECG report very well. Figure 15 The enlarged view shows the waveforms of adjacent leads where the waveforms overlap, where it can be seen that the exact location of the peak in the identified overlapping waveform matches the location of the corresponding peak in the original paper ECG report very well.

[0084] See Figure 16 The illustration schematically shows a comparison between the waveforms of each lead plotted using the electrocardiogram report digitization method of this disclosure and the waveforms of each lead in the original paper ECG report. Figure 16 As shown, the waveforms of each lead plotted based on the digitization results also match the waveforms of each lead in the original paper ECG report very well.

[0085] See Figure 17 The diagram schematically illustrates an electrocardiogram report digitization processing apparatus according to an exemplary embodiment of the present disclosure. Figure 17As shown, the electrocardiogram (ECG) report digitization processing device 500 includes: an image preprocessing module 510, a to-be-processed image region generation module 520, an overlapping waveform processing module 530, a lead category determination module 540, a data extraction module 550, and a digitized report generation module 560. Specifically, the image preprocessing module 510 can be configured to preprocess a paper ECG report to obtain basic report information and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed. The to-be-processed image region generation module 520 can be configured to perform connected component detection on the lead waveform images to be processed, dividing the lead waveform images to be processed into multiple to-be-processed image regions. The overlapping waveform processing module 530 can be configured to: when adjacent lead waveform images in the to-be-processed image regions have overlapping waveforms, move a portion of the waveform including the overlapping peaks relative to the overlapping waveforms, so as to determine the accurate position of the peaks in the overlapping waveforms based on the composite similarity between the portion of the waveforms and the overlapping waveforms. The lead category determination module 540 can be configured to determine the lead category corresponding to the lead waveform image in each image region to be processed based on the position of each image region to be processed in the lead waveform image to be processed. The data extraction module 550 can be configured to extract the electrocardiogram data corresponding to the lead category from the lead waveform image in each image region to be processed based on the reference pulse waveform in the lead waveform image to be processed. The digital report generation module 560 can be configured to generate a digital report corresponding to the paper electrocardiogram report based on the basic report information and the electrocardiogram data. The electrocardiogram report digitization processing apparatus 500 can apply the electrocardiogram report digitization processing method 100 and its exemplary embodiments described in detail above.

[0086] In this manner, the ECG report digitization processing apparatus 500 according to this disclosure can provide fully automatic batch digitization processing of paper ECG reports, and can simultaneously solve the problems of insufficient overlapping area extraction capability and high computational cost with image processing algorithms with low computational overhead and adaptive composite correlation algorithms, and can also customize adjustable parameters to be compatible with multiple ECG report formats.

[0087] It should be understood that, Figure 17 The modules shown in the text relate to the topics discussed above. Figure 1The operation of the corresponding steps in the ECG report digitization processing method 100 and its exemplary embodiments described herein will not be repeated here. Furthermore, the various modules described in this disclosure can be implemented in hardware or in hardware combined with software and / or firmware. For example, these modules can be implemented as computer-executable code / instructions configured to execute in one or more processors and stored in a computer-readable storage medium. Alternatively, these modules can be implemented as hardware logic / circuit. As a non-limiting example, one or more of these modules can be implemented in a system-on-a-chip (SoC). An SoC may include an integrated circuit chip (which includes a processor (e.g., a central processing unit, microcontroller, microprocessor, digital signal processor, etc.), memory, one or more communication interfaces, and / or one or more components of other circuitry) and may optionally execute received program code and / or include embedded firmware to perform functions.

[0088] See Figure 18 This schematically illustrates the structure of a computer device according to an exemplary embodiment of the present disclosure. Figure 18 As shown, the computer device 700 can be used in the various exemplary embodiments described in this disclosure, and it can implement the corresponding steps in the electrocardiogram report digitization processing method 100 and its various exemplary embodiments described in this disclosure.

[0089] Computer device 700 may include at least one processor 702, memory 704, multiple communication interfaces 706, display device 708, other input / output (I / O) devices 710, and one or more mass storage devices 712 that are capable of communicating with each other, such as via system bus 714 or other suitable means.

[0090] Processor 702 may be a single processing unit or multiple processing units, and all processing units may include single or multiple computing units or multiple cores. Processor 702 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuits, and / or any device that manipulates signals based on operating instructions. Among other capabilities, processor 702 may be configured to acquire and execute computer-readable instructions stored in memory 704, mass storage device 712, or other computer-readable storage media, such as program code of operating system 716, program code of application program 718, program code of other program 720, etc.

[0091] Memory 704 and mass storage device 712 are examples of computer-readable storage media for storing instructions that can be executed by processor 702 to perform the various functions described above. For example, memory 704 can generally include both volatile and non-volatile memory (e.g., RAM, ROM, etc.). Furthermore, mass storage device 712 can generally include hard disk drives, solid-state drives, removable media, including external and removable drives, memory cards, flash memory, floppy disks, optical disks (e.g., CDs, DVDs), storage arrays, network-attached storage, storage area networks, etc. Both memory 704 and mass storage device 712 can be collectively referred to herein as computer-readable memory or computer-readable storage media, and can be non-transitory media capable of storing computer-readable, processor-executable program instructions as computer-executable code that can be executed by processor 702 as a specific machine configured to perform the operations and functions described in the various exemplary embodiments of this disclosure.

[0092] Multiple program modules can be stored on a mass storage device 712. These program modules may include an operating system 716, one or more application programs 718, other programs 720, and program data 722, and they can be executed by a processor 702. Examples of such application programs or program modules may include, for example, computer program logic (e.g., computer-executable code or instructions) for implementing the following components / functions: an image preprocessing module 510, a to-be-processed image region generation module 520, an overlapping waveform processing module 530, a lead category determination module 540, a data extraction module 550, and a digitization report generation module 560.

[0093] Although Figure 18The modules are illustrated as being stored in memory 704 of computer device 700, but the various modules or portions thereof described above can be implemented using any form of computer-readable storage medium accessible by computer device 700. As used herein, "computer-readable storage medium" includes at least two types of computer-readable storage media: computer storage media and communication media. Computer storage media includes volatile and non-volatile, removable and non-removable media implemented by any method or technique for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, Digital Universal Disc (DVD), or other optical storage devices, magnetic cassettes, magnetic tapes, disk storage devices or other magnetic storage devices, or any other non-transmission medium that can be used to store information for access by a computing device. In contrast, communication media can embody computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms. Computer storage media as defined in this disclosure do not include communication media.

[0094] Computer device 700 may also include one or more communication interfaces 706 for exchanging data with other devices, such as via a network, direct connection, etc. Communication interface 706 can facilitate communication across various network and protocol types, including wired networks (e.g., LAN, cable, etc.) and wireless networks (e.g., WLAN, cellular, satellite, etc.), the Internet, etc. Communication interface 706 can also provide communication with external storage devices (not shown), such as storage arrays, network-attached storage, storage area networks, etc.

[0095] In some examples, computer device 700 may also include a display device 708, such as a monitor, for displaying information and images. Other I / O devices 710 may be devices that receive various inputs from a target object and provide various outputs to the target object, including but not limited to touch input devices, gesture input devices, cameras, keyboards, remote controls, mice, printers, audio input / output devices, and so on.

[0096] This disclosure also relates to a computer-readable storage medium configured to store computer-executable instructions configured to, when executed on a processor, cause the processor to perform the electrocardiogram report digitization processing method according to various exemplary embodiments of this disclosure. It should be understood that the computer-readable storage medium should be any suitable storage medium, including, but not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD), or other optical storage devices, magnetic cassettes, magnetic tapes, disk storage devices or other magnetic storage devices, or semiconductor media (e.g., solid-state drives), or any other non-transmission medium that can be used to store information for access by a computing device. This disclosure does not limit the scope of the computer-readable storage medium.

[0097] Furthermore, this disclosure also relates to a computer program product including computer-executable instructions configured to, when executed on a processor, cause the processor to perform the electrocardiogram report digitization processing method according to various exemplary embodiments of this disclosure.

[0098] When implemented using software, the electrocardiogram (ECG) report digitization processing method according to this disclosure can be implemented, in whole or in part, as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the exemplary embodiments of this disclosure are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0099] The terminology used in this disclosure is for the purpose of describing embodiments thereof and is not intended to limit the disclosure. As used herein, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the terms “comprising” and “including,” as used herein, refer to the presence of the stated feature but do not exclude the presence of one or more other features or the addition of one or more other features. As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items. It will be understood that although the terms “first,” “second,” “third,” etc., may be used in this disclosure to describe various features, these features should not be limited by these terms. These terms are used only to distinguish one feature from another.

[0100] Unless otherwise defined, all terms used in this disclosure (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It should also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the relevant field and / or the context of this specification, and will not be interpreted in an idealized or overly formal sense unless expressly defined in this disclosure.

[0101] In this disclosure, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of a larger module or unit that includes the functionality of that module or unit.

[0102] In the description of this disclosure, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction or violation of technical principles, those skilled in the art can combine and integrate different embodiments or examples and features of different embodiments or examples described in this specification, or omit some technical features from different embodiments or examples described in this specification, and embodiments or examples obtained based on such combinations, arrangements, or omissions are also considered to fall within the scope of this disclosure.

[0103] The methods described in this disclosure include one or more steps or actions. These method steps and / or actions do not necessarily have to be performed in the order described in this disclosure, but can be performed in different orders. For example, these method steps and / or actions can be performed simultaneously or in reverse order, as long as this does not contradict the principles of the technical solutions described in this disclosure. Furthermore, depending on actual needs, the steps and / or actions in the methods described in this disclosure can be replaced with different steps and / or actions, or additional steps and / or actions may be included.

[0104] While at least one exemplary embodiment has been described in the foregoing detailed description, it should be understood that numerous variations exist. It should also be understood that the one or more exemplary embodiments described herein are merely examples and are not intended to limit the scope, applicability, or construction of this application in any way. Rather, the foregoing detailed description will provide convenient guidance to those skilled in the art in implementing one or more exemplary embodiments. It should be understood that various changes, modifications, or alterations can be made to the function and arrangement of elements without departing from the scope of this application as set forth by the appended claims and their equivalents. The scope of protection of this disclosure is defined only by the appended claims.

Claims

1. A method for digital processing of electrocardiogram (ECG) reports, characterized in that, include: The paper electrocardiogram report is preprocessed to obtain basic information of the report and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed. Connectivity detection is performed on the lead waveform image to be processed, dividing the lead waveform image to be processed into multiple image regions to be processed; when there are overlapping waveforms in adjacent lead waveform images in the image regions to be processed, a portion of the waveform including the overlapping peaks is moved relative to the overlapping waveform, so as to determine the accurate position of the peaks in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform. Based on the position of each image region to be processed in the lead waveform image to be processed, the lead category corresponding to the lead waveform image in each image region to be processed is determined. Based on the reference pulse waveform in the image to be processed of the lead waveform, the electrocardiogram data of the corresponding lead category is extracted from the lead waveform image in each image region to be processed; Based on the basic information in the report and the electrocardiogram data, a digital report corresponding to the paper electrocardiogram report is generated.

2. The method for digital processing of electrocardiogram reports according to claim 1, characterized in that, The preprocessing of the paper electrocardiogram report to obtain basic report information and lead waveform images to be processed includes: The paper electrocardiogram report is scanned and identified to obtain the basic information of the report and the original images of the lead waveforms. Convert the original image of the lead waveform from the RGB color space to the HSV color space; Adjust the hue, saturation, and brightness of the red color in the original image of the lead waveform to determine the position of the grid lines; The original image of the lead waveform is binarized to generate a binarized image of the original lead waveform. In the binarized image, the pixel values ​​at the positions of the grid lines are set to zero, thereby removing the grid lines and generating the lead waveform image to be processed.

3. The method for digital processing of electrocardiogram reports according to claim 1, characterized in that, The step of performing connected component detection on the lead waveform image to be processed, and dividing the lead waveform image to be processed into multiple image regions to be processed, includes: The continuous lead waveform images in the image to be processed are divided into separate image regions to be processed; Adjacent lead waveform images with overlapping waveforms in the image to be processed are divided into the same image region.

4. The method for digital processing of electrocardiogram reports according to claim 1, characterized in that, When adjacent lead waveforms in the image region to be processed overlap, the step of shifting a portion of the waveform including the overlapping peak relative to the overlapping waveform, so as to determine the accurate position of the peak in the overlapping waveform based on the composite similarity between the portion of the waveform and the overlapping waveform, includes: In the image region to be processed, an overlapping lead connectivity region is determined, wherein the overlapping lead connectivity region surrounds at least one pair of overlapping troughs and peaks; In the overlapping lead connectivity region, a non-overlapping region is determined, wherein the non-overlapping region is formed by the preliminarily determined lead waveform including the peak and the boundary of the overlapping lead connectivity region and includes the preliminary position of the peak; The non-overlapping region is shifted along the direction indicated by the wave crest to form an offset non-overlapping region; The Manhattan distance, cross-union ratio, Hamming distance, and correlation coefficient between each offset non-overlapping region and the overlapping lead connected region are weighted and summed to obtain the composite similarity between each offset non-overlapping region and the overlapping lead connected region. The exact position of the peak in the overlapping waveform is determined by adding the offset corresponding to the minimum value of the composite similarity to the initial position of the peak.

5. The method for digital processing of electrocardiogram reports according to claim 1, characterized in that, The determination of the lead category corresponding to each lead waveform image in each lead waveform image region based on the position of each lead waveform image region in the lead waveform image to be processed includes: In the image of the lead waveform to be processed, the pixel values ​​of each row of pixels are summed to generate the pixel intensity of each row of pixels; The row corresponding to the peak pixel intensity in all pixel intensities is determined as the row index in the image of the lead waveform to be processed. Calculate the centroid of the pixel distribution in each region of the image to be processed; The column corresponding to each centroid is determined as the column index in the image of the lead waveform to be processed; Based on the row index and column index, as well as prior knowledge about the distribution of different types of lead waveforms in paper electrocardiogram reports, the lead category corresponding to the lead waveform image in each image region to be processed is determined.

6. The method for digital processing of electrocardiogram reports according to claim 1, characterized in that, The step of extracting ECG data corresponding to the lead category from the lead waveform images in each image region to be processed, based on the reference pulse waveform in the image to be processed, includes: Based on the reference pulse waveform, determine the ratio between pixels and voltage values ​​in the image to be processed of the lead waveform; Based on the determined ratio between pixels and voltage values, the lead waveform images in each image region to be processed are converted into corresponding voltage value data.

7. A digital processing device for electrocardiogram (ECG) reports, wherein the ECG report digital processing method according to any one of claims 1 to 6 is characterized in that, include: The image preprocessing module is configured to: preprocess the paper electrocardiogram report to obtain basic information of the report and lead waveform images to be processed, wherein the lead waveform images to be processed are binarized images with grid lines removed; The image region generation module is configured to: perform connected component detection on the lead waveform image to be processed, and divide the lead waveform image to be processed into multiple image regions to be processed; An overlapping waveform processing module is configured to: when adjacent lead waveform images in the image region to be processed have overlapping waveforms, move a portion of the waveform including the overlapping peaks relative to the overlapping waveforms, so as to determine the accurate position of the peaks in the overlapping waveforms based on the composite similarity between the portion of the waveforms and the overlapping waveforms. The lead category determination module is configured to determine the lead category corresponding to the lead waveform image in each image region to be processed based on the position of each image region to be processed in the lead waveform image to be processed. The data extraction module is configured to: extract electrocardiogram data of the corresponding lead category from the lead waveform images in each image region to be processed, based on the reference pulse waveform in the lead waveform image to be processed; The digital report generation module is configured to generate a digital report corresponding to the paper electrocardiogram report based on the basic information of the report and the electrocardiogram data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the electrocardiogram report digitization processing method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the electrocardiogram report digitization processing method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the electrocardiogram report digitization processing method according to any one of claims 1 to 6.

Images