CN117351221A - Method for extracting key points of paper simulation verification electrocardiogram - Google Patents
Method for extracting key points of paper simulation verification electrocardiogram Download PDFInfo
- Publication number
- CN117351221A CN117351221A CN202311271505.7A CN202311271505A CN117351221A CN 117351221 A CN117351221 A CN 117351221A CN 202311271505 A CN202311271505 A CN 202311271505A CN 117351221 A CN117351221 A CN 117351221A
- Authority
- CN
- China
- Prior art keywords
- key
- test
- sift
- electrocardiogram
- points
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 73
- 238000012795 verification Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000012360 testing method Methods 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000000605 extraction Methods 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000001914 filtration Methods 0.000 claims description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 238000013135 deep learning Methods 0.000 claims description 6
- 238000002565 electrocardiography Methods 0.000 claims 1
- 230000033764 rhythmic process Effects 0.000 description 7
- 238000003672 processing method Methods 0.000 description 4
- 230000000747 cardiac effect Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/353—Detecting P-waves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Cardiology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a paper simulation verification electrocardiogram key point extraction method, which comprises the steps of carrying out digital image processing on a paper simulation verification electrocardiogram data set, and carrying out corner detection on the paper simulation verification electrocardiogram data set after the digital image processing; selecting a plurality of paper simulation verification electrocardiograph key points, taking the key points as labels of a deep Labcut paper simulation verification electrocardiograph training data set, and training; the method comprises the steps of testing a paper simulation test electrocardiogram data set by adopting deep Labcut, and predicting the positions of a plurality of selected paper simulation test electrocardiogram key points; searching SIFT corner points near the key points of the electrocardiogram predicted by the deep Labcut, and selecting the key points of the paper simulation verification electrocardiogram through the distances between the predicted key points and the SIFT corner points.
Description
Technical Field
The invention relates to the technical field of metering, in particular to a paper simulation electrocardiogram key point extraction method based on deep learning and high-confidence corner searching.
Background
The paper simulation test electrocardiogram is an important carrier for the verification of the digital electrocardiograph, and according to the verification rules, the distance between key points in the paper simulation test electrocardiogram is measured to obtain verification parameters, so as to judge whether the digital electrocardiograph meets the specified requirements. The standard matching equipment for electrocardiograph measurement required by the JJG 1041-2008 digital electrocardiograph calibration procedure comprises a signal generator, an analog impedance circuit, a divider and a graduated scale, and a calibrating staff measures waveform pattern characteristic parameters by using the divider and the graduated scale for calibration. The verification data to be detected in the actual operation are more and complex, and the problems of easy error, large workload, long training time of manual operation, higher cost, low efficiency and the like exist, so that how to realize the automatic and intelligent extraction of key points in the paper simulation verification electrocardiogram becomes a problem to be solved urgently.
The invention CN115690435A, CN115272112A and the like are used for digitizing the paper centroid by adopting the traditional digital image processing method, and extracting characteristic parameters of the paper centroid by adopting the traditional digital image processing method; patent CN116012568A, CN115512130a adopts a deep learning method to extract characteristic parameters of paper centroid map.
The specific patent reference documents are as follows:
1) The invention discloses a system for acquiring cardiac rhythm information by photographing an electrocardiogram, which is disclosed in patent number CN116012568A, wherein an cardiac region extraction unit is responsible for acquiring a 12-lead paper electrocardiogram report photographed by a mobile phone camera, removing irrelevant additional information and extracting an cardiac region based on a target detection neural network; the image processing unit adopts algorithms such as image denoising, inclination correction, shadow removal, data enhancement and the like to preprocess the image; the electrocardiographic rhythm information acquisition and abnormality classification unit acquires electrocardiographic rhythm information based on a convolutional neural network and classifies rhythm abnormality of the electrocardiographic rhythm information; and the android mobile phone application program functional unit is used for feeding the classification result of the electrocardiographic rhythm information of the user back to the result feedback module of the visiting user. The invention can acquire and identify and classify the electrocardiographic rhythm information by taking a paper electrocardiogram report photo, is a manageable, portable and low-cost tool, and is helpful for improving electrocardiographic interpretation in doctor aided diagnosis. The key point and characteristic parameter extraction method of the invention is different from the above patent, and adopts deep Labcut deep learning and searching high confidence SIFT angular point detection method to extract key points.
2) "a paper centroid electrogram digitizing method based on classical statistical method", patent number CN115690435a, the application discloses a paper centroid electrogram digitizing method based on classical statistical method, the steps include: acquiring an image to be processed, carrying out graying treatment on the image to be processed to obtain a gray image, and obtaining an electrocardiogram waveform based on the gray image; based on the gray level image, a kernel density estimation graph is made, partial gray level data is removed according to the kernel density estimation graph, and a first image is obtained; performing first FoF clustering on the first image to obtain a second image; removing part of gray data according to the kernel density estimation graph for a plurality of maximum noise-containing classes in the second image to obtain a third image; performing second FoF clustering on the third image, and further screening according to the dispersion degree to obtain a fourth image; removing grid noise intersecting the fourth image and the electrocardiogram waveform to obtain a fifth image; and carrying out FoF clustering on the fifth image for the third time to obtain an image after noise removal. The digital display device is simple in structure, fast and effective, and extremely high in digital efficiency. The method for extracting the key points and the characteristic parameters is different from the above patent, and the method is focused on digitizing the paper centroid map without extracting the position information of the characteristic points.
3) The invention relates to a paper centroid electrogram waveform parameter automatic measurement method, in particular to a paper centroid electrogram waveform parameter automatic measurement method, which belongs to the technical field of digital image processing and comprises electrocardiograph image acquisition, electrocardiograph image inclination detection and correction, electrocardiograph removal grid characteristics, electrocardiograph waveform extraction and electrocardiograph data calculation. The key point and characteristic parameter extraction method of the invention is different from the above patent, the invention combines the deep Labcut algorithm with the SIFT algorithm, and adopts a method of searching adjacent areas to determine the key point of the paper centroid electrogram, which is different from the AI algorithm matched with the industrial camera in the invention.
4) "a paper centroid electrogram digitizing method and device", patent number CN115272112a, the invention relates to a paper centroid electrogram digitizing method and device, comprising: preprocessing an electronic image of a paper electrocardiogram containing twelve lead areas to obtain an electrocardiogram; separating individual lead regions of the electrocardiographic image; based on an eight-neighborhood sparse outlier removal algorithm and a prestored refinement algorithm, refining the electrocardiographic images of each lead region to obtain electrocardiographic waveform curves of each lead region; and connecting the electrocardiographic wave curves of the lead areas in a lead dimension mode to obtain a digitalized result of the paper centroid map. The invention converts the paper centroid electrogram continuously generated into the electronic electrocardiogram in a way of digitizing the effective information of the central electric signal of the paper centroid electrogram, solves the problem of insufficient body quantity of the electronic electrocardiogram, and promotes the development of the automatic electrocardiograph analysis field. The key point and characteristic parameter extraction method is different from the patent, and the deep LabCut deep learning algorithm is adopted to combine the digital image processing method and the SIFT corner extraction method, so that the method is different from the traditional image processing methods such as preprocessing, connected domain, thinning and the like, and has stronger generalization capability.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a paper simulation verification electrocardiogram key point extraction method based on deep learning and high-confidence-degree corner searching.
The aim of the invention is achieved by the following technical scheme:
a paper simulation verification electrocardiogram key point extraction method comprises the following steps:
step A, carrying out digital image processing on a paper simulation test electrocardiographic data set, and carrying out corner detection on the paper simulation test electrocardiographic data set subjected to the digital image processing;
b, selecting a plurality of paper simulation verification electrocardiograph key points, taking the key points as labels of a DeepLabcut paper simulation verification electrocardiograph training data set, and training;
step C, adopting deep Labcut to test the paper simulation test electrocardiograph data set, and predicting the positions of the selected multiple paper simulation test electrocardiograph key points;
and D, searching SIFT corner points near the key points of the electrocardiogram predicted by the deep Labcut, and selecting the key points of the paper simulation verification electrocardiogram through the distances between the predicted key points and the SIFT corner points.
One or more embodiments of the present invention may have the following advantages over the prior art:
by introducing high-confidence SIFT angular point search for deep Labcut key point prediction, the deep Labcut key point prediction deviation caused by inaccurate manual labeling of a training set is reduced, the key point prediction precision and robustness are improved, for key points with unobvious SIFT angular point characteristics, the deep Labcut key point is directly adopted to be beneficial to retaining original characteristics of the key points, and the method meets the pre-arranged requirements of intelligent verification of JJG 1041-2008 digital electrocardiograph verification rules.
Drawings
FIG. 1 is a flow chart of a method for extracting key points of a paper simulation verification electrocardiogram;
FIG. 2 is a schematic diagram of 17 key points of a paper simulation verification electrocardiogram key point extraction method;
fig. 3 is a schematic diagram of deep labcut prediction key points and high confidence SIFT corner points in the paper simulation verification electrocardiogram key point extraction method.
Wherein 31 is a deep Labcut prediction key point; 32 is the high confidence SIFT corner.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following examples and the accompanying drawings.
As shown in FIG. 1, the method for extracting the key points of the electrocardiograph for paper simulation verification comprises the following steps:
step 10, carrying out digital image processing on the paper simulation test electrocardiographic data set, and carrying out corner detection on the paper simulation test electrocardiographic data set subjected to the digital image processing;
step 20, selecting a plurality of paper simulation verification electrocardiograph key points, and taking the key points as labels of a deep Labcut paper simulation verification electrocardiograph training data set and training;
step 30, testing the paper simulation test electrocardiograph data set by adopting deep Labcut, and predicting the positions of the selected multiple paper simulation test electrocardiograph key points;
and step 40, searching SIFT corner points near the key points of the electrocardiogram predicted by the deep Labcut, and selecting the key points of the paper simulation verification electrocardiogram according to the distances between the predicted key points and the SIFT corner points. Establishing an image pixel coordinate system for carrying out SIFT angular point detection coordinate positioning, wherein the origin of coordinates O is the top left vertex of the image, the positive x-axis direction is horizontal to the right, and the positive y-axis direction is vertical to the lower; for gray scale images, binarized images, median filtered images, the pixel coordinate system is fixed.
The step 10 specifically includes performing gray level, binarization and median filtering processing on the digital image of the paper simulation test electrocardiogram data set to obtain I respectively grey Image dataset, I binary Image dataset, I median filtering Image data set to facilitate SIFT angular point feature extraction, and performing angular point detection on the paper simulation verification electrocardiogram data set processed by the digital image by using SIFT angular point detection algorithm to obtain SIFT angular point subset i SIFT 。
Paper simulation test of electrocardiographic data set as I origin ，I origin The image dataset is { F origin-1 ,F origin-2 ,F origin-3 ,……,F origin-N (wherein N is I) origin The number of images; pair I origin Gray processing to obtain I grey ,I grey The image dataset is { F grey-1 ,F grey-2 ,F grey-3 ,……,F grey-N (wherein N is I) grey The number of images; pair I grey Binarizing to obtain I binary Image dataset, I binary The image dataset is { F binary-1 ,F binary-2 ,F binary-3 ,……,F binary-N (wherein N is I) binary The number of images; pair I binary Median filtering to obtain I median filtering Image dataset, I median filtering The image dataset is { F median filtering-1 ,F median filtering-2 ,F median filtering-3 ,……,F median filtering-N (wherein N is I) median filtering Image number of the image. The object detected by the SIFT corner detection algorithm is I median filtering (I grey Image dataset and I binary The image data set does not carry out angular point detection, graying and binarization are only used for preparing median filtering, and the SIFT angular point detection is only carried out by I median filtering ) Processed by SIFT angular point detection algorithm, I SIFT The image dataset is { F SIFT-1 ,F SIFT-2 ,F SIFT-3 ,……,F SIFT-N (wherein N is I) SIFT Image number of (F) for each F SIFT-i I epsilon N, contains SIFT corner images and coordinates, where { P }, where SIFT-1 ,P SIFT-2 ,P SIFT-3 ,……,P SIFT-n Form SIFT corner subset i SIFT . Pair I using SIFT corner detection algorithm median filtering The corner detection is divided into construction of a scale space, detection of extreme points, positioning of key points, direction distribution of the key points and generation of key point descriptors.
In the step 20, 17 key points (17 key points are taken as an example in this embodiment, but not limited to this, and are used as labels of the deep labcut paper simulation verification electrocardiogram training data set) and are trained (as shown in fig. 2); the method specifically comprises the following steps:
at I origin Image dataset { F origin-1 ,F origin-2 ,F origin-3 ,……,F origin-N F of } origin-i Selecting 17 key points i of paper simulation test electrocardiogram key-label Obtaining coordinates of 17 key points, i key-label Comprises { P ] key-label-1 ,P key-label-2 ,P key-label-3 ,……,P key-label-17 For each I } origin F in image dataset origin-i All have a corresponding subset of keypoints i key-label . For P key-label-i Includes an x-coordinate and a y-coordinate, which can be expressed as { (x) key-label-1 ,y key-label-1 ),(x key-label-2 ,y key-label-2 ),(x key-label-3 ,y key-label-3 ),……,(x key-label-17 ,y key-label-17 )}。
In the step 30, the paper simulation test electrocardiogram data set is tested by deep Labcut to predict 17The positions of the key points can be detected { P } key-test-1 ,P key-test-2 ,P key-test-3 ,……,P key-test-17 -a }; the method specifically comprises the following steps:
will I origin And corresponding key point subset i key-label Converting into training set format, inputting into deep Labcut for training, and performing key point detection on paper simulation check electrocardiogram by deep Labcut model after training, to obtain { P } key-test-1 ,P key-test-2 ,P key-test-3 ,……,P key-test-17 For P } key-test-i Includes an x-coordinate and a y-coordinate, which can be expressed as { (x) key-test-1 ,y key-test-1 ),(x key-test-2 ,y key-test-2 ),(x key-test-3 ,y key-test-3 ),……,(x key-test-17 ,y key-test-17 )}。
As shown in fig. 3, in the step 40, SIFT corner points are searched near the 17 key points 31 predicted by deep labcut, if the distance is smaller than the threshold value, the SIFT corner points 32 with high confidence coefficient are selected as the key points of the paper simulation verification electrocardiogram; if the distance is greater than the threshold value, directly selecting the key points predicted by the deep Labcut as the key points of the paper simulation verification electrocardiogram to obtain a subset i of key points with high confidence coefficient of the paper simulation verification electrocardiogram key ，i key Is { P key-1 ,P key-2 ,P key-3 ,……,P key-17 -a }; the method specifically comprises the following steps:
obtaining 17 key feature points { (x) based on DeepLabcut key-test-1 ,y key-test-1 ),(x key-test-2 ,y key-test-2 ),(x key-test-3 ,y key-test-3 ),……,(x key-test-17 ,y key-test-17 ) In P }, at key-test-i Search for the center of a circle for the vicinity { P SIFT-1 ,P SIFT-2 ,P SIFT-3 ,……,P SIFT-n }：
Wherein P is SIFT-i Is the distance (x) key-test-i ,y key-test-i ) The nearest SIFT corner, called the high confidence corner, P SIFT-i And (x) key-test-i ,y key-test-i ) Is the distance of (2)
Obtaining a subset i of key points with high confidence coefficient of paper simulation verification of electrocardiogram key ，i key Is { P key-1 ,P key-2 ,P key-3 ,……,P key-17 }。
Although the embodiments of the present invention are described above, the embodiments are only used for facilitating understanding of the present invention, and are not intended to limit the present invention. Any person skilled in the art can make any modification and variation in form and detail without departing from the spirit and scope of the present disclosure, but the scope of the present disclosure is still subject to the scope of the appended claims.
Claims (10)
1. The method for extracting the key points of the paper simulation verification electrocardiogram is characterized by comprising the following steps of:
step A, carrying out digital image processing on a paper simulation test electrocardiographic data set, and carrying out corner detection on the paper simulation test electrocardiographic data set subjected to the digital image processing;
b, selecting a plurality of paper simulation verification electrocardiograph key points, taking the key points as labels of a DeepLabcut paper simulation verification electrocardiograph training data set, and training;
step C, adopting deep Labcut to test the paper simulation test electrocardiograph data set, and predicting the positions of the selected multiple paper simulation test electrocardiograph key points;
and D, searching SIFT corner points near the key points of the electrocardiogram predicted by the deep Labcut, and selecting the key points of the paper simulation verification electrocardiogram through the distances between the predicted key points and the SIFT corner points.
2. The method for extracting key points of paper simulation verification electrocardiography according to claim 1, wherein in the step a: the digital image processing of the paper simulation checking electrocardiogram data set comprises graying, binarization and median filtering, and the digital image processing sequentially obtains a gray image data set, a binarization image data set and a median filtering image data set through graying, binarization and median filtering;
the paper simulation test electrocardiographic data set is I origin ，I origin The image dataset is { F origin-1 ,F origin-2 ,F origin-3 ,……,F origin-N (wherein N is I) origin Image number of the image.
3. The method for extracting key points of paper simulation verification electrocardiogram according to claim 2, wherein,
examination of Electrocardiogram dataset I from paper simulation origin Gray image data set I obtained by graying grey ，I grey Is { F grey-1 ,F grey-2 ,F grey-3 ,……,F grey-N (wherein N is I) grey The number of images; from I grey Binarizing to obtain binarized image data set I binary ，I binary Is { F binary-1 ,F binary-2 ,F binary-3 ,……,F binary-N (wherein N is I) binary The number of images; pair I binary Median filtering to obtain median filtered image dataset I median filtering ,I median filtering Is { F median filtering-1 ,F median filtering-2 ,F median filtering-3 ,……,F median filtering-N (wherein N is I) median filtering Image number of the image.
4. The method for extracting key points of a paper simulation verification electrocardiogram according to claim 1, wherein the method further comprises: and establishing an image pixel coordinate system to enable SIFT angular point detection coordinates to be positioned, wherein the origin of coordinates O is the top left vertex of the image, the positive direction of the x axis is horizontal to the right, and the positive direction of the y axis is vertical to the down.
5. The method for extracting key points of paper simulation verification electrocardiogram according to claim 1, wherein in the a, a SIFT corner detection algorithm is used to perform corner detection on the paper simulation verification electrocardiogram data set processed by the digital image, wherein the SIFT corner detection algorithm detects an image data set I median filtering Processed by SIFT angular point detection algorithm, I SIFT The image dataset is { F SIFT-1 ,F SIFT-2 ,F SIFT-3 ,……,F SIFT-N (wherein N is I) SIFT Image number of (F) for each F SIFT-i Comprises SIFT corner images and coordinates, wherein { P } SIFT-1 ,P SIFT-2 ,P SIFT-3 ,……,P SIFT-n Form SIFT corner subset i SIFT ，i∈N。
6. The method for extracting key points of paper simulation verification electrocardiogram according to claim 1, wherein in the step B, in I origin Image dataset { F origin-1 ,F origin-2 ,F origin-3 ,……,F origin-N F of } origin-i Selecting 17 key points i of paper simulation test electrocardiogram key-label Obtaining coordinates of 17 key points, i key-label Comprises { P ] key-label-1 ,P key-label-2 ,P key-label-3 ,……,P key-label-17 Each I origin F in image dataset origin-i All have a corresponding subset of keypoints i key-label ；P key-label-i Includes an x-coordinate and a y-coordinate, which can be expressed as { (x) key-label-1 ,y key-label-1 ),(x key-label-2 ,y key-label-2 ),(x key-label-3 ,y key-label-3 ),……,(x key-label-17 ,y key-label-17 )}。
7. The method for extracting key points of paper simulation test electrocardiogram according to claim 1, wherein in said C, the paper simulation test electrocardiogram data set I is obtained origin And corresponding key point subset i key-label Converting into training set format, inputting into deep Labcut for training, and performing key point detection on paper simulation test electrocardiogram by using the deep Labcut model after training to obtain { P } key-test-1 ,P key-test-2 ,P key-test-3 ,……,P key-test-17 }，P key-test-i Includes an x-coordinate and a y-coordinate, which can be expressed as { (x) key-test-1 ,y key-test-1 ),(x key-test-2 ,y key-test-2 ),(x key-test-3 ,y key-test-3 ),……,(x key-test-17 ,y key-test-17 )}。
8. The method for extracting key points of a paper simulation verification electrocardiogram according to claim 1, wherein the step D specifically comprises: searching SIFT corner points near the 17 deep Labcut predicted key points, and if the distance is smaller than a threshold value, selecting the SIFT corner points with high confidence as key points for paper simulation verification of an electrocardiogram; if the distance is larger than the threshold value, directly selecting the key point predicted by deep Labcut as the key point of the paper simulation test electrocardiogram.
9. The method for extracting key points of paper simulation verification electrocardiogram according to claim 7, wherein 17 key feature points { (x) are obtained based on deep Labcut key-test-1 ,y key-test-1 ),(x key-test-2 ,y key-test-2 ),(x key-test-3 ,y key-test-3 ),……,(x key-test-17 ,y key-test-17 ) In P }, at key-test-i Search for the center of a circle for the vicinity { P SIFT-1 ,P SIFT-2 ,P SIFT-3 ,……,P SIFT-n }：
Wherein P is SIFT-i Is the distance (x) key-test-i ,y key-test-i ) The nearest SIFT corner, called the high confidence corner, P SIFT-i And (x) key-test-i ,y key-test-i ) Is R is a distance of i min ；
Obtaining a subset i of key points with high confidence coefficient of paper simulation verification of electrocardiogram key ，i key Is { P key-1 ,P key-2 ,P key-3 ,……,P key-17 }。
10. The paper simulation verification electrocardiogram key point extraction method according to any one of claims 1 to 9, wherein the paper simulation verification electrocardiogram key point extraction method is realized based on deep learning and high-confidence corner search.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311271505.7A CN117351221A (en) | 2023-09-27 | 2023-09-27 | Method for extracting key points of paper simulation verification electrocardiogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311271505.7A CN117351221A (en) | 2023-09-27 | 2023-09-27 | Method for extracting key points of paper simulation verification electrocardiogram |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117351221A true CN117351221A (en) | 2024-01-05 |
Family
ID=89370303
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311271505.7A Pending CN117351221A (en) | 2023-09-27 | 2023-09-27 | Method for extracting key points of paper simulation verification electrocardiogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117351221A (en) |
-
2023
- 2023-09-27 CN CN202311271505.7A patent/CN117351221A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20130051676A1 (en) | Microcalcification Detection Classification in Radiographic Images | |
| Pang et al. | Automatic detection and quantification of hand movements toward development of an objective assessment of tremor and bradykinesia in Parkinson's disease | |
| CN111583320B (en) | Breast cancer ultrasonic image typing method and system fusing deep convolutional network and image omics characteristics and storage medium | |
| WO2005083635A1 (en) | Method and system for detection of bone fractures | |
| WO2019238104A1 (en) | Computer apparatus and method for implementing classification detection of pulmonary nodule images | |
| CN111627555B (en) | Intelligent inspection diagnosis system based on deep learning | |
| CN113160153A (en) | Lung nodule screening method and system based on deep learning technology | |
| CN115527065A (en) | Hip joint typing method, device and storage medium | |
| CN111445386A (en) | Image correction method based on four-point detection of text content | |
| CN117333489B (en) | Film damage detection device and detection system | |
| Alshudukhi et al. | Convolutional Neural Network Architectures to Solve a Problem of Tuberculosis Classification Using X-Ray Images of the Lungs | |
| CN116725546A (en) | Digital electrocardiograph verification method based on deep learning and corner detection | |
| Wei et al. | Multi-feature fusion for ultrasound breast image classification of benign and malignant | |
| CN117351221A (en) | Method for extracting key points of paper simulation verification electrocardiogram | |
| Liang et al. | Relative saliency model over multiple images with an application to yarn surface evaluation | |
| CN113537407B (en) | Image data evaluation processing method and device based on machine learning | |
| CN114898347A (en) | Machine vision identification method for pointer instrument | |
| CN113658238A (en) | Near-infrared vein image high-precision matching method based on improved feature detection | |
| CN110349662A (en) | The outliers across image collection that result is accidentally surveyed for filtering pulmonary masses find method and system | |
| Jing-Shi et al. | Symmetry theory based classification algorithm in brain computed tomography image database | |
| Lence et al. | Automatic digitization of paper electrocardiograms–A systematic review | |
| CN116521912B (en) | Ultrasonic data storage management system and method based on artificial intelligence | |
| CN111709946A (en) | Image data analysis method and device, terminal equipment and storage medium | |
| CN110889836A (en) | Image data analysis method and device, terminal equipment and storage medium | |
| Fang et al. | A Multitarget Interested Region Extraction Method for Wrist X-Ray Images Based on Optimized AlexNet and Two-Class Combined Model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |