CN102188240A

CN102188240A - Electrocardiographic data sampling method and device

Info

Publication number: CN102188240A
Application number: CN2010101185083A
Authority: CN
Inventors: 张嘉伟; 董军
Original assignee: East China Normal University
Current assignee: East China Normal University
Priority date: 2010-03-05
Filing date: 2010-03-05
Publication date: 2011-09-21

Abstract

The invention provides an electrocardiographic data sampling method, which is to sample the following data from an electrocardiograph: starting and ending points of a P wave, starting and ending points of a quantum resonance spectrometer (QRS) wave, starting and ending points of a T wave, form of the P wave, base line of the P wave, position of subwave of the QRS wave, form of the QRS wave, base line of QRS wave, form of the T wave, base line of the T wave, and heart beat rate diagnosis result. The invention also provides an electrocardiographic data sampling device for sampling the data. An electrocardiographic database obtained by using the sampling method and the sampling device provides a test platform with better automatic electrocardiograph characteristic identification and a better diagnosis method.

Description

A kind of ECG data method of sampling and device thereof

Technical field

The present invention relates to a kind of method of sampling and employed device of ECG data.

Background technology

Electrocardiogram (Electrocardiogram is called for short ECG) has reflected human body central authorities blood circulation, the i.e. active characteristics of heart.Electrocardiographic signal can provide the bulk information of the normal and ill physiological activity of heart.Therefore, electrocardiogram is a kind of clinical important diagnostic tool of heart disease of non-intrusion type.The early stage arrhythmia of finding fast especially at the critically ill patient in the intensive care unit, is crucial.Before fourth century, computer-aided diagnosis system (Computer-aideddiagnostic system) is applied to Electrocardiographic classification, and its technology that adopts is varied.Before the aid system practical application of any electrocardiogram classification, all must detect its effect by ecg database.At present, the standard database of main flow has three kinds.

Massachusetts reason arrhythmia data base (MIT-BIH Arrhythmia database) comprises 48 electrocardiographic recorders, every 30 minutes, comprise 2 data of leading.This data base is collected from 47 objects 1975 to 1979 by Bostonian Beth-Israel hospital arrhythmia laboratory.These electrocardiographic recording sample rates are 360 hertz, and data precision is 11.Every record is independently marked by the electrocardiogram expert more than 2 or 2, and finally reaches an agreement, and obtains the heart bat annotation results that computer can be read.

The multi-lead data base of the European Community (CSE) approximately provides 1000 short electrocardiographic recorders (be mainly 12 lead record).Every is recorded as 10 seconds length, and 500 hertz of sample rates have marked the beginning and the end position of P-QRS-T ripple.Because strict checking process, these annotation results are very accurate.Yet each Electrocardiographic heart is clapped diagnostic result and is not provided.

The American Heart Association (AHA) that is set up by National Heart, Lung, and Blood Institute at the year ends 1970, has set up AHA standard database construction project at the beginning of 1980 by patronage University of Washington, is devoted to the detection of ventricular arrhythmia.This data base has announced its first's labeled data in nineteen eighty-two, and finishes all mark work in 1985 the most finally.Comprise 155 2 electrocardiographic recordings that lead in the up-to-date AHA java standard library DVD release version, wherein related to 8 class cardiac arrhythmias.The record sample rate is 250 hertz, and data precision is 12.Every record reaches 3 hours, but has only last halfhour part to be marked.Table 1 has carried out simple relative analysis to these 3 data bases.

Table 1. three big mainstream standard ecg databases

	?MIT-BIH	CSE	AHA
				The quantity of leading	?2	12	2
Sample rate	?360	500	250
				Record length	30 minutes	10 seconds	3 hours
Record quantity	?48	About 1000	155

Price	Freely	Expensive	Expensive
				The mark section	All	All	Last 30 minutes

1997 American National cardiopulmonary Blood Research Institute (NHLBI) the standard cardioelectric data in future have been proposed following requirement: the electrocardiogram (ECG) data sample rate of collection is at least more than 500 hertz; Be at least 12 electrocardiogram (ECG) datas that lead; Possesses complete feature markup information; Every section electrocardiogram (ECG) data length needs more than 10 seconds; Comprise the essential information of being gathered the people; And the suggestion labeled data can be downloaded by the Internet.Though MIT-BIH data base full disclosure in three big standard databases is freely used for researcher, its data are 2 electrocardiogram (ECG) datas that lead, and sample rate only is 360 hertz, and only comprise R crest value position and the heart is clapped the mark of type.The AHA data base not only needs pay to buy, and its sample rate only is 250 hertz, and the markup information that is comprised is also identical with MIT-BIH.CSE database data sample rate meets recommended standard, its markup information comprises the beginning and the end position of P-QRS-T ripple, but the type that does not provide the heart to clap, this makes this can't carry out the electrocardiogram effect detection of classification automatically the data base, and the CSE data base costs an arm and a leg.

In three present big mainstream standard ecg databases, having only the MIT-BIH java standard library is to be full disclosure, has attracted a large amount of researcheres with its test platform as Characteristics of electrocardiogram identification and automatic classification method.In other two gyp java standard libraries, the AHA java standard library not only needs to buy, and its quality is not as the MIT-BIH java standard library; Electrocardiographic recording sample rate in the CSE java standard library is higher, meets the suggestion of NHLBI, but does not comprise the diagnostic result of clapping for the heart in its mark, and this method that makes this data base can only be used for feature identification detects.MIT-BIH data base's defective not only is the electrocardiographic recording parameter level that it is lower, and its less feature marks kind, and (only comprise QRS master's ripple, generally leading at II is the R ripple; Clap diagnostic message with the heart) also be a major issue, this makes other the conventional characteristic recognition methods in the electrocardiogram not detect by this data base.The CSE data base provides high-quality conventional feature markup information really, but can't be used for the test and validation of automatic diagnosis method.

In view of the deficiency of existing data base, consider the importance of electrocardiogram form feature simultaneously, a kind of new ECG data is sampled and the method for setting up a kind of new data base is badly in need of very much for research work.

Electrocardiographic feature diagnostic characteristic comprises conventional feature and this two big class of morphological characteristic.Wherein conventional feature is the basic foundation of electrocardiogram expert diagnosis.Use and morphological characteristic mostly is experienced doctor, be used for the case of efficient diagnosis complexity.

According to definition medically, the most important information of electrocardiogram concentrates on P ripple, QRS ripple and T ripple.These information spinners will comprise the position and the amplitude of QRS, PR, QT, ST interval and PR, ST section, calculate (beginning of P-QRS-T and end position) by datum mark.For example, the PR interval is that the starting position of P ripple is to the distance between the starting position of QRS ripple; The ST section is that QRS ripple end position is to the distance between the T ripple end position; The QRS interval is that the starting position of QRS ripple is to the distance between the end position of QRS ripple.

Based on above feature, the doctor just can judge the kind of heart disease substantially.Fig. 1 has provided the relation of these datum marks and conventional feature.

Electrocardiographic morphological characteristic is mainly reflected among P ripple, QRS ripple and the T ripple.Wherein the morphological characteristic of QRS ripple comprises: QRS waveform morphotype formula and QRS morphological character.

Before QRS waveform morphotype formula is discussed, need introduce the combination and the name of different positive directions of QRS ripple and negative direction wavelet in detail:

1) first downward wavelet is called the Q ripple

2) first wavelet that makes progress is called the R ripple

3) first the downward wavelet after the R ripple becomes the S ripple

4) first wavelet that makes progress after the S ripple is called R ' ripple

5) first downward wavelet is called S ' ripple after the R ' ripple

By the combination of different Q, R, S, R ' and S ', can produce many different QRS shape mode.For example, the QRS ripple of a complete forward is called the R type; A complete downward QRS ripple is called the QR type; Other pattern also has QR, qRs, RSR ' etc.The wavelet letter of capitalization is used to the wavelet of representing that amplitude is bigger, and the letter of small letter indicates the less wavelet of amplitude.But medical circle does not also have to provide the big and little accurate boundary value of unified amplitude at present, can provide a large amount of examples by setting up the ECG data storehouse in this.Fig. 2 has provided some relatively more typical QRS shape mode.

The QRS ripple also has two kinds of morphological characters in addition: thick blunt and incisura.Present diagnostic method is only paid close attention to the thick blunt and incisura that occurs in R and the S wavelet.The V-type vestige has appearred in incisura exactly in signal waveform, thereby and when this V-type vestige merged make primary signal increase in the signal slightly to be referred to as slightly blunt.Fig. 3 has provided in R, the S wavelet form of thick blunt and incisura.

The form of P ripple and T ripple also is determined with direction by the order of waveform turning point (wavelet), and its shape mode comprises: upright, inversion and two-way also have special form mould-bimodal (claiming the M type again) in addition for the P ripple.For two-way be like this definition: P or T ripple part are on baseline, and another part is under baseline.Fig. 4 has showed the shape mode of some P ripples and T ripple.

Provide two diagnostic rules here, be used to illustrate that how employed by the electrocardiogram expert these morphological characteristics are.

Left bundle branch block:

1) the P ripple of not losing promptly all can have a P ripple before each QRS ripple

2) the PR interval,, PR interval was meant the beginning to QRS of beginning of P between 0.12s-0.20

3) QRS ripple:＞0.11s

4) V1 presents rS

5) V6 has slightly blunt (end) at the R ripple; Perhaps V6 presents Rsr ', and at the R ripple incisura is arranged, and I and V5 present rSR ' and RSR ' respectively simultaneously, and V5 leads, and upward the R ripple is slightly blunt.

Premature atrioventricular junctional beat:

1) QRS ripple＜0.11s

2) times normal RR in current RR interval+next RR interval＜=2 interval

3) the PP interval stable, II leads and goes up the P ripple and be inverted, and than normal P ripple position in advance

4) the P ripple is inverted on I, II, III, aVF, V1, V5 and V6 lead, and is upright on aVR and aVL lead

5) II leads and presents the qRs type, and V1 leads and present the rS type, and V5 leads and presents qRs

Summary of the invention

The invention provides a kind of ECG data method of sampling, it is in the following data of electrocardiogram up-sampling: the P ripple begins, finishes, the QRS ripple begins, finishes, the T ripple begins, finishes, P marble ripple position, P ripple baseline, QRS marble ripple position, QRS waveform attitude, QRS baseline, T marble ripple position, T ripple baseline, the heart are clapped diagnostic result.

In addition, the present invention also provides a kind of ECG data sampling apparatus, comprise the online long distance monitoring of ECG center, it is in the following data of electrocardiogram up-sampling: the P ripple begins, finishes, the QRS ripple begins, finishes, the T ripple begins, finishes, P waveform attitude, and P ripple baseline, QRS marble ripple position, QRS waveform attitude, QRS baseline, T waveform attitude, T ripple baseline, the heart are clapped diagnostic result.

Effect of the present invention:

We have not only adopted the suggestion of NHLBI, and more detailed Characteristics of electrocardiogram mark is provided, and have especially added morphological characteristic.By paying attention to electrocardiogram expert's diagnosis experience, we find that morphological characteristic is effectively applied in the field, are important component parts in the electrocardiogram expert diagnosis cognitive model.

And the resulting ecg database of the method for sampling of the present invention provides identification of electrocardiogram automated characterization and the better test platform of diagnostic method.We provide all be 12 lead, the electrocardiographic recording of 500 hertz of samplings, and all be high-quality real patient data.More complete mark feature is provided simultaneously, comprises: the beginning of P-QRS-T ripple and end position; All morphological characteristics of leading of P-QRS-T ripple; Clap diagnostic result with the heart.Table 2 has provided the tabulation of the feature that we provided, and contrasts with existing storehouse.

The improvements of table 2. ecg database

	?MIT-BIH	CSE	?AHA	The method of sampling of the present invention institute The data base who obtains ^*
					The P ripple begins	Do not have	Have	Do not have	Have
P marble ripple position ^*	Do not have	Do not have	Do not have	Have
					The P ripple finishes	Do not have	Have	Do not have	Have
The QRS ripple begins	Do not have	Have	Do not have	Have
					QRS marble ripple position ^*	R ripple only	Do not have	R ripple only	All wavelets
QRS waveform attitude ^**	Do not have	Do not have	Do not have	Have
					The QRS ripple finishes	Do not have	Have	Do not have	Have
The T ripple begins	Do not have	Have	Do not have	Have
					T marble ripple position ^*	Do not have	Do not have	Do not have	Have
The T ripple finishes	Do not have	Have	Do not have	Have
					The heart is clapped diagnostic result	Have	Do not have	Have	Have
Baseline ^*	Do not have	Do not have	?N	Have

The morphological characteristic of P ripple and T ripple is to provide indirectly by the direction of its wavelet and order.Each wavelet position (Q, R, S, R ' and S ') of QRS ripple is provided by the resulting ecg database of the present invention with the shape mode of QRS and morphological character (slightly blunt, incisura) simultaneously.Baseline is because its necessary property when calculating amplitude, also has been incorporated into first among the resulting data base of the present invention.Fig. 5 has shown the Electrocardiographic example as a result behind the mark that obtains after the method for sampling according to the present invention is sampled.

Description of drawings

The conventional feature and the datum mark of Fig. 1 displayed map;

Fig. 2. show typical QRS shape mode;

Fig. 3 represents the thick blunt and incisura in the QRS ripple;

Fig. 4 represents the shape mode of P ripple and T ripple;

Fig. 5 is that the heart after finishing is clapped the annotation results example;

Fig. 6 represents ECG-Online long distance monitoring center;

Fig. 7 representation feature task topological order;

Fig. 8 shows electric data base's structure;

Fig. 9 A-9D represents management end and the partial function that marks the end aid.

The specific embodiment

The ECG data method of sampling and the device thereof of this aspect are described below in conjunction with the drawings and specific embodiments

1) data of database source

Carry out cardiac electrical sampling in order to set up, we have set up a cover system, and this cover system is called as ECG-Online long distance monitoring center, and its structure as shown in Figure 6.

2) Data base's structure

Adopt the background support system of MySQL data base, be used for depositing, inquire about and obtaining of management data as ecg database.MySQL data base high speed, stable, flexible, and be freeware.It is a multi-user, the SQL of multithreading (Structured Query Language) data server, and ecg database is very suitable for making up.

At the design aspect of database structure, follow simple suitable design guideline.Because this ecg database is not the demand that is used to handle hospital's complex environment, extract and sorting technique but be used to detect the electrocardiogram automated characterization, comprise necessary information and get final product.Fig. 8 has showed the core list structure in ECG data storehouse, and this structure mainly is divided into the two large divisions: electrocardiographic recorder and electrocardiogram mark.This data base's structural design not only can be managed electrocardiographic recording and annotation results thereof, and has taken into full account the management process of mark flow process.

Wherein the electrocardiographic recording part is made up of " Patient " and " ECGRecord " table.A patient can have many electrocardiographic recordings, and the attribute that comprises in " Patient " table has: Real Name (can not reveal during the data issue), sex, date of birth, nationality and historical case notes." ECGRecord " table has comprised the relevant important information of electrocardiographic recorder.By the sampling time and the date of birth that is sampled the people of record, this patient's age can obtain when electrocardiographic recording obtained.The primary signal data of electrocardiographic recording directly are not stored in the field of tables of data, but the mode of the associated data files of having sampled can be avoided the restriction of database table field greatest length like this." datasource " understands the source of electrocardiographic recording and the file format of storage with " dataformat " two field lists in the table.Although the overwhelming majority or whole data will come from our system, can not get rid of the probability of introducing other electrocardiographic recordings fully.

Another part, promptly the electrocardiogram mark is mainly concerned with " User ", " Annotation " and " AnnotationTask " these three tables.At first, create a mark task, describe the details of task in detail, and the original electrocardiographicdigital record packing that will need to mark sends to corresponding electrocardiogram expert (also being a kind of User).After mark work was done, its result was kept in " Annotation " table.Comprise two class roles in " User " table here, electrocardiogram expert and system manager, " type " distinguishes by its attribute.An electrocardiographic recording of finally including in ecg database must have its annotation results correspondence.Same annotation results is also deposited by the mode of associated with, rather than directly is stored among the field of table.Comprised the characteristic information of listing in all tables 2 in the mark file.

3) Labeled data form and computer program accesses interface

Many researchers is devoted to represent to store electrocardiographic recording with extend markup language (XML), and as ecgML and XML-ECG, it all respectively has superiority.These researchs mainly are in order to allow electrocardiographic recording that a kind of unified exchange and storage standards are arranged.Benefit with the XML storage is exactly to be easy to describe patient's the relevant information and the relevant parameter of electrocardiographic recording, but considers the preservation of a large amount of ecg wave form signal datas, will be a big problem if adhere to adopting the XML form.For example, for 500 hertz of 12 electrocardiographic recording that leads of 20 seconds, adopt the form of XML to store the space that will take greater than 50M.So 2 system contents of enormous amount have not only been flooded originally XML file structure clearly, and greatly reduce document analysis and the speed that reads, and under similar circumstances, the 2 system files of sampling are preserved only needs 1/4 space.So the electrocardiographic recording associated with among we data base adopts self-defining 2 system files to preserve.Also possess simultaneously the ability of supporting other kind electrocardiogram file formats because when database design, kept " datasource " and " data format " these two field attributes.

At present, because we have added the requirement of distinctive Characteristics of electrocardiogram mark, existing mark file memory format all can't satisfy our ecg database.The mark file format that therefore we formulate is based on the XML form.Showed below this form is how to preserve annotation results, and wherein " qR～s^ " expression QRS ripple is the qRs shape mode, and its R ripple of while slightly blunt and s ripple has incisura.

<？xml?version＝″1.0″encoding＝″utf-8″standalone＝″no″？>

-<Annotation?MaxS＝″1″Name＝″203_V0″>

-<Cycle>

-<P?has＝″true″>

<Onset?s＝″1″>295</Onset>

<Offset?s＝″1″>330</Offset>

-<Lead?type＝″II″>

<Deflections?s＝″1″>310</Deflections>

<Zero?s＝″1″>0.12</Zero>

</Lead>

+<Lead?type＝″III″>

</P>

-<QRS?has＝″true″>

<Onset?s＝″1″>350</Onset>

<Offset?s＝″1″>385</Offset>

-<Lead?type＝″II″>

<Deflections?s＝″1″>365，370，373</Deflections>

<Morphology?s＝″1″>qR～s^</Morphology>

<Zero?s＝″1″>0.12</Zero>

</Lead>

-<Lead?type＝″III″>

<Deflections?s＝″1″>370</Deflections>

<Morphology?s＝″1″>QS</Morphology>

<Zero?s＝″1″>0.12</Zero>

</Lead>

+<Lead?type＝″V1″>

</QRS>

-<T?has＝″true″>

<Onset?s＝″1″>439</Onset>

<Offset?s＝″1″>516</Offset>

-<Lead?type＝″II″>

<Deflections?s＝″1″>468，500</Deflections>

<Zero?s＝″1″>0.12</Zero>

</Lead>

+<Lead?type＝″III″>

</T>

</Cycle>

</Annotation>

Characteristic scalar explanatory notes spare form based on the XML form

The inquiry and the download service of electrocardiographic recording and mark finally are provided by internet website by the ecg database of MySQL database server governs.Institute's downloaded files comprises the necessary information that all MySQL data bases preserve.Every electrocardiographic recording is represented by 3 files: * .info, * .ecg and * .ann.* the .info file provides corresponding informance in database table " Patient " and " ECGRecord " table with the information of XML file, and it comprises the relevant information of patient and electrocardiographic recording.* preserved electrocardiographic recording primary signal data among the .ecg, its file format is by providing in the * .info file.Last * .info file is the annotation results file, its form with between the description that provides consistent.

4) Management end and mark end aid

In order to carry out data sampling and storage, the applicant has also designed management end and mark end aid.

Fig. 9 has showed the partial function of this instrument.The electrocardiogram mark task of whole issue can be subdivided into less one by one mark subtask.These subtasks have: mark " the QRS ripple starts from end position ", " the T ripple begins and end position ", " the P ripple begins and end position ", " shape mode of QRS and characteristic " on each leads, " QRS baseline " on each leads, " T waveform attitude " on each leads, " heart bat diagnostic result " etc.The order of these subtasks is determined according to the precedence relationship of its existence condition, for example: beginning and the end position that should determine the QRS ripple before mark QRS shape mode and characteristic earlier.

After mark end aid is opened, can import a task bag (task of packed original electrocardiographicdigital diagram data and mark).

Electrocardiogram comprises 12 and leads, and has showed the cardiomotility state that arrives that synchronization is gathered from different directions.Feature comprises in 2 on the scope of leading:

1. the feature of leading entirely

12 lead goes up unified feature, and it is whole unanimities on leading that for example beginning of P ripple and end position, the same heart are clapped at 12.

Belong to having of the feature of leading entirely: the P ripple begins, and the P ripple finishes, and the QRS ripple begins, and the QRS ripple finishes, and the T ripple begins, and the T ripple finishes, and the heart is clapped diagnostic result.

2. the feature of singly leading

Each leads and goes up inconsistent feature, QRS waveform morphotype formula for example, and in the same heart was clapped, it was qRs that II leads, it is RS that III leads.

Belong to having of the feature of singly leading: P marble ripple position, QRS marble ripple position, QRS waveform morphotype formula, QRS waveform step response, T marble ripple position, baseline.

5) The concrete method of sampling

Introduce concrete data sampling method below:

At first according to some dependences, necessity with in order to avoid the principle of artificial incorrect operation changes the above feature of standard that needs into the mark subtask, as shown in Figure 7, the order of carrying out on these feature The Logic of Tasks:

1.P ripple begins, finish (begin and finish to finish together, easy to operate, and event memory is also relatively easy.Be difficult for makeing mistakes, begin and forget to mark the mistake of end) in order to avoid the P ripple occurs only having marked

2.QRS ripple begins, and finishes

3.T ripple begins, and finishes

4.P waveform attitude (its preserving type is P marble ripple position and amplitude thereof)

5.P ripple baseline

6.QRS marble ripple position

7.QRS waveform attitude (pattern+characteristic)

8.QRS baseline

9.T waveform attitude (its preserving type is T marble ripple position and amplitude thereof)

10.T ripple baseline

11. the heart is clapped diagnostic result

Fig. 7 has described a prerequisite (topological order) that the feature task can be performed, and promptly a task can be performed, and then its prerequisite task is finished (those tasks of pointing to it need be finished earlier).For example, QRS waveform attitude will be carried out, and at first QRS ripple baseline and QRS marble ripple position all will be done, and will finish this two tasks, and then at first the QRS ripple begins, and end position will be marked and finish.Promptly order begins for the QRS ripple, end-＞QRS ripple baseline and QRS marble ripple position-＞QRS waveform attitude.And the task of satisfying the topological order condition can be performed, and does not have the requirement of sequencing, and promptly the P ripple begins to finish at first, and the QRS ripple begins to finish, and the T ripple begins to finish and heart bat diagnostic result can arbitrarily be carried out, and does not force sequencing.

These sequence requirement mainly are constraints in logic, also help the accuracy of standard simultaneously, and making simultaneously is the easier realization of marking program, effectively the result of management mark.

Here:

The P-QRS-T ripple need mark out beginning and end position earlier, and the feature (baseline and wavelet position) in the follow-up like this P ripple just is limited among this scope, helps the accuracy that the expert marks work.This also is the form of combined standard storage file, and such flow process is more reasonable.Otherwise marked P ripple baseline earlier, belonged to the P ripple that heart claps and to waste time and energy for locating this baseline afterwards.

For this order of QRS waveform attitude, be because after mark baseline and wavelet position, can verify the result of mark mutually by marking several forms that assistant software provides reference automatically as selection.For example, QRS wavelet quantity is 2, and that just can not be forms such as qRs, RSR ', and should be QR, Qr etc.

The enforceability of these tasks is handled by the mark assistant software, and for example: originally P baseline task can't enter, and after having finished P and beginning to end task, P baseline taskbar is just clicked effectively, can be performed.

By the ECG data method of sampling of the present invention and device, resulting database data type is more comprehensive, for diagnosis provides more information, and the resulting ecg database of the method for sampling of the present invention provides identification of electrocardiogram automated characterization and the better test platform of diagnostic method.

Claims

1. ECG data method of sampling, it is characterized in that, in the following data of electrocardiogram up-sampling: the P ripple begins, finishes, the QRS ripple begins, finishes, the T ripple begins, finishes, P waveform attitude, P ripple baseline, QRS marble ripple position, QRS waveform attitude, QRS baseline, T waveform attitude, T ripple baseline, the heart are clapped diagnostic result.

2. the ECG data method of sampling as claimed in claim 1 is characterized in that, described QRS waveform attitude comprises pattern and characteristic.

3. the ECG data method of sampling as claimed in claim 1 is characterized in that, in the following order sampled data:

1) the P ripple begins, finishes

2) the QRS ripple begins, finishes

3) the T ripple begins, finishes

4) P waveform attitude

5) P ripple baseline

6) QRS marble ripple position

7) QRS waveform attitude

8) QRS baseline

9) T waveform attitude

10) T ripple baseline

11) heart is clapped diagnostic result.

4. the ECG data method of sampling as claimed in claim 1 is characterized in that, the data of being sampled are preserved with the XML language.

5. ECG data sampling apparatus, comprise the online long distance monitoring of ECG center, it is characterized in that, this device is in the following data of electrocardiogram up-sampling: the P ripple begins, finishes, the QRS ripple begins, finishes, the T ripple begins, finishes, P waveform attitude, and P ripple baseline, QRS marble ripple position, QRS waveform attitude, QRS baseline, T waveform attitude, T ripple baseline, the heart are clapped diagnostic result.

6. ECG data sampling apparatus as claimed in claim 5 is characterized in that, described QRS waveform attitude comprises pattern and characteristic.

7. ECG data sampling apparatus as claimed in claim 5 is characterized in that, in the following order sampled data:

1) the P ripple begins, finishes

2) the QRS ripple begins, finishes

3) the T ripple begins, finishes

4) P waveform attitude

5) P ripple baseline

6) QRS marble ripple position

7) QRS waveform attitude

8) QRS baseline

9) T waveform attitude

10) T ripple baseline

11) heart is clapped diagnostic result.