CN104834921B - Electrocardio just/abnormal big data processing method and processing device - Google Patents
Electrocardio just/abnormal big data processing method and processing device Download PDFInfo
- Publication number
- CN104834921B CN104834921B CN201510268861.2A CN201510268861A CN104834921B CN 104834921 B CN104834921 B CN 104834921B CN 201510268861 A CN201510268861 A CN 201510268861A CN 104834921 B CN104834921 B CN 104834921B
- Authority
- CN
- China
- Prior art keywords
- data
- electrocardiogram
- ecg
- wave
- sorted
- 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.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
- G06F2218/16—Classification; Matching by matching signal segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Abstract
The invention discloses a kind of electrocardio just/abnormal big data processing method and processing device, this method includes normal ecg database and following step：S1, clapped by the heart and split electrocardiogram (ECG) data to be sorted, normalized then is done to length and amplitude respectively, forms some bat Wave datas；S2, the extraction electrocardiogram (ECG) data to be sorted achievement data；S3, confidential interval is determined according to the achievement data of the database purchase, and by the achievement data of the electrocardiogram (ECG) data to be sorted extracted compared with the confidential interval, export comparative result；S4, the similarity for calculating the Wave data that some bat Wave datas heart corresponding to the electrocardiogram (ECG) data of the database purchase being divided into from electrocardiogram (ECG) data to be sorted is clapped, the output comparative result compared with similarity threshold.The device includes normal ecg database and multiple modules for realizing above-mentioned steps.The present invention can reliably be sorted out examination by computer to ND electrocardiogram, avoid false negative.
Description
Technical field
The present invention relates to pattern-recognition, big data analysis, medical signals process field, specifically a kind of electrocardio just/it is abnormal
Big data processing method and processing device.
Background technology
Electrocardiogram is a clinical conventional inspection, by recording the situation of each cycle electrical activity of heart, can be helped
It is abnormal to diagnose arrhythmia cordis, impatient ischemic, miocardial infarction, atrioventricular hypertrophy, block, premature beat etc., it may also be used for judge electrolysis
The influence of matter and medicine to heart etc..
At present, the popularization of electrocardio examination, it is significantly limited by the Cardiologists quantity that can understand electrocardiogram.In
It is the research and development that many research institutions are directed to electrocardio auto-check system.Realize that electrocardio is classified automatically for computer both at home and abroad,
A variety of research methods are employed, such as artificial neural network, fuzzy clustering, logic discrimination tree, template matches ..., electrocardio is divided
During class, abnormal electrocardiogram is divided into multiple species by some according to different diseases；(such as room property is early for certain specific exceptions for some
Fight) it is classified.There has been no a kind of more ripe method, by identifying whether electrocardiographic abnormality supports electrocardio to sieve
Look into.
The content of the invention
It is an object of the invention to provide a kind of electrocardio just/abnormal big data processing method and processing device, enabling pass through meter
Calculation machine is reliably sorted out examination to ND electrocardiogram (electrocardiogram (ECG) data to be sorted), and avoiding false negative, (abnormal electrocardiogram is sentenced
Break as normal electrocardio) so that doctor only needs, to being judged as that abnormal progress diagnoses again, to reduce the workload of doctor.
The present invention concrete technical scheme be：
A kind of electrocardio just/abnormal big data processing method, the processing method includes normal ecg database, data stock
Normal electrocardiogram (ECG) data as much as possible is contained, every normal electrocardiogram (ECG) data includes achievement data and presses heart bat to electrocardiogram (ECG) data segmentation
Some bat Wave datas obtained；The processing method comprises the following steps：
S1, clapped by the heart and split electrocardiogram (ECG) data to be sorted, normalized then is done to length and amplitude respectively, if being formed
It is dry to clap Wave data；
S2, the extraction electrocardiogram (ECG) data to be sorted achievement data；
S3, confidential interval, and the electrocardio to be sorted that will be extracted are determined according to the achievement data of the database purchase
The achievement data of data exports comparative result compared with the confidential interval；And
S4, the electrocardio number for calculating some bat Wave datas being divided into from electrocardiogram (ECG) data to be sorted and the database purchase
The similarity for the Wave data that the corresponding heart is clapped, the output comparative result compared with similarity threshold in；
The achievement data includes between QRS wave segment length, PR at least one of phase between phase and RR between phase, QT.
Above-mentioned electrocardio just/abnormal big data processing method in, it is preferable that in the normal ecg database, every
Electrocardiogram (ECG) data, which corresponds to each heart and clapped, includes multistage Wave data, the multistage Wave data equal length, most overlapping, institutes
If the center for stating multistage Wave data is located at respectively at waveform peak and the front and rear of the waveform peak differs only by between each other
Dry data point；The step S4 comprises the following steps：
S41, by from the electrocardiogram (ECG) data to be sorted be divided into one clap Wave data respectively with the database purchase
The multistage Wave data that the corresponding heart is clapped in electric data of uniting as one calculates, and obtains relative to the multiple of the multistage Wave data
Similarity；
S42, minimum value is chosen from the multiple similarities relative to the multistage Wave data obtained, as from described
The bat Wave data that electrocardiogram (ECG) data to be sorted is divided into and phase in the electric data of uniting as one of the database purchase
Answer the similarity of the Wave data of heart bat；
S43, circulation perform the step S41 and S42, calculate described one be divided into from the electrocardiogram (ECG) data to be sorted and clap
The similarity for the Wave data that the Wave data heart corresponding to other electrocardiogram (ECG) datas of the database purchase is clapped；And
S44, circulation perform described step S41, S42, S43, obtain be divided into from the electrocardiogram (ECG) data to be sorted it is other
Clap the similarity for the Wave data that the Wave data heart corresponding to the electrocardiogram (ECG) data of the database purchase is clapped.
Above-mentioned electrocardio just/abnormal big data processing method in, it is preferable that the processing method is additionally included in segmentation and referred to
The step of electrocardiogram (ECG) data to be sorted is pre-processed before mark extraction.
Above-mentioned electrocardio just/abnormal big data processing method in, it is preferable that the comparative result bag of the step S4 outputs
Include：Normal electrocardiogram (ECG) data or abnormal electrocardiogram data, when the comparative result of output is abnormal electrocardiogram data, the comparative result of output
Also include the abnormal cycle being present, the cycle that exception be present refers to some bat waveform numbers in the electrocardiogram (ECG) data to be sorted
In, the similarity of the Wave data that corresponding to each bar electrocardiogram (ECG) data of the database purchase heart is clapped is all higher than described similar
Spend the heart corresponding to the Wave data of threshold value and clap the cycle.
Above-mentioned electrocardio just/abnormal big data processing method in, it is preferable that in the step S4, Similarity Measure bag
Include：The corresponding points of two sections of Wave datas to be compared are sought into difference one by one；And summing value after being taken absolute value to each difference, by this
With similarity of the value as two sections of Wave datas.
A kind of electrocardio just/abnormal big data processing unit, the processing unit includes：
Normal ecg database, the database purchase have normal electrocardiogram (ECG) data as much as possible, every normal electrocardiogram (ECG) data
Including achievement data and press some bat Wave datas of the heart bat to electrocardiogram (ECG) data segmentation acquisition；
Segmentation module, electrocardiogram (ECG) data to be sorted is split for being clapped by the heart, then length and amplitude normalized respectively
Processing, forms some bat Wave datas；
Index extraction module, for extracting the achievement data of the electrocardiogram (ECG) data to be sorted；
Targets match module, for determining confidential interval according to the achievement data of the database purchase, and will extraction
To electrocardiogram (ECG) data to be sorted achievement data compared with the confidential interval, export comparative result；And
Waveform Matching module, for calculating from some bat Wave datas that electrocardiogram (ECG) data to be sorted is divided into and the data
The similarity for the Wave data that the corresponding heart is clapped, the output comparative result compared with similarity threshold in the electrocardiogram (ECG) data of library storage；
The achievement data includes between QRS wave segment length, PR at least one of phase between phase and RR between phase, QT.
Above-mentioned electrocardio just/abnormal big data processing unit in, it is preferable that in the normal ecg database, every
Electrocardiogram (ECG) data, which corresponds to each heart and clapped, includes multistage Wave data, the multistage Wave data equal length, most overlapping, institutes
If the center for stating multistage Wave data is located at respectively at waveform peak and the front and rear of the waveform peak differs only by between each other
Dry data point；The Waveform Matching module includes：
First module, for will from the electrocardiogram (ECG) data to be sorted be divided into one clap Wave data respectively with the data
The multistage Wave data that the corresponding heart is clapped in the electric data of uniting as one of library storage calculates, and obtains relative to the multistage waveform number
According to multiple similarities；
Second module, for choosing minimum value from the multiple similarities relative to the multistage Wave data obtained,
Wave data is clapped with being united as one described in the database purchase as described one be divided into from the electrocardiogram (ECG) data to be sorted
The similarity for the Wave data that the corresponding heart is clapped in electric data；
3rd module, for the first module described in recursive call and the second module, calculate from the electrocardiogram (ECG) data to be sorted
Described one be divided into claps the Wave data that the Wave data heart corresponding to other electrocardiogram (ECG) datas of the database purchase is clapped
Similarity；And
4th module, for the first module, the second module and the 3rd module described in recursive call, obtain from described to be sorted
The Wave data that other bat Wave datas heart corresponding to the electrocardiogram (ECG) data of the database purchase that electrocardiogram (ECG) data is divided into is clapped
Similarity.
Above-mentioned electrocardio just/abnormal big data processing unit in, it is preferable that the processing unit also includes pretreatment mould
Block, for being pre-processed electrocardiogram (ECG) data to be sorted before segmentation and index extraction.
Above-mentioned electrocardio just/abnormal big data processing unit in, it is preferable that the comparison of the Waveform Matching module output
As a result include：Normal electrocardiogram (ECG) data or abnormal electrocardiogram data, when the comparative result of output is abnormal electrocardiogram data, the ratio of output
Relatively result also includes the abnormal cycle being present, and the cycle that exception be present refers to some bats in the electrocardiogram (ECG) data to be sorted
In Wave data, the similarity for the Wave data that the heart corresponding to each bar electrocardiogram (ECG) data of the database purchase is clapped is all higher than institute
State the heart corresponding to the Wave data of similarity threshold and clap the cycle.
Above-mentioned electrocardio just/abnormal big data processing unit in, it is preferable that in the Waveform Matching module, similarity
Calculating includes：The corresponding points of two sections of Wave datas to be compared are sought into difference one by one；And summed after being taken absolute value to each difference
Value, using this and it is worth the similarity as two sections of Wave datas.
For the present invention by the law mining to normal ecg database and analysis, the matching algorithm of feature based and waveform is real
Existing electrocardiogram (ECG) data classification, can reliably identify normal electrocardiogram (ECG) data, avoiding the occurrence of false negative, (abnormal electrocardiogram is judged as the normal heart
Electricity), therefore, it is possible to aid in doctor to carry out electrocardio examination, doctor is only needed to being judged as that abnormal progress diagnoses again, so as to
Greatly reduce the workload of doctor.
Brief description of the drawings
Fig. 1 be electrocardio of the present invention just/flow charts of abnormal big data processing method some embodiments；
Fig. 2-Fig. 6 is that one be divided into from electrocardiogram (ECG) data to be sorted is clapped in the electric data of uniting as one of waveform and database purchase
The match condition for five sections of waveforms that the corresponding heart is clapped.
Embodiment
The present invention is further described with reference to the accompanying drawings and examples.These more detailed descriptions are intended to help and understand this
Invention, and should not be taken to be limiting the present invention.According to present disclosure, it will be understood by those skilled in the art that can be not required to
Some or all these specific details is wanted to implement the present invention.And in other cases, in order to avoid innovation and creation are light
Change, well-known operating process is not described in detail.
As shown in figure 1, the electrocardio of some embodiments just/abnormal big data processing method includes normal ecg database, should
Database purchase has normal electrocardiogram (ECG) data as much as possible, and every normal electrocardiogram (ECG) data includes achievement data and presses heart bat to electrocardio
Some bat Wave datas that data sectional obtains；Wherein, the achievement data include between QRS wave segment length, PR between phase, QT the phase and
Phase between RR.
As a rule, can be by the collection and arrangement to up to ten thousand normal electrocardiogram (ECG) datas, to establish normal electrocardiogram (ECG) data
Storehouse, for the summary and analysis to normal electrocardiogram (ECG) data rule.It is as follows that normal ecg database specifically establishes process：
Step S100, each cycle of every normal electrocardiogram (ECG) data is intercepted out by peak-value detection method (i.e. normal
Electrocardiogram (ECG) data presses the heart and claps segmentation).Assuming that in a data, the data point of a certain cycle interception is kth point~kth+N points, then
By kth point~kth+N points, the point of kth -2~kth+N-2 points, the point of kth -1~kth+N-1, the point of kth+1~kth+N+1 points, kth+2
The methods of point~segment data of kth+N+2 points five is all stored in database respectively, and deposit passes through interpolation before does every section of interception
Length normalization method, and the amplitude of different segment datas is normalized.Therefore every electrocardiogram (ECG) data corresponds to each heart and claps bag
Include multistage (being five sections in this example) Wave data, the multistage Wave data equal length, most overlapping, multistage ripples
The center of graphic data is located at respectively at waveform peak and the front and rear of the waveform peak differs only by some data points between each other.
Step S200, by the index such as phase is deposited between phase, RR between phase, QT between the QRS wave segment length of every normal electrocardiogram (ECG) data, PR
Enter in database.
The electrocardio of some embodiments just/abnormal big data processing method comprises the following steps：
Step S1, clapped by the heart and split electrocardiogram (ECG) data to be sorted, normalized, shape then are done to length and amplitude respectively
Into some bat Wave datas.Specifically, the result that can be identified according to QRS complex, every bat waveform is split.Can during segmentation
The methods of with using dual threshold, determine that each cycle splits beginning and end.Assuming that the bat being divided into from electrocardiogram (ECG) data to be sorted
Wave data (a certain cycle data) is a={ a1, a2..., an, in order to make with every section of Wave data length one in database
Cause, the operation such as row interpolation is entered to it, the data after processing are
Step S2, the achievement data of the electrocardiogram (ECG) data to be sorted is extracted.The achievement data include QRS wave segment length,
Phase between phase and RR between phase, QT between PR.
Step S3, confidential interval is determined according to the achievement data of the database purchase, and it is to be sorted by what is extracted
The achievement data of electrocardiogram (ECG) data exports comparative result compared with the confidential interval.That is, first by targets match, to described
The whether normal of electrocardiogram (ECG) data to be sorted makes preliminary judgement.Specifically, can be according to the QRS of normal electrocardiogram (ECG) data in database
The indication range such as phase, average, variance between phase, RR between phase, QT between wave band length, PR, determine confidential interval, for the heart to be detected
The whether normal of electric data makes preliminary judgement.
Step S4, the heart from some bat Wave datas that electrocardiogram (ECG) data to be sorted is divided into and the database purchase is calculated
The similarity for the Wave data that the corresponding heart is clapped, the output comparative result compared with similarity threshold in electric data.I.e. Waveform Matching walks
Suddenly.
In Waveform Matching step mainly by do not diagnose the waveform of electrocardiogram (electrocardiogram (ECG) data to be sorted) with it is known just
Whether normal cardiac electrical similarity, normally judge to it.
It is closely related in view of similarity analysis and the relative position of two waveforms by compared with, so carrying out similarity
Obtained during calculating, while during calculating two waveform peaks alignment and when peak value relative position moves some data points multiple similar
Degree, therefrom takes the similarity that minimum value is final as this two waveforms.Therefore, the step S4 preferably includes following steps：
S41, by from the electrocardiogram (ECG) data to be sorted be divided into one clap Wave data respectively with the database purchase
The multistage Wave data that the corresponding heart is clapped in electric data of uniting as one calculates, and obtains relative to the multiple of the multistage Wave data
Similarity.Specifically, it is represented by below equation
Wherein,Represent to be divided into from the electrocardiogram (ECG) data to be sorted one claps Wave data, bI, j, k, wRepresent database
The multistage Wave data of j-th of cycle (heart bat) in i-th electrocardiogram (ECG) data of middle storage, p=1,2 ..., N, k=1,2 ...,
N represents the length of Wave data, and w=1,2 ..., 5 represents the hop count of multistage Wave data, and specifically corresponding to each cycle (has set
Initial point is kth point) it is stored in 5 segment datas (k-2~k+N-2 points, k-1 points~k+N-1, k~k+N points, k+1~k+N in database
+ 1 point, k+2 points~k+N+2 points).
Therefore in the present embodiment, Similarity Measure includes：By the corresponding points of two sections of Wave datas to be compared one by one
Seek difference；And summing value after being taken absolute value to each difference, using this and it is worth the similarity as two sections of Wave datas.But this hair
It is bright to be not limited to this, such as averaged after can also being taken absolute value to each difference, using the average as two sections of Wave datas
Similarity, other similar approach can also be used.
Uniting as one for bat waveform 1 and the database purchase being divided into from electrocardiogram (ECG) data to be sorted is shown in Fig. 2-Fig. 6
The match condition for five sections of waveforms 2,3,4,5,6 that the corresponding heart is clapped in electric data.
S42, minimum value is chosen from the multiple similarities relative to the multistage Wave data obtained, as from described
The bat Wave data that electrocardiogram (ECG) data to be sorted is divided into and phase in the electric data of uniting as one of the database purchase
Answer the similarity of the Wave data of heart bat.It is represented by below equation
S43, circulation perform the step S41 and S42, calculate described one be divided into from the electrocardiogram (ECG) data to be sorted and clap
The similarity for the Wave data that the Wave data heart corresponding to other electrocardiogram (ECG) datas of the database purchase is clapped.That is, above-mentioned
I, j are traveled through in formula.By the similarity Δ of calculatingI, jCompared with similarity threshold θ, when Δ being presentI, jDuring ＞ θ, the Duan Bo is judged
ShapeFor a normal cardiac electrical cycle, wherein similarity threshold θ can be set based on experience value.
S44, circulation perform described step S41, S42, S43, obtain be divided into from the electrocardiogram (ECG) data to be sorted it is other
Clap the similarity for the Wave data that the Wave data heart corresponding to the electrocardiogram (ECG) data of the database purchase is clapped.If all sections of phase
Similarity threshold θ is respectively less than like degree, then, Waveform Matching output result is that the electrocardiogram (ECG) data is normal electrocardiogram (ECG) data, is otherwise exported
The electrocardiogram (ECG) data is abnormal electrocardiogram data, and is exported in the presence of the abnormal cycle.
When These parameters matching and Waveform Matching result all for it is normal when, judge the electrocardiogram (ECG) data to be sorted be it is normal, it is no
Then to be abnormal.
Further, in certain embodiments, electrocardiogram (ECG) data to be sorted is carried out before being additionally included in segmentation and index extraction
The step S1 ' of pretreatment.These pretreatments can include：By basic electrocardiographicdata data Fast Classification, such as when heart rate exception, directly
Connect and judge electrocardiogram (ECG) data exception；For the normal electrocardiogram of basic electrocardiographicdata data, if baseline drift or noise are serious, it is passed through
The modes such as filtering carry out denoising.
Particularly point out, the sequence number explanation merely for convenience of step, does not represent specific ordinal relation, example in the present invention
Such as, step S1, S2, S3, S4 will not necessarily put in order execution by this in method is stated, and step S1 and S2 order can be right
Adjust, step S1 can also be placed on behind step S3, etc..
With some above-mentioned embodiment electrocardios just/the corresponding electrocardio of abnormal big data processing method just/abnormal big data handles
Device, including：
Normal ecg database, the database purchase have normal electrocardiogram (ECG) data as much as possible, every normal electrocardiogram (ECG) data
Including achievement data and press some bat Wave datas of the heart bat to electrocardiogram (ECG) data segmentation acquisition；
Segmentation module, electrocardiogram (ECG) data to be sorted is split for being clapped by the heart, then length and amplitude normalized respectively
Processing, forms some bat Wave datas；
Index extraction module, for extracting the achievement data of the electrocardiogram (ECG) data to be sorted；
Targets match module, for determining confidential interval according to the achievement data of the database purchase, and will extraction
To electrocardiogram (ECG) data to be sorted achievement data compared with the confidential interval, export comparative result；And
Waveform Matching module, for calculating from some bat Wave datas that electrocardiogram (ECG) data to be sorted is divided into and the data
The similarity for the Wave data that the corresponding heart is clapped, the output comparative result compared with similarity threshold in the electrocardiogram (ECG) data of library storage；
The achievement data includes between QRS wave segment length, PR at least one of phase between phase and RR between phase, QT.
In the normal ecg database, every electrocardiogram (ECG) data, which corresponds to each heart and clapped, includes multistage Wave data, described more
Section Wave data equal length, the overwhelming majority is overlapping, and the center of the multistage Wave data is located at waveform peak respectively
And the front and rear of the waveform peak differs only by some data points between each other；The Waveform Matching module includes：
First module, for will from the electrocardiogram (ECG) data to be sorted be divided into one clap Wave data respectively with the data
The multistage Wave data that the corresponding heart is clapped in the electric data of uniting as one of library storage calculates, and obtains relative to the multistage waveform number
According to multiple similarities；
Second module, for choosing minimum value from the multiple similarities relative to the multistage Wave data obtained,
Wave data is clapped with being united as one described in the database purchase as described one be divided into from the electrocardiogram (ECG) data to be sorted
The similarity for the Wave data that the corresponding heart is clapped in electric data；
3rd module, for the first module described in recursive call and the second module, calculate from the electrocardiogram (ECG) data to be sorted
Described one be divided into claps the Wave data that the Wave data heart corresponding to other electrocardiogram (ECG) datas of the database purchase is clapped
Similarity；And
4th module, for the first module, the second module and the 3rd module described in recursive call, obtain from described to be sorted
The Wave data that other bat Wave datas heart corresponding to the electrocardiogram (ECG) data of the database purchase that electrocardiogram (ECG) data is divided into is clapped
Similarity.
The processing unit also includes pretreatment module, for carrying out electrocardiogram (ECG) data to be sorted before segmentation and index extraction
Pretreatment.
The comparative result of the Waveform Matching module output includes：Normal electrocardiogram (ECG) data or abnormal electrocardiogram data, work as output
Comparative result when being abnormal electrocardiogram data, the comparative result of output also includes the abnormal cycle being present, it is described exist it is abnormal
Cycle refers in some bat Wave datas of the electrocardiogram (ECG) data to be sorted, each bar electrocardiogram (ECG) data with the database purchase
In the similarity of Wave data clapped of the corresponding heart be all higher than the heart corresponding to the Wave data of the similarity threshold and clap the cycle.
In the Waveform Matching module, Similarity Measure includes：By the corresponding points of two sections of Wave datas to be compared one by one
Seek difference；And summing value after being taken absolute value to each difference, using this and it is worth the similarity as two sections of Wave datas.
In the above-described embodiments, by the law mining to normal ecg database and analysis, feature based and waveform
Matching algorithm realizes that electrocardiogram (ECG) data is classified, and can reliably identify normal electrocardiogram (ECG) data, avoid the appearance of false negative.Therefore, may be used
To aid in doctor to diagnose a large amount of electrocardiogram (ECG) datas, i.e., before doctor diagnoses to a large amount of electrocardiogram (ECG) datas, first by meter
Calculation machine carries out automatic examination by the above method or device, and the abnormal electrocardiogram data that doctor need to only go out to examination diagnose.
Claims (4)
1. a kind of electrocardio just/abnormal big data processing unit, it is characterised in that the processing unit includes：
Normal ecg database, the database purchase have normal electrocardiogram (ECG) data as much as possible, and every normal electrocardiogram (ECG) data includes
Achievement data and press the heart clap to electrocardiogram (ECG) data be segmented obtain some bat Wave datas；
Segmentation module, electrocardiogram (ECG) data to be sorted is split for being clapped by the heart, normalized then is done to length and amplitude respectively,
Form some bat Wave datas；
Index extraction module, for extracting the achievement data of the electrocardiogram (ECG) data to be sorted；
Targets match module, for determining confidential interval according to the achievement data of the database purchase, and it will extract
The achievement data of electrocardiogram (ECG) data to be sorted exports comparative result compared with the confidential interval；And
Waveform Matching module, for calculating from some bat Wave datas that electrocardiogram (ECG) data to be sorted is divided into and the data stock
The similarity for the Wave data that the corresponding heart is clapped, the output comparative result compared with similarity threshold in the electrocardiogram (ECG) data of storage；
The achievement data includes between QRS wave segment length, PR at least one of phase between phase and RR between phase, QT；
In the normal ecg database, every electrocardiogram (ECG) data, which corresponds to each heart and clapped, includes multistage Wave data, the multistage ripple
Graphic data equal length, the overwhelming majority is overlapping, and the center of the multistage Wave data at waveform peak and is somebody's turn to do respectively
The front and rear of waveform peak differs only by some data points between each other；
The Waveform Matching module includes：
First module, for will from the electrocardiogram (ECG) data to be sorted be divided into one clap Wave data respectively with the data stock
The multistage Wave data that the corresponding heart is clapped in the electric data of uniting as one of storage calculates, and obtains relative to the multistage Wave data
Multiple similarities；
Second module, for choosing minimum value from the multiple similarities relative to the multistage Wave data obtained, as
From the electrocardiogram (ECG) data to be sorted the bat Wave data being divided into and electric number of being united as one described in the database purchase
The similarity for the Wave data that the corresponding heart is clapped in；
3rd module, for the first module described in recursive call and the second module, calculate and split from the electrocardiogram (ECG) data to be sorted
To described one clap the bat of Wave data corresponding in other electrocardiogram (ECG) datas of the database purchase heart Wave data it is similar
Degree；And
4th module, for the first module, the second module and the 3rd module described in recursive call, obtain from the electrocardio to be sorted
The phase for the Wave data that other bat Wave datas heart corresponding to the electrocardiogram (ECG) data of the database purchase that data are divided into is clapped
Like degree.
2. electrocardio according to claim 1 just/abnormal big data processing unit, it is characterised in that the processing unit is also wrapped
Pretreatment module is included, for electrocardiogram (ECG) data to be sorted being pre-processed before segmentation and index extraction.
3. electrocardio according to claim 1 just/abnormal big data processing unit, it is characterised in that the Waveform Matching mould
The comparative result of block output includes：Normal electrocardiogram (ECG) data or abnormal electrocardiogram data, when the comparative result of output is abnormal electrocardiogram number
According to when, the comparative result of output also includes the abnormal cycle being present, and described have an abnormal cycle and refer in the heart to be sorted
In some bat Wave datas of electric data, the Wave data of heart bat corresponding to each bar electrocardiogram (ECG) data of the database purchase
Similarity is all higher than the heart corresponding to the Wave data of the similarity threshold and claps the cycle.
4. electrocardio according to claim 1 just/abnormal big data processing unit, it is characterised in that the Waveform Matching mould
In block, Similarity Measure includes：The corresponding points of two sections of Wave datas to be compared are sought into difference one by one；And each difference is taken absolutely
To summing value after value, using this and it is worth the similarity as two sections of Wave datas.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510268861.2A CN104834921B (en) | 2015-05-22 | 2015-05-22 | Electrocardio just/abnormal big data processing method and processing device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510268861.2A CN104834921B (en) | 2015-05-22 | 2015-05-22 | Electrocardio just/abnormal big data processing method and processing device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104834921A CN104834921A (en) | 2015-08-12 |
| CN104834921B true CN104834921B (en) | 2018-01-16 |
Family
ID=53812799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510268861.2A Expired - Fee Related CN104834921B (en) | 2015-05-22 | 2015-05-22 | Electrocardio just/abnormal big data processing method and processing device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104834921B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018166136A1 (en) * | 2017-03-14 | 2018-09-20 | 华为技术有限公司 | Method and apparatus for processing waveform signal |
| CN108090509B (en) * | 2017-12-13 | 2021-10-08 | 四川大学 | Data length self-adaptive electrocardiogram classification method |
| CN108227930A (en) * | 2018-01-18 | 2018-06-29 | 四川斐讯信息技术有限公司 | The gestural control method and wearable device of a kind of wearable device |
| CN109754877B (en) * | 2018-01-23 | 2023-12-22 | 上海移视网络科技有限公司 | Twelve-lead standard electrocardiogram acute myocardial infarction intelligent discrimination system based on artificial intelligence |
| CN108903956A (en) * | 2018-05-22 | 2018-11-30 | 上海掌门科技有限公司 | It detects a lie device and method |
| CN113164071A (en) * | 2018-08-07 | 2021-07-23 | 金达中华有限公司 | Health map for navigating health space |
| CN113349791A (en) * | 2021-05-31 | 2021-09-07 | 平安科技（深圳）有限公司 | Abnormal electrocardiosignal detection method, device, equipment and medium |
| CN113576422B (en) * | 2021-10-08 | 2021-12-24 | 隽智生物医学研究实验室(佛山)有限公司 | Method, system and equipment for simplifying processing and quantizing physiological signal waveform |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103110417A (en) * | 2013-02-28 | 2013-05-22 | 华东师范大学 | Automatic electrocardiogram recognition system |
| CN104102915A (en) * | 2014-07-01 | 2014-10-15 | 清华大学深圳研究生院 | Multiple-template matching identity recognition method based on ECG (Electrocardiogram) under electrocardiogram abnormality state |
| CN104473633A (en) * | 2014-12-31 | 2015-04-01 | 广州视源电子科技股份有限公司 | Judging method and device of abnormal electrocardio data |
| CN104523265A (en) * | 2014-12-31 | 2015-04-22 | 广州视源电子科技股份有限公司 | Judging method and device for abnormal electrocardio data |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7136695B2 (en) * | 2001-10-12 | 2006-11-14 | Pless Benjamin D | Patient-specific template development for neurological event detection |
-
2015
- 2015-05-22 CN CN201510268861.2A patent/CN104834921B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103110417A (en) * | 2013-02-28 | 2013-05-22 | 华东师范大学 | Automatic electrocardiogram recognition system |
| CN104102915A (en) * | 2014-07-01 | 2014-10-15 | 清华大学深圳研究生院 | Multiple-template matching identity recognition method based on ECG (Electrocardiogram) under electrocardiogram abnormality state |
| CN104473633A (en) * | 2014-12-31 | 2015-04-01 | 广州视源电子科技股份有限公司 | Judging method and device of abnormal electrocardio data |
| CN104523265A (en) * | 2014-12-31 | 2015-04-22 | 广州视源电子科技股份有限公司 | Judging method and device for abnormal electrocardio data |
Non-Patent Citations (1)
| Title |
|---|
| 智能处理在ECG检测中的应用研究;聂桥桥;《中国优秀硕士学位论文全文数据库》;20130115;第I136-123页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104834921A (en) | 2015-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104834921B (en) | Electrocardio just/abnormal big data processing method and processing device | |
| Acharya et al. | A deep convolutional neural network model to classify heartbeats | |
| CN107358196B (en) | Heart beat type classification method and device and electrocardiograph | |
| Fang et al. | Human identification by quantifying similarity and dissimilarity in electrocardiogram phase space | |
| US7769434B2 (en) | Method of physiological data analysis and measurement quality check using principal component analysis | |
| CN103110417B (en) | Automatic electrocardiogram recognition system | |
| CN102835954B (en) | Method for generating heart beating waveform template and treating ventricular premature beat detection and device | |
| Li et al. | Inter-patient arrhythmia classification with improved deep residual convolutional neural network | |
| Ye et al. | Combining general multi-class and specific two-class classifiers for improved customized ECG heartbeat classification | |
| Bhoi et al. | Arrhythmia and ischemia classification and clustering using QRS-ST-T (QT) analysis of electrocardiogram | |
| CN110367968B (en) | Right bundle branch retardation detection method, device, equipment and storage medium | |
| Kaya et al. | Feature selection using genetic algorithms for premature ventricular contraction classification | |
| CN109077720B (en) | Signal processing method, device, equipment and storage medium | |
| Poddar et al. | Automated classification of hypertension and coronary artery disease patients by PNN, KNN, and SVM classifiers using HRV analysis | |
| CN109770861B (en) | Method, device, equipment and storage medium for training and detecting cardioelectric rhythm model | |
| CN110226928A (en) | Atrial fibrillation is early with room and atrial fibrillation is with the poor recognition methods passed and device | |
| Yang et al. | Heartbeat classification using discrete wavelet transform and kernel principal component analysis | |
| CN113876335B (en) | Electrocardiogram and heart beat classification method, device, electronic equipment and medium | |
| CN115486855A (en) | Electrocardiogram heart beat classification method based on QRS wave group indefinite cycle superposition | |
| Afsar et al. | Detection of ST segment deviation episodes in ECG using KLT with an ensemble neural classifier | |
| Dhyani et al. | Analysis of ECG-based arrhythmia detection system using machine learning | |
| Karnaukh et al. | Evaluation of machine learning techniques for ecg t-wave alternans | |
| Lamba et al. | A survey of algorithms for feature extraction and feature classification methods | |
| Mao et al. | Automated detection of first-degree atrioventricular block using ECGs | |
| Subashini et al. | Identification and classification of heart beat by analyzing ecg signal using naive bayes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180116 Termination date: 20210522 |