CN109009074A - A kind of sudden cardiac death auxiliary prior-warning device based on deep learning - Google Patents
A kind of sudden cardiac death auxiliary prior-warning device based on deep learning Download PDFInfo
- Publication number
- CN109009074A CN109009074A CN201810808839.6A CN201810808839A CN109009074A CN 109009074 A CN109009074 A CN 109009074A CN 201810808839 A CN201810808839 A CN 201810808839A CN 109009074 A CN109009074 A CN 109009074A
- Authority
- CN
- China
- Prior art keywords
- module
- deep learning
- sudden cardiac
- electrocardiogram
- warning device
- 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
Classifications
-
- 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]
-
- 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
-
- 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
- A61B5/35—Detecting specific parameters of the electrocardiograph cycle by template matching
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient ; user input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Heart & Thoracic Surgery (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Cardiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Multimedia (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The present invention relates to a kind of, and the sudden cardiac death based on deep learning assists prior-warning device, including electrocardiogram acquisition module, image pre-processing module, model judgment module and abnormal alarm module, the electrocardiogram acquisition module is connect by image pre-processing module with model judgment module, and the model judgment module is connect with abnormal alarm module；The electrocardiogram acquisition module acquires electrocardiogram in real time and is sent to image pre-processing module, the image pre-processing module takes ROI region using sliding window, and ROI region is input to model judgment module, the embedded convolutional neural networks model based on deep learning of the model judgment module judges input data, if it is determined that abnormal electrocardiogram, abnormal electrocardiogram number ycount increases by 1, when ycount quantity is 3, the model judgment module control abnormal alarm module carries out and alarm.Compared with prior art, the present invention has many advantages, such as accuracy rate and high sensitivity.
Description
Technical field
The present invention relates to a kind of medical apparatus and instruments, assist more particularly, to a kind of sudden cardiac death based on deep learning
Prior-warning device.
Background technique
Sudden cardiac death (Sudden Cardiac Death, SCD) is a kind of disease for seriously endangering human health, the world
Annual therefore disease sudden death patient is there are about 12,000,000, and wherein SCD accounts for 75%.The death caused by the heart disease of developing country
In event, ratio shared by sudden cardiac death is about 50%.It is more up in the total number of persons that China dies of sudden cardiac death every year
More than 50 ten thousand people, average minute clock just have 3 people dead because of cardiac reasons morbidity.The status of China's sudden cardiac death is severe and does not allow to find pleasure in
See, this is not only occupied first of the world because of annual sudden death total number of persons, but also cause many factors such as social population to die suddenly aging,
Various incidence of cardiovascular disease are not effectively controlled and have exacerbation trend, and sudden cardiac death is to China's public health and economic development
With very big harmfulness.
The human body electrocardio figure (ECG) for being applied to clinical heart disease's diagnosis for the first time by Dutch physiologist Einthoven is the heart
Dirty electrical activity body surface general performance, wherein the physiology and pathological information that contain can reflect cardiac rhythm and electrical conduction feelings
Condition, the electrocardiogram by monitoring patient can be found that work as the untimely anomalous ecg performance of heart occur.It and blood glucose, blood pressure one
Most important three indexs for being listed in cardiovascular and cerebrovascular disease monitoring and diagnosis are acted, are clinically widely used.
The research history of convolutional neural networks can trace back to the last century 60's, it uses three kinds of basic conceptions and comes
Help improves machine learning system: local receptor field (local receptive fields), shared weight (shared
Weights) and pond dimensionality reduction (pooling), due to above feature, convolutional neural networks convert translation, scaling etc.
It can keep height invariance.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind to be based on deep learning
Sudden cardiac death assist prior-warning device.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of sudden cardiac death auxiliary prior-warning device based on deep learning, including electrocardiogram acquisition module, image preprocessing
Module, model judgment module and abnormal alarm module, the electrocardiogram acquisition module are sentenced by image pre-processing module and model
Disconnected module connection, the model judgment module are connect with abnormal alarm module；
The electrocardiogram acquisition module acquires electrocardiogram in real time and is sent to image pre-processing module, and the image is located in advance
It manages module and takes ROI region using sliding window, and ROI region is input to model judgment module, the model judgment module
The embedded convolutional neural networks model based on deep learning judges input data, if it is decided that it is abnormal electrocardiogram, it is different
Chang Xin electricity number ycount increases by 1, when ycount quantity is 3, the model judgment module control abnormal alarm module into
Capable and alarm.
Preferably, the convolutional neural networks model construction based on deep learning are as follows: set using deep learning technology
The convolutional neural networks framework for being applied to sudden cardiac death auxiliary early warning is counted, and in normal sinus rhythm database
Processing conclusion is carried out to ecg wave form on the basis of nsrdb and sudden cardiac death database sddb, is built by data enhancing
Trained test set is found, to carry out training convolutional neural networks model, using trained convolutional neural networks model to test number
According to being tested.
Preferably, the convolutional neural networks mode input image is 32 × 32.
Preferably, the convolutional neural networks model shares 4 groups of network blocks composition, first 3 groups use 5 × 5 convolution kernel,
4th group use 3 × 3 convolution kernel.
Preferably, the pond mode of the convolutional neural networks model is that " maximum dimensionality reduction-maximum dimensionality reduction-is averaged dimensionality reduction-
The mode of average dimensionality reduction ", sets maximum dimensionality reduction mode for two network blocks of beginning, as the 3rd, 4 network block output is special
The size of sign mapping reduces, and ecg characteristics tend to sparsity expression, the pixel value of surrounding is replaced using average value, effective
Guarantee that main effective information is not lost in ecg characteristics expression while controlling over-fitting.
Preferably, the activation primitive of the convolutional neural networks model chooses ReLU function.
Preferably, data enhancing specifically: electrocardiogram (ECG) data is enhanced using scale transformation.
Preferably, electrocardio ROI is obtained using translation gliding window strategy, the acquisition time period is 6 seconds, and in electrocardio ROI
In screened.
Preferably, the screening specifically: in the sample that the acquired period is 6 seconds, the abnormal data time is big
In the data equal to 4.8 seconds as abnormal electrocardiogram sample, the normal electrocardio time is not less than 4.8 seconds in acquired sample
Data are as normal electrocardio sample.
Compared with prior art, the invention has the following advantages that
1) accuracy rate and high sensitivity, accuracy rate have reached 91.75%, sensitivity 94%, and two indices are superior to pair
Traditional machine learning method than in experiment.
2) practical, the single picture testing time is 44ms under GTX1050 environment, and it is pre- to meet sudden cardiac death auxiliary
Alert requirement of real-time, it is practical.
3) integration is good, it is only necessary to the model proposed in the present invention is integrated in electrocardiogram acquisition equipment, after succeeding after debugging
Electrocardiogram acquisition equipment can have electrocardio monitoring function, and installing alarm module additional can work.
Detailed description of the invention
Fig. 1 is work flow diagram of the invention；
Fig. 2 is the principle of the present invention block diagram；
Fig. 3 is convolutional neural networks structure chart；
Fig. 4 is the present invention and other method index comparison diagrams.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiment is a part of the embodiments of the present invention, rather than whole embodiments.Based on this hair
Embodiment in bright, those of ordinary skill in the art's every other reality obtained without making creative work
Example is applied, all should belong to the scope of protection of the invention.
The present invention designs a kind of sudden cardiac death auxiliary prior-warning device based on deep learning on the basis of electrocardiogram,
The device devises the convolutional neural networks framework for being applied to sudden cardiac death auxiliary early warning using deep learning technology, and
Ecg wave form is carried out on the basis of normal sinus rhythm database (nsrdb) and sudden cardiac death database (sddb)
Processing conclude, by data enhancing establish trained test set, using convolutional neural networks training model to test data into
Row test, achieves ideal auxiliary early warning effect.Ecg characteristics can be automatically extracted using the present apparatus and are judged different
Normal electrocardiosignal, the shortcomings that overcoming professional's lazy weight and experience and knowledge to be short of and limitation, increase substantially analysis
Efficiency promotes to be widely used.It helps many patients to predict with preferable practicability and actual effect, prevent, manage sudden cardiac disease
Disease and the harm for reducing sudden cardiac death reduce the death rate.Simultaneously using can effectively mitigate seeing for medical staff in institute
Pressure is protected, nurse efficiency is improved, prevents the generation of fortuitous event.
The lossless auxiliary prior-warning device groundwork of sudden cardiac death of the present invention includes: to assist early warning problem for SCD, simultaneous
While caring for accuracy and real-time, a kind of convolutional neural networks model for sudden cardiac death auxiliary early warning is proposed.Knot
It closes medicine priori knowledge and establishes the test training set of cardiac sudden death auxiliary early warning, and utilize data enhancement methods expanding data
Collection, for verifying the accuracy of device.Auxiliary early warning effect based on this device compares the conventional machines such as hog+svm, BP
The auxiliary early warning effect of habit, it was demonstrated that the validity of device.
The convolutional neural networks mode input image is 32 × 32, and network training speed can be improved in small image input
Degree meets monitoring requirement of real-time.Network shares 4 groups of network blocks composition, first 3 groups use 5 × 5 convolution kernel, main cause is
In order to obtain the biggish receptive field of range, the linear feature variation of electrocardiogram is preferably extracted.4th group use 3 × 3 convolution
Core, main cause are to adopt by the data transformation of preceding 3 network blocks and feature extraction, the Feature Mapping of output having a size of 4 × 4
It is expressed with the sparsity that the convolution kernel of 3 × 3 sizes can continue to obtain in the case where not introducing 0 pixel of redundancy ecg characteristics.
Pond mode is the mode of " maximum dimensionality reduction-maximum dimensionality reduction-be averaged dimensionality reduction-be averaged dimensionality reduction ", compared to maximum dimensionality reduction mode, is averaged
Dimensionality reduction is computationally intensive, noise is readily incorporated, therefore set maximum dimensionality reduction mode for two network blocks of beginning, with the 3rd, 4
The size of a network block output Feature Mapping reduces, and ecg characteristics tend to sparsity expression, replaces surrounding using average value
Pixel value, effectively control over-fitting while guarantee ecg characteristics expression in main effective information do not lose.Activate letter
Number chooses ReLU function, and relative to activation primitives such as Sigmoid, TanH, ReLU function has relatively broad excited boundary, protects
The sparse activity of network has been demonstrate,proved, while having improved the convergence rate of network.
The test training set of the cardiac sudden death auxiliary early warning is based primarily upon the normal sinus rhythm in PhysioBank
The purpose of two ecg databases of database (nsrdb) and sudden cardiac death database (sddb), data enhancing is effective
Avoid data volume it is less caused by network overfitting problem generation, electrocardiogram (ECG) data is increased using scale transformation herein
By force, the reason is that scale transformation zooms in or out image according to a certain percentage, carrying out data enhancing with this method will not change
The sequential relationship for becoming time shaft and amplitude axis, more meets electrocardio feature of image.
The present invention and other traditional machine learning method comparisons mainly include pair with two methods of hog+svm and BP
Than Hog feature, that is, histograms of oriented gradients, it is constituted by calculating with the gradient orientation histogram of statistical picture regional area
Feature, svm are that Vpanik et al. is calculated based on the theoretical machine learning with structural risk minimization of VC dimension in statistical learning
Method；BP neural network is a kind of Multi-layered Feedforward Networks by Back Propagation Algorithm training, is current most widely used nerve
One of network model.
As shown in Figure 1, acquiring electrocardiogram in real time using electrocardiogram acquisition equipment in the present embodiment, and taken using sliding window
ROI region, time cycle are 6 seconds, and ROI is input in electrocardiogram model and is judged, if it is decided that it is abnormal electrocardiogram, it is abnormal
Electrocardio number ycount increase by 1, when ycount quantity be 3 when and alarm.
As shown in Figure 2,3, a use is designed while taking into account accuracy and real-time for SCD auxiliary early warning problem
In the convolutional neural networks model of sudden cardiac death auxiliary early warning, using data enhancement methods EDS extended data set and verification method
The indexs such as accuracy rate, sensitivity.
As shown in figure 4, the auxiliary prior-warning device designed in the present invention is accurate compared to hog+svm and BP neural network
Dominant in rate and sensitivity, device assists early warning effect preferable.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can readily occur in various equivalent modifications or replace
It changes, these modifications or substitutions should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with right
It is required that protection scope subject to.
Claims (9)
1. a kind of sudden cardiac death based on deep learning assists prior-warning device, which is characterized in that including electrocardiogram acquisition module, figure
As preprocessing module, model judgment module and abnormal alarm module, the electrocardiogram acquisition module pass through image pre-processing module
It is connect with model judgment module, the model judgment module is connect with abnormal alarm module；
The electrocardiogram acquisition module acquires electrocardiogram in real time and is sent to image pre-processing module, the image preprocessing mould
Block takes ROI region using sliding window, and ROI region is input to model judgment module, and the model judgment module is embedded
The convolutional neural networks model based on deep learning input data is judged, if it is decided that be abnormal electrocardiogram, the abnormal heart
Electric number ycount increases by 1, when ycount quantity is 3, the model judgment module control abnormal alarm module carry out and
Alarm.
2. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
The convolutional neural networks model construction based on deep learning are as follows: devise one using deep learning technology and be applied to the heart
The convolutional neural networks framework of dirty property sudden death auxiliary early warning, and die of illness in normal sinus rhythm database nsrdb and sudden cardiac
It dies and processing conclusion is carried out to ecg wave form on the basis of database sddb, trained test set is established by data enhancing, thus
Carry out training convolutional neural networks model, test data is tested using trained convolutional neural networks model.
3. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
The convolutional neural networks mode input image is 32 × 32.
4. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
The convolutional neural networks model shares 4 groups of network blocks composition, first 3 groups use 5 × 5 convolution kernel, the 4th group using 3 × 3
Convolution kernel.
5. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
The pond mode of the convolutional neural networks model is the side of " maximum dimensionality reduction-maximum dimensionality reduction-be averaged dimensionality reduction-be averaged dimensionality reduction "
Two network blocks of beginning are set maximum dimensionality reduction mode by formula, and the size with the 3rd, 4 network block output Feature Mapping subtracts
Small, ecg characteristics tend to sparsity expression, and the pixel value of surrounding is replaced using average value, in the same of effectively control over-fitting
When guarantee ecg characteristics expression in main effective information do not lose.
6. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
The activation primitive of the convolutional neural networks model chooses ReLU function.
7. a kind of sudden cardiac death based on deep learning according to claim 2 assists prior-warning device, which is characterized in that
The data enhancing specifically: electrocardiogram (ECG) data is enhanced using scale transformation.
8. a kind of sudden cardiac death based on deep learning according to claim 1 assists prior-warning device, which is characterized in that
Electrocardio ROI is obtained using translation gliding window strategy, the acquisition time period is 6 seconds, and is screened in electrocardio ROI.
9. a kind of sudden cardiac death based on deep learning according to claim 8 assists prior-warning device, which is characterized in that
The screening specifically: in the sample that the acquired period is 6 seconds, the abnormal data time is more than or equal to 4.8 seconds
Data are as abnormal electrocardiogram sample, and the normal electrocardio time, the data not less than 4.8 seconds were as the normal heart in acquired sample
Electric sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810808839.6A CN109009074A (en) | 2018-07-19 | 2018-07-19 | A kind of sudden cardiac death auxiliary prior-warning device based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810808839.6A CN109009074A (en) | 2018-07-19 | 2018-07-19 | A kind of sudden cardiac death auxiliary prior-warning device based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109009074A true CN109009074A (en) | 2018-12-18 |
Family
ID=64645006
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810808839.6A Pending CN109009074A (en) | 2018-07-19 | 2018-07-19 | A kind of sudden cardiac death auxiliary prior-warning device based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109009074A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109805898A (en) * | 2019-03-22 | 2019-05-28 | 中国科学院重庆绿色智能技术研究院 | Critical illness Mortality Prediction method based on attention mechanism timing convolutional network algorithm |
| CN109903841A (en) * | 2019-03-01 | 2019-06-18 | 中山大学肿瘤防治中心 | A kind of the abnormality reminding method and device of superior gastrointestinal endoscope image |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392137A (en) * | 2014-12-02 | 2015-03-04 | 北京海思敏医疗技术有限公司 | Signal abnormity detecting method and device based on entropy |
| CN106308791A (en) * | 2016-08-29 | 2017-01-11 | 郑州大学第附属医院 | Auxiliary data long distance transmission method and device for electrocardiosignal recording system |
| CN106599549A (en) * | 2016-11-25 | 2017-04-26 | 上海联影医疗科技有限公司 | Computer-aided diagnosis system and method, and medical system |
| CN106725426A (en) * | 2016-12-14 | 2017-05-31 | 深圳先进技术研究院 | A kind of method and system of electrocardiosignal classification |
| CN106909778A (en) * | 2017-02-09 | 2017-06-30 | 北京市计算中心 | A kind of Multimodal medical image recognition methods and device based on deep learning |
| CN107103182A (en) * | 2017-03-28 | 2017-08-29 | 南京医科大学 | A kind of heart disease Warning System and method based on deep learning algorithm |
| CN107832737A (en) * | 2017-11-27 | 2018-03-23 | 乐普（北京）医疗器械股份有限公司 | Electrocardiogram interference identification method based on artificial intelligence |
| CN107945168A (en) * | 2017-11-30 | 2018-04-20 | 上海联影医疗科技有限公司 | The processing method and magic magiscan of a kind of medical image |
-
2018
- 2018-07-19 CN CN201810808839.6A patent/CN109009074A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392137A (en) * | 2014-12-02 | 2015-03-04 | 北京海思敏医疗技术有限公司 | Signal abnormity detecting method and device based on entropy |
| CN106308791A (en) * | 2016-08-29 | 2017-01-11 | 郑州大学第附属医院 | Auxiliary data long distance transmission method and device for electrocardiosignal recording system |
| CN106599549A (en) * | 2016-11-25 | 2017-04-26 | 上海联影医疗科技有限公司 | Computer-aided diagnosis system and method, and medical system |
| CN106725426A (en) * | 2016-12-14 | 2017-05-31 | 深圳先进技术研究院 | A kind of method and system of electrocardiosignal classification |
| CN106909778A (en) * | 2017-02-09 | 2017-06-30 | 北京市计算中心 | A kind of Multimodal medical image recognition methods and device based on deep learning |
| CN107103182A (en) * | 2017-03-28 | 2017-08-29 | 南京医科大学 | A kind of heart disease Warning System and method based on deep learning algorithm |
| CN107832737A (en) * | 2017-11-27 | 2018-03-23 | 乐普（北京）医疗器械股份有限公司 | Electrocardiogram interference identification method based on artificial intelligence |
| CN107945168A (en) * | 2017-11-30 | 2018-04-20 | 上海联影医疗科技有限公司 | The processing method and magic magiscan of a kind of medical image |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109903841A (en) * | 2019-03-01 | 2019-06-18 | 中山大学肿瘤防治中心 | A kind of the abnormality reminding method and device of superior gastrointestinal endoscope image |
| CN109805898A (en) * | 2019-03-22 | 2019-05-28 | 中国科学院重庆绿色智能技术研究院 | Critical illness Mortality Prediction method based on attention mechanism timing convolutional network algorithm |
| CN109805898B (en) * | 2019-03-22 | 2024-04-05 | 中国科学院重庆绿色智能技术研究院 | Critical death prediction method based on attention mechanism time sequence convolution network algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109886986A (en) | A kind of skin lens image dividing method based on multiple-limb convolutional neural networks | |
| CN110010219A (en) | Optical coherence tomography image retinopathy intelligent checking system and detection method | |
| CN110084259A (en) | A kind of facial paralysis hierarchical synthesis assessment system of combination face texture and Optical-flow Feature | |
| CN108143411A (en) | A kind of tranquillization state brain electricity analytical system towards Autism Diagnostic | |
| CN110246577B (en) | Method for assisting gestational diabetes genetic risk prediction based on artificial intelligence | |
| CN110120055B (en) | Fundus fluorography image non-perfusion area automatic segmentation method based on deep learning | |
| CN109558912A (en) | A kind of Alzheimer's disease classification method separating convolution based on depth | |
| CN106127769A (en) | A kind of brain Forecasting Methodology in age connecting network based on brain | |
| CN109726743A (en) | A kind of retina OCT image classification method based on Three dimensional convolution neural network | |
| CN109509552A (en) | A kind of mental disease automatic distinguishing method of the multi-level features fusion based on function connects network | |
| CN111951219B (en) | Thyroid eye disease screening method, system and equipment based on orbit CT image | |
| CN104840186A (en) | Evaluation method of autonomic nervous function of patient suffering from CHF (Congestive Heart-Failure) | |
| CN116386860A (en) | Diabetes and complications intelligent auxiliary prediction and diagnosis platform based on multiple modes | |
| Jayanthi et al. | Automatic diagnosis of retinal diseases from color retinal images | |
| CN110379509A (en) | A kind of Breast Nodules aided diagnosis method and system based on DSSD | |
| CN110400301A (en) | A kind of cerebral apoplexy artificial intelligence screening method based on eye feature | |
| CN114038564A (en) | Noninvasive risk prediction method for diabetes | |
| CN115312198A (en) | Deep learning brain tumor prognosis analysis modeling method and system combining attention mechanism and multi-scale feature mining | |
| CN114549520A (en) | Retina pathological section analysis system based on full convolution attention enhancement network | |
| CN109009074A (en) | A kind of sudden cardiac death auxiliary prior-warning device based on deep learning | |
| CN114398979A (en) | Ultrasonic image thyroid nodule classification method based on feature decoupling | |
| CN116110597A (en) | Digital twinning-based intelligent analysis method and device for patient disease categories | |
| CN116363145A (en) | Fundus OCT image retina layering method based on double edge representation | |
| CN110236497A (en) | A kind of fatty liver prediction technique based on tongue phase and BMI index | |
| Xu et al. | A Multi-scale Attention-based Convolutional Network for Identification of Alzheimer's Disease based on Hippocampal Subfields |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |