A paper electrocardiogram digitization method and system based on slope adaptive correction

Abstract

The present application relates to the technical field of automatic electrocardiogram analysis, and particularly relates to a paper electrocardiogram digitization method and system based on slope adaptive correction. The method comprises the following steps: obtaining a paper electrocardiogram, and performing binaryzation processing on the paper electrocardiogram; determining the positions of each lead and QRS wave peak in the electrocardiogram, and completing lead segmentation; extracting digital information of each lead, and converting the binaryzation electrocardiogram of each lead into a digital reconstruction electrocardiogram signal; dynamically generating a left adaptive slope and a right adaptive slope of the QRS wave; and reconstructing the QRS wave according to the left adaptive slope and the right adaptive slope of the QRS wave. The method provided by the present application generates a dynamic slope according to the morphology of the QRS wave, and then reconstructs the QRS wave according to the dynamic slope, which can effectively remove the influence of lead adhesion on the QRS wave peak, and at the same time reflect the characteristic differences of each heart signal, so as to ensure that the digitized data can accurately reflect the original information of the paper electrocardiogram.

Description

Technical Field

[0001] This invention relates to the field of automated electrocardiogram (ECG) analysis technology, and in particular to a method and system for digitizing paper ECGs based on slope adaptive correction. Background Technology

[0002] Electrocardiograms (ECGs), as core clinical data recording cardiac electrical activity, play an irreplaceable role in the diagnosis, monitoring, and evaluation of treatment effects in cardiovascular diseases. Currently, a large number of paper ECG records remain in clinical practice. These records contain long-term diagnostic and treatment data and are important resources for retrospective studies, chronic disease follow-up, and multi-center data integration. With the advancement of medical informatization, the digitization of paper ECGs has become an inevitable trend. Digitization not only solves the problems of large storage space, susceptibility to environmental factors (such as moisture and aging), and low retrieval efficiency of paper records, but also provides a foundation for subsequent computer-aided analysis (such as heart rate variability analysis and automatic arrhythmia identification), remote consultation, and cross-institutional data sharing. It is a key link in achieving precision diagnosis and treatment of cardiovascular diseases and smart healthcare. The digitization of paper ECGs typically involves five core steps: image acquisition, preprocessing, lead localization, waveform extraction, and parameter quantification. Among these, preprocessing and waveform extraction directly determine the accuracy of the digitization results.

[0003] Chinese patent application number ZL202510318904.7, entitled "Method and System for Reconstructing Paper Electrocardiogram Voltage Values ​​Based on Dynamic Diffusion Threshold," discloses a method for digitizing paper electrocardiograms, including the following steps: acquiring a standard paper electrocardiogram and a paper electrocardiogram to be processed, and extracting paper electrocardiogram parameters; iteratively optimizing based on a preset threshold to obtain an optimal threshold for binarizing the paper electrocardiogram to be processed, thereby obtaining a binarized paper electrocardiogram; locating the electrocardiogram waveform leads; converting the electrocardiogram waveform information of the lead image region into a digital signal; converting the acquired digital signal into a voltage signal; and resampling the voltage signal to obtain a digitized reconstructed electrocardiogram signal.

[0004] This invention can make the edges of the ECG waveform clearer during the digitization process and alleviate problems such as paper stains, scanning noise, and baseline drift. However, lead adhesion, as a complex interference unique to paper ECGs, is still a key bottleneck restricting the accuracy of digitization. The overlapping of the edges of adjacent leads can easily cause the QRS waveform, especially the peak, to be deformed after the adhesion area is segmented, resulting in the loss of key clinical information.

[0005] Static correction strategies are insufficient to achieve high-precision correction of lead adhesions. There is an urgent need for a method that can dynamically adjust the correction waveform to solve this core bottleneck problem in the digitization of paper electrocardiograms. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes a paper electrocardiogram digitization method and system based on slope adaptive correction. This method enables adaptive, high-precision, and automated lead adhesion correction, thereby ensuring that the digitized electrocardiogram data accurately reflects the original information of the paper electrocardiogram.

[0007] In a first aspect, this invention proposes a method for digitizing paper electrocardiograms based on slope adaptive correction, the steps of which include:

[0008] Obtain a paper electrocardiogram (ECG) and perform binarization on it;

[0009] Determine the location of each lead and the position of the QRS wave peak in the electrocardiogram, and complete the lead segmentation;

[0010] The digital information of each lead is extracted, and the binary electrocardiogram of each lead is converted into a digitally reconstructed electrocardiogram signal. The digitally reconstructed electrocardiogram signal of each lead contains multiple heartbeat cycles, and each heartbeat cycle contains a QRS wave.

[0011] The method is characterized in that, after converting the binary electrocardiogram of each lead into a digitally reconstructed electrocardiogram signal, the method further includes:

[0012] Dynamically generate the left and right adaptive slopes of the QRS wave;

[0013] QRS wave reconstruction is performed based on the left and right adaptive slopes of the QRS wave.

[0014] Furthermore, the dynamic generation of the left adaptive slope and right adaptive slope of the QRS wave specifically includes:

[0015] The baseline R_ECG-baseline of the lead is determined based on its position.

[0016] Select a QRS wave from the digitally reconstructed electrocardiogram signal;

[0017] Based on the position of the baseline R_ECG-baseline and the peak position of the QRS wave, remove the portion of the QRS wave affected by lead adhesion;

[0018] The leftward search is performed from the peak position of the QRS wave to obtain two dynamic points on the left side, and an adaptive slope of the left side of the QRS wave is generated based on the two dynamic points on the left side.

[0019] By performing a right-side search from the peak position of the QRS wave, two dynamic points on the right side are obtained, and an adaptive slope on the right side of the QRS wave is generated based on the two dynamic points on the right side.

[0020] Repeat the above four steps until all QRS waves generate left-side adaptive slopes and right-side adaptive slopes.

[0021] Furthermore, the step of removing the portion of the QRS wave affected by lead adhesion based on the baseline position R_ECG-baseline and the peak position of the QRS wave specifically includes:

[0022] Determine whether the wavelet containing the QRS crest is a forward or reverse wave;

[0023] If it is a positive wave, the baseline R_ECG-baseline is moved up by a preset distance to form R'_ECG-baseline, and the signal of the sub-wavelength where the QRS peak is located above the R'_ECG-baseline is removed;

[0024] If it is a reverse wave, the baseline R_ECG-baseline is lowered by a preset distance to form R'_ECG-baseline, and the signal of the sub-wavelength where the QRS peak is located below the R'_ECG-baseline is removed.

[0025] Furthermore, the step of performing a leftward search from the peak position of the QRS wave to obtain two dynamic points on the left, and generating a leftward adaptive slope of the QRS wave based on the two dynamic points on the left, specifically includes:

[0026] Starting from the peak position of the QRS wave, the horizontal axis is moved to the left until the QRS wave curve intersects with the R'_ECG-baseline for the first time. This intersection point is taken as the first dynamic point on the left side of the QRS wave.

[0027] Starting from the peak position of the QRS wave, the horizontal axis is moved to the left until the QRS wave curve produces the first extreme point. This extreme point is taken as the second left dynamic point of the QRS wave.

[0028] Based on the first and second left-side dynamic points, the left-side adaptive slope of the QRS wave is generated.

[0029] Furthermore, the step of performing a right-side search from the peak position of the QRS wave to obtain two dynamic points on the right side, and generating a right-side adaptive slope of the QRS wave based on the two dynamic points on the right side, specifically includes:

[0030] Starting from the peak position of the QRS wave, the horizontal axis is moved to the right until the QRS wave curve intersects with the R'_ECG-baseline for the first time. This intersection point is taken as the first right dynamic point of the QRS wave.

[0031] Starting from the peak position of the QRS wave, the horizontal axis is moved to the right until the QRS wave curve produces the first extreme point. This extreme point is taken as the second right-side dynamic point of the QRS wave.

[0032] Based on the first and second right-side dynamic points, the right-side adaptive slope of the QRS wave is generated.

[0033] Furthermore, the QRS wave reconstruction based on the left and right adaptive slopes of the QRS wave specifically includes:

[0034] A straight line is extended based on the left and right adaptive slopes of each QRS wave until the left and right extensions of each QRS wave intersect.

[0035] QRS wave reconstruction is performed based on the two intersecting extensions.

[0036] Secondly, this invention proposes a paper electrocardiogram (ECG) digitization system based on slope adaptive correction, to implement the paper ECG digitization method based on slope adaptive correction as described in the first aspect of this invention. The system includes:

[0037] Data acquisition module: Acquires standard paper electrocardiograms and paper electrocardiograms to be processed, and extracts paper electrocardiogram parameters;

[0038] Binarization module: Performs binarization processing on paper electrocardiograms to obtain binarized paper electrocardiograms;

[0039] Lead segmentation module: Extracts each lead image region from a binary paper electrocardiogram to achieve lead location of the electrocardiogram waveform;

[0040] Digital signal conversion module: converts the ECG waveform information of the lead image region into a digital reconstructed ECG signal;

[0041] QRS wave reconstruction module: Based on the digitized reconstructed ECG signal and lead position, dynamically generate adaptive slopes on both sides of the QRS wave, and reconstruct the QRS wave according to the adaptive slopes of the QRS wave.

[0042] The beneficial effects of this invention are as follows:

[0043] The slope adaptive correction method employed in this invention adaptively generates a dynamic slope specific to each heartbeat based on the waveform of the QRS wave near the baseline of a single heartbeat. Then, it regenerates the peak of the QRS wave based on this dynamic slope, thereby eliminating the influence of lead adhesion on the peak portion of the QRS wave and dynamically adjusting the detailed information of the digitized QRS wave. Compared to statically formulating a general correction strategy, the method provided by this invention effectively removes the influence of lead adhesion while specifically utilizing the characteristic differences of each heartbeat signal to reconstruct the QRS waveform. This makes the regeneration of each peak more closely resemble reality, resulting in richer details in the ECG waveform during digitization and ensuring that the digitized data accurately reflects the original information of the paper ECG. Attached Figure Description

[0044] Figure 1 A flowchart of the paper electrocardiogram digitization method based on slope adaptive correction in this embodiment of the invention;

[0045] Figure 2 A schematic diagram of the lead positioning and digitization process of paper electrocardiogram waveforms in an embodiment of the present invention;

[0046] Figure 3 A schematic diagram of the adaptive slope dynamic generation and reconstruction process of the digitized QRS wave in this embodiment of the invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0048] like Figure 1-3 As shown, a paper electrocardiogram digitization method based on slope adaptive correction includes the following steps:

[0049] S1. Obtain a paper electrocardiogram (ECG) and perform binarization processing on the paper ECG, specifically including:

[0050] S11. Obtain a standard paper electrocardiogram with length X and width Y;

[0051] S12. Obtain the acquisition rate of standard paper electrocardiograms. In a paper electrocardiogram, each large grid corresponds to a time of S seconds and a voltage value of T millivolts. A standard paper electrocardiogram has 0.2 seconds and 0.5 millivolts.

[0052] S13. Remove the background grid of the standard paper electrocardiogram using the background grid removal method to obtain a binarized paper electrocardiogram.

[0053] S2. Determine the location of each lead in the electrocardiogram and complete lead segmentation, specifically including:

[0054] S21. Horizontally process the binarized paper electrocardiogram;

[0055] S22. Summate the row and column pixels of the horizontally binarized paper electrocardiogram to obtain a list of row and column pixels. and column pixels and list Because the waveform of a binarized electrocardiogram is represented as a series of continuous black or white pixels in the image, therefore Each value in the list, i.e., the sum of pixels in each row, reflects the intensity or density of the ECG waveform in that row. Its peak position can be considered the baseline position of a certain lead. Similarly, Each value in the list, i.e. the sum of pixels in each column, can reflect the intensity or density of the ECG waveform in that column, and its peak position can be considered as the peak position of the QRS wave.

[0056] S23, Combining row pixels and list Column pixels and lists The baseline position and QRS peak position of each lead are obtained by the peak detection algorithm, and the position coordinates of each lead are located based on the baseline position and QRS peak position of each lead;

[0057] S24. Based on the position coordinates of each lead, extract the image region and pixel-time-voltage conversion grid region of the lead from the horizontally binarized paper electrocardiogram. The image region of lead I is... The image region of lead II is The image region of lead III is The aVL lead image region is The aVR lead image area is The aVF lead image region is The image area of ​​lead V1 is The image area of ​​lead V2 is The image area of ​​lead V3 is The image area of ​​lead V4 is The image area of ​​lead V5 is The image area of ​​lead V6 is .

[0058] S3. Extract the digital information from each lead and convert the binary electrocardiogram of each lead into a digitally reconstructed electrocardiogram signal, specifically including:

[0059] S31. Image of the lead region with length E and width F. Coordinate representation, with length as the x-axis and width as the y-axis, and the bottom left corner set as the origin of the coordinate axes. ;

[0060] S32. Starting from left to right, calculate the mean of all elements on the vertical axis for each horizontal axis position to obtain the lead area image. Central electrical waveform position space mapping array ;

[0061] S33. Map the ECG waveform location space array. Elements with a value of 0 are discarded, and the digitized signal representing the pixel position is obtained. ;

[0062] S34. Based on the large-grid time corresponding to the standard paper electrocardiogram. - Voltage standard: each large division corresponds to a time of S seconds and a voltage value of T millivolts. This is used to calculate the time-voltage value for each pixel. The formula for calculating the time for each pixel is: The formula for calculating the time corresponding to each pixel is: millivolt;

[0063] S35. The calculation formula for the time corresponding to each pixel is as follows: millivolts can digitize signals Converted to voltage signal The formula for calculating the time corresponding to each pixel is as follows: The digital signal can be calculated in seconds. Recording time ;

[0064] S36. Based on the recorded time and sampling rate voltage signal Length resampling to Obtain digitally reconstructed electrocardiogram signals ;

[0065] S37. Repeat steps S31 to S36 to obtain the reconstructed ECG signal in lead I. Reconstructed ECG signal in lead II Reconstructed ECG signal in lead III aVL lead reconstructed ECG signal aVR lead reconstruction of ECG signal Reconstructing ECG signals in lead aVF Reconstructing ECG signals in lead V1 Reconstructing ECG signals in lead V2 Reconstructing ECG signals in lead V3 Reconstructing ECG signals in lead V4 Reconstructing ECG signals in lead V5 Reconstructing ECG signals in lead V6 .

[0066] S4. Dynamically generate the left and right adaptive slopes of the QRS wave, specifically including:

[0067] S41. Determine the baseline R_ECG-baseline of the lead based on its position;

[0068] S42. Select a QRS wave from the reconstructed ECG signal;

[0069] S43. Based on the baseline R_ECG-baseline position and the peak position of the QRS wave, remove the portion of the QRS wave affected by lead adhesion, as follows:

[0070] S431. Determine whether the wavelet containing the QRS peak is a forward wave or a reverse wave.

[0071] S432. If it is a positive wave: Calculate the distance L between the baseline R_ECG-baseline and the upper edge of the digitally reconstructed ECG signal in this lead; shift the baseline R_ECG-baseline up by L / 10 to form R'_ECG-baseline; remove the signal of the sub-wave containing the QRS peak located above the R'_ECG-baseline.

[0072] S433. If it is a reverse wave: Calculate the distance L' from the baseline R_ECG-baseline to the lower edge of the digitally reconstructed ECG signal in this lead; shift the baseline R_ECG-baseline down by L' / 10 to form R'_ECG-baseline; remove the signal of the sub-wave containing the QRS peak located below the R'_ECG-baseline.

[0073] S44. Perform a leftward search from the peak position of the QRS wave to obtain two dynamic points on the left side. Generate an adaptive slope for the left side of the QRS wave based on the two dynamic points on the left side, as follows:

[0074] S441. Starting from the peak position of the QRS wave, move the horizontal axis to the left until the QRS wave curve intersects with R'_ECG-baseline for the first time. Take this intersection point as the first left dynamic point of the QRS wave.

[0075] S442. Starting from the peak position of the QRS wave, move the horizontal axis to the left until the QRS wave curve produces the first extreme point. Use this extreme point as the second left dynamic point of the QRS wave.

[0076] S443. Based on the first left dynamic point and the second left dynamic point, generate the left adaptive slope of the QRS wave;

[0077] S45. Perform a right-side search from the peak position of the QRS wave to obtain two dynamic points on the right side. Generate an adaptive right-side slope for the QRS wave based on the two dynamic points on the right side, as follows:

[0078] S451. Starting from the peak position of the QRS wave, move the horizontal axis to the right until the QRS wave curve intersects with R'_ECG-baseline for the first time. Take this intersection point as the first right dynamic point of the QRS wave.

[0079] S452. Starting from the peak position of the QRS wave, move the horizontal axis to the right until the QRS wave curve produces the first extreme point. Use this extreme point as the second right-side dynamic point of the QRS wave.

[0080] S453. Based on the first right-side dynamic point and the second right-side dynamic point, generate the right-side adaptive slope of the QRS wave;

[0081] S46. Repeat steps S42-S45 until all QRS waves generate left-side adaptive slope and right-side adaptive slope.

[0082] S5. Reconstruct the QRS wave based on the adaptive slope of the QRS wave, as follows:

[0083] S51. Based on the direction of the adaptive slope of the QRS wave, QRS completion is achieved by extending until the two slope extension lines on the left and right sides of each QRS wave intersect.

[0084] S52. Reconstruct QRS waveform based on two intersecting straight lines;

[0085] S53. Repeat S4 and S5 until the QRS wave waveform reconstruction is completed in all leads.

[0086] The present invention also provides a paper electrocardiogram digitization system based on slope adaptive correction, for implementing the paper electrocardiogram digitization method based on slope adaptive correction as described in the first aspect of the present invention, the system comprising:

[0087] Data acquisition module: Acquires standard paper electrocardiograms and paper electrocardiograms to be processed, and extracts paper electrocardiogram parameters;

[0088] Binarization module: Performs binarization processing on paper electrocardiograms to obtain binarized paper electrocardiograms;

[0089] Lead segmentation module: Extracts each lead image region from a binary paper electrocardiogram to achieve lead location of the electrocardiogram waveform;

[0090] Digital signal conversion module: converts the ECG waveform information of the lead image region into a digital reconstructed ECG signal;

[0091] QRS wave reconstruction module: Based on the digitized reconstructed ECG signal and lead position, dynamically generate adaptive slopes on both sides of the QRS wave, and reconstruct the QRS wave according to the adaptive slopes of the QRS wave.

[0092] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this hardware-software interchangeability, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

Claims

1. A method for digitizing paper electrocardiograms based on slope adaptive correction, comprising the following steps: Obtain a paper electrocardiogram (ECG) and perform binarization on it; Determine the location of each lead and the position of the QRS wave peak in the electrocardiogram, and complete the lead segmentation; The digital information of each lead is extracted, and the binary electrocardiogram of each lead is converted into a digitally reconstructed electrocardiogram signal. The digitally reconstructed electrocardiogram signal of each lead contains multiple heartbeat cycles, and each heartbeat cycle contains a QRS wave. The method is characterized in that, after converting the binary electrocardiogram of each lead into a digitally reconstructed electrocardiogram signal, the method further includes: The dynamic generation of the left and right adaptive slopes of the QRS wave includes the following steps: S41: Determine the baseline R_ECG-baseline of the lead based on its position; S42: Select a QRS wave from the digitally reconstructed ECG signal; S43: Based on the position of the baseline R_ECG-baseline and the peak position of the QRS wave, remove the portion of the QRS wave affected by lead adhesion; S44: Perform a left search from the peak position of the QRS wave to obtain two dynamic points on the left side, and generate an adaptive slope of the left side of the QRS wave based on the two dynamic points on the left side. S45: Perform a right search from the peak position of the QRS wave to obtain two dynamic points on the right side, and generate an adaptive slope on the right side of the QRS wave based on the two dynamic points on the right side. S46: Repeat steps S42-S45 until all QRS waves generate left-side adaptive slope and right-side adaptive slope; QRS wave reconstruction is performed based on the left and right adaptive slopes of the QRS wave, specifically including the following steps: S51: Extend a straight line based on the left and right adaptive slopes of each QRS wave until the left and right extensions of each QRS wave intersect. S52: QRS wave reconstruction based on the two intersecting extensions.

2. The paper electrocardiogram digitization method based on slope adaptive correction according to claim 1, characterized in that... The step of removing the portion of the QRS wave affected by lead adhesion, based on the baseline position R_ECG-baseline and the peak position of the QRS wave, specifically includes: Determine whether the wavelet containing the QRS peak is a forward or reverse wave; If it is a positive wave, the baseline R_ECG-baseline is moved up by a preset distance to form R'_ECG-baseline, and the signal of the sub-wavelength where the QRS peak is located above the R'_ECG-baseline is removed; If it is a reverse wave, the baseline R_ECG-baseline is lowered by a preset distance to form R'_ECG-baseline, and the signal of the sub-wavelength where the QRS peak is located below the R'_ECG-baseline is removed.

3. The paper electrocardiogram digitization method based on slope adaptive correction according to claim 2, characterized in that... The preset upward shift distance is calculated as L, which is the distance L between the baseline R_ECG-baseline and the upper edge of the digitally reconstructed ECG signal map of this lead. The baseline R_ECG-baseline is shifted upward by L / 10. The preset distance for the descent is calculated as the distance L' between the baseline R_ECG-baseline and the lower edge of the digitally reconstructed ECG signal map of this lead, and the baseline R_ECG-baseline is shifted down by L' / 10.

4. The paper electrocardiogram digitization method based on slope adaptive correction according to claim 2, characterized in that... The step of performing a leftward search from the peak position of the QRS wave to obtain two dynamic points on the left, and generating a leftward adaptive slope of the QRS wave based on the two dynamic points on the left, specifically includes: Starting from the peak position of the QRS wave, the horizontal axis is moved to the left until the QRS wave curve intersects with the R'_ECG-baseline for the first time. This intersection point is taken as the first dynamic point on the left side of the QRS wave. Starting from the peak position of the QRS wave, the horizontal axis is moved to the left until the QRS wave curve produces the first extreme point. This extreme point is taken as the second left dynamic point of the QRS wave. Based on the first and second left-side dynamic points, the left-side adaptive slope of the QRS wave is generated.

5. The paper electrocardiogram digitization method based on slope adaptive correction according to claim 2, characterized in that... The step of performing a right-side search from the peak position of the QRS wave to obtain two dynamic points on the right side, and generating a right-side adaptive slope of the QRS wave based on the two dynamic points on the right side, specifically includes: Starting from the peak position of the QRS wave, the horizontal axis is moved to the right until the QRS wave curve intersects with the R'_ECG-baseline for the first time. This intersection point is taken as the first right dynamic point of the QRS wave. Starting from the peak position of the QRS wave, the horizontal axis is moved to the right until the QRS wave curve produces the first extreme point. This extreme point is taken as the second right-side dynamic point of the QRS wave. Based on the first and second right-side dynamic points, the right-side adaptive slope of the QRS wave is generated.

6. The paper electrocardiogram digitization method based on slope adaptive correction according to claim 1, characterized in that, The process of determining the position of each lead and the QRS peak position in the electrocardiogram to complete lead segmentation specifically includes: Horizontalization processing was performed on the binarized paper electrocardiogram; Summing the row and column pixels of a horizontally binarized paper electrocardiogram yields a list of row and column pixel sums. With column pixels and list ; Combining row pixels and list Column pixels and lists The baseline position and QRS peak position of each lead are obtained by the peak detection algorithm, and the position coordinates of each lead are located based on the baseline position and QRS peak position of each lead.

7. A paper electrocardiogram digitization system based on slope adaptive correction, used to implement the paper electrocardiogram digitization method based on slope adaptive correction as described in any one of claims 1 to 6, characterized in that, The system includes: Data acquisition module: Acquires standard paper electrocardiograms and paper electrocardiograms to be processed, and extracts paper electrocardiogram parameters; Binarization module: Performs binarization processing on paper electrocardiograms to obtain binarized paper electrocardiograms; Lead segmentation module: Extracts each lead image region from a binary paper electrocardiogram to achieve lead location of the electrocardiogram waveform; Digital signal conversion module: converts the ECG waveform information of the lead image region into a digital reconstructed ECG signal; QRS wave reconstruction module: Based on the digitized reconstructed ECG signal and lead position, dynamically generate adaptive slopes on both sides of the QRS wave, and reconstruct the QRS wave according to the adaptive slopes of the QRS wave.

Images