CN101773385B - Intelligent chinese medicine pulse-taking system - Google Patents
Intelligent chinese medicine pulse-taking system Download PDFInfo
- Publication number
- CN101773385B CN101773385B CN2010100345244A CN201010034524A CN101773385B CN 101773385 B CN101773385 B CN 101773385B CN 2010100345244 A CN2010100345244 A CN 2010100345244A CN 201010034524 A CN201010034524 A CN 201010034524A CN 101773385 B CN101773385 B CN 101773385B
- Authority
- CN
- China
- Prior art keywords
- microbridge
- pick
- distance
- pulse wave
- wave spectrum
- 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
Images
Landscapes
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The invention discloses an intelligent Chinese medicine pulse-taking system, which comprises a microbridge sensor array which is formed by arranging a plurality of microbridge sensor standard cells, wherein in the microbridge sensor array, the distance between the adjacent two microbridge sensor standard cells presents in an exponential type distribution from the center to the edge. The invention is in accordance to the intensity distribution of the real pulse signal, the middle of which is strong and dense while the edge thereof is weak and sparse, and signal transmission accords with the features of a certain exponential law, and the invention can real-timely and accurately realize Chinese medicine pulse examination. A main control chip comprises a core processor and multi-coprocessors, wherein the multi-coprocessors exchange information with the core processor by an internal bus, and thereby the operating speed of the system is improved.
Description
Technical field
The present invention relates to a kind of pulse wave spectrum equipment, relate in particular to a kind of intelligent pulse wave spectrum system.
Background technology
Pulse wave spectrum is with a long history, abundant in content, is most characteristic in a China's traditional medicine diagnostic method, but because " the arteries and veins reason is precise and tiny, and its body difficulty distinguishes that easy at the heart, difficulty is bright under referring to " affects objectifying and scientific development of pulse study for a long time.Exploring aspect three ones traditional of cun,guan,chis feel the pulse, find that by clinical trial cunkou pulse distributes internal organs that certain clinical meaning is arranged.But all have certain problem, main cause be to nonstationary random signal obtain, the shortcomings and deficiencies of analysis, Recognition Theory and technology.
In the prior art, the pulse wave spectrum system adopts the single-point acquiring pulse condition, adopts wavelet transformation, PRS that signal is handled, remedy to nonstationary random signal obtain, the shortcomings and deficiencies of analysis, Recognition Theory and technology.
As shown in Figure 1, it to the pulse signal processing procedure is: application sensors changes into the signal of telecommunication to pulse signal, adopt default threshold that signal is carried out denoising Processing, utilize wavelet transformation to extract the characteristic parameter of pulse condition, again the further Neural Network Optimization of pulse condition eigenvalue, mould are known identification, and the query calls data base, finally on touch screen, show diagnostic result.
There is following shortcoming at least in above-mentioned prior art:
The single-point acquiring pulse condition can not imitate the true shape of finger solid surface, so the lost part disease information; In addition, the medical pulse pressure signal of sensor of single-point is very faint, the influence of signal easily is interfered, pulse signal is changed a series of processing procedures through pressure transducer to A/D, the accuracy rate of gathering will reduce greatly, because wavelet transformation, pattern recognition, Neural Network Optimization, and computing complexity such as specialist system, the arithmetic speed and the precision of system are lower, are difficult to reach the requirement of real-time, accuracy again.
Summary of the invention
The intelligent pulse wave spectrum system that the purpose of this invention is to provide that a kind of energy in real time, accurately realizes feeling the pulse.
The objective of the invention is to be achieved through the following technical solutions:
Intelligent pulse wave spectrum of the present invention system comprises the microbridge sensor array that is arranged to by a plurality of microbridge pick off standard blocks, and in the described microbridge sensor array, the distance between adjacent two microbridge pick off standard blocks is exponential type and distributes from middle mind-set edge.
As seen from the above technical solution provided by the invention, intelligent pulse wave spectrum of the present invention system, owing to comprise the microbridge sensor array that is arranged to by a plurality of microbridge pick off standard blocks, in the described microbridge sensor array, the distance between adjacent two microbridge pick off standard blocks is exponential type and distributes from middle mind-set edge.Realistic pulse signal intensity distributions is strong and intensive in the middle of being, and the edge is weak and sparse, and the signal transmission meets the characteristics of certain index law, can realize in real time, accurately that the traditional Chinese medical science feels the pulse.
Description of drawings
Fig. 1 is the signal processing flow figure of prior art pulse wave spectrum system;
Fig. 2 is the structural representation of the specific embodiment of array microbridge pick off among the present invention;
Fig. 3 is the theory diagram of the specific embodiment of reading circuit among the present invention;
Fig. 4 is the schematic diagram of the specific embodiment of main control chip among the present invention;
Fig. 5 is the schematic diagram of the specific embodiment of main control chip internal bus among the present invention.
The specific embodiment
Intelligent pulse wave spectrum of the present invention system, its preferable specific embodiment is:
Comprise the microbridge sensor array that is arranged to by a plurality of microbridge pick off standard blocks, in the described microbridge sensor array, the distance between adjacent two microbridge pick off standard blocks is exponential type and distributes from middle mind-set edge.
Described exponential type distributes and can comprise:
Mind-set number of edges from described microbridge sensor array, the distance at first microbridge pick off standard block matrix row center is 2
0, second microbridge pick off standard block is 2 apart from the distance of first microbridge pick off standard block
1, the distance of second microbridge pick off standard block of the 3rd microbridge pick off standard block distance is 2
2, the distance of the 3rd microbridge pick off standard block of the 4th microbridge pick off standard block distance is 2
3, and the like.
Also can adopt other exponential type to distribute.
Described microbridge sensor array can be square array or circular array etc.Such as, adopt 8 * 8 square array.
Described microbridge sensor array is connected with reading circuit, and described reading circuit can comprise successively the integrating amplification circuit that connects, sampling hold circuit, output buffer, multi-channel transmission channel, A/D change-over circuit etc.
Intelligent pulse wave spectrum of the present invention system comprises main control chip, and described main control chip comprises a core processor and many coprocessors, and described coprocessor is by internal bus and core processor interactive information.
Such as, described coprocessor has 4, is respectively applied for to realize following date processing:
Wavelet transformation, mould are known identification, Neural Network Optimization, specialist system.
Described internal bus can comprise many data channel that be arranged in parallel, and the data transfer direction in many passages between the adjacency channel is opposite.
The intelligence system of pulse wave spectrum of the present invention at the characteristics of pulse signal, has proposed to use the method for array microbridge pick off, this array pitch is not unified, according to the characteristics of pulse signal, design spacing for the array that the exponential type rule distributes, be particularly useful for responding to pulse signal; Array-type sensor can be provided with the reference of pick off, eliminates the incidental error that is brought by spot measurement, and systematic analysis software can also carry out the multicomponent analysis simultaneously to the signal that array-type sensor obtains, thereby increases work efficiency widely; The actual pulse signal intensity profile is strong and intensive in the middle of being, and the edge is weak and sparse, and the signal transmission meets certain index law, among the present invention in the sensor array design of each collection point also meet this exponential rule.
The present invention is directed to the requirement of signal processing system, designed the high-performance CPU that has many coprocessors, improved the arithmetic speed of system.
Below by specific embodiment, and in conjunction with the accompanying drawings, the present invention is described in detail:
The specific embodiment of array microbridge pick off, as shown in Figure 2:
Under the situation that does not influence certainty of measurement, in order to reduce sampling number, therefrom mind-set is outer with the exponential form decay, the structure that sensor design is 8 * 8 arrays, intermediate density is big, marginal density is little according to the pulse oscillation intensity.
Because about the about 4mm of diameter of adult's cun,guan,chi tremulous pulse, the finger area is about 1cm
2, the pulse oscillation intensity therefrom decays with exponential form outside the mind-set.Measure when realizing body surface multiple spot large tracts of land physical quantity, with the microbridge sensor array design become the intermediate density of 8 * 8 arrays big, the marginal density minor structure.
Whole sensitive zones area is about 10mm, and establishing d is parasang, and the size of single microbridge pick off standard block is about about 2 * 2d, and then parasang d draws by (1) formula:
2d(1+2
1+2
2+2
3+2)≈10mm
d≈0.29412mm (1)
For convenient processing d gets 0.3mm, then the overall dimensions of microbridge sensor array is about 10.2mm * 10.2mm.Outside sensitive zones, reserve the marginal area of 0.5mm again, so whole array microbridge pick off is of a size of 11.2 * 11.2mm.
In the manufacturing process, can at first design microbridge sensor array standard block, with the standard block expansion, form 8 * 8 array structures, and use pin interconnection then, at last lead-in wire is drawn out to press welding block.In the design of topological structure, each cell block has two pins between standard block, is expanded into 8 * 8 array.
The specific embodiment of reading circuit, as shown in Figure 3:
Because pick off has charge generation under the pulse signal effect, this quantity of electric charge is very little can not directly be measured, and must the quantity of electric charge be converted into voltage through signal amplifying apparatus and just can gather, so reading circuit should contain integral amplifier.Pulse signal is very faint, be subject to electromagnetic interference, need be integrated into A/D converter in the reading circuit, the analog output signal that is subject to electromagnetic interference is transferred to the strong digital signal of capacity of resisting disturbance, to improve the overall performance of system, can also simplify the interface of array and system.Simultaneously pulse signal is time dependent, and if in transformation process signal level change, transformation result can have than mistake with specifying instantaneous analogue signal.In order to reduce error, an integrated sampling hold circuit (S/H circuit) behind integral amplifier, when " sampling " instruction arrived, switch S was connected circuit and is entered sampling configuration, and the output and the input signal of S/H circuit together change.When " maintenance " instruction arrived, switch S disconnected, and voltage on the capacitor C and S/H circuit output voltage all remain on the analog signal values that " maintenance " instruction arrives moment.And go here and there the conversion back by an output lead output, and the driving force that every row output need be certain, therefore the cross-talk buffer that constitutes with device with the one-level source not only can reduce the signal cross-talk between row and row, column and the row, and can drive bigger lead-in wire load.External with the current source biasing that device constitutes cross-talk buffer stage circuit, can suitably adjust the size and the output voltage swing of drive current by the source.Integration amplifies, sampling keeps, exports buffering, multiplexing, A/D conversion etc. so reading circuit should comprise.
Because the complexity of pathology, sometimes need be in planar array windowing arbitrarily, examine interested local message, in order to realize this function, the shift register of sequential scanning is changed into the random address generator, as to the random address generator of calculator memory addressing.
The specific embodiment of main control chip:
Main control chip carries out date processing to the gained signal, mainly comprises four parts:
1, wavelet transformation:
1. adopt wavelet function sym8 that signal is carried out the default threshold denoising Processing;
2. utilize wavelet transformation to obtain best time domain resolution of signal and frequency domain resolution;
3. adopt wavelet analysis that the information extraction of pulse signal different frequency section is come out, describe the feature of pulse signal;
4. measure the cycle of pulse condition;
5. adopt wavelet basis function db1 small echo to extract pulse condition temporal signatures parameter.
6. press the energy computing formula
Calculate the energy on each yardstick.Be used for the pulse condition identification of next step neutral net.
2, mould identification:
1. make up rational neural network structure.Use forward-propagating process and error back propagation process repeatedly, carry out e-learning.Since the logarithm sigmoid function with neuronic input range (∞ ,+∞) be mapped to (0 ,+1), and this function can be little, so be transfer function with logarithm S shape transfer function, and adds factor of momentum in training, adopts self adaptation to regulate learning rate.
2. choose less learning rate for the stability of assurance system, adopt the strategy of the learning rate self adaptation adjustment that adds momentum term, the convergence rate that overcomes learning algorithm is slow.
3. with the frequency spectrum 2 energy new feature values of pulse condition as neuronic input, improve not training the classification accuracy rate of sample.Frequency spectrum 2 energy features, the energy of choosing each yardstick behind the wavelet transformation are as input: the spectrum energy that (1) is total; (2) signal fundamental frequency; (3) difference at the interval of local maximum energy place frequency; (4) the local maximum energy value appears for the first time; (5) the local maximum energy value appears for the second time.
3, Neural Network Optimization:
1. further be optimized,, improve pulse analysis result's credibility by objectifying of membership function type determined to carry out the weight regulative mode with fuzzy inference rule by the fuzzy reasoning process of fuzzy neural network to the pulse signal Classification and Identification.
2. further by the study of fuzzy neural network, adapt to the variation of pulse signal data set; By increase and decrease, the polymerization of pulse condition fuzzy input variable degree of membership type scope and weights adjusting, the adjusting of fuzzy rule weights and fuzzy rule, improve the accuracy of pulse condition fuzzy classification.
3. adopt the fuzzy neural network clustering algorithm, determine the variate-value scope more accurately, select the parameter value of membership function type and the corresponding membership function of calculating with objectifying.And the local learning algorithm of use fuzzy neural network, finish fuzzy inference rule and carry out the weight adjusting, improve pulse analysis result's credibility precision, simultaneously along with new fuzzy inference rule is learnt in the expansion of data set.
4, specialist system.
Date processing mainly is divided into four parts as the above analysis, and the operand of each part is all very big.
As shown in Figure 4, for improving the arithmetic speed of system, core is handled in 5 of the system integrations, and wherein 4 is coprocessor, and an other PPE has the complete function that modern processors has.
Contain cache and the conversion table of supporting real-time operation among the PPE, can the internal resource of entire chip be managed, improve arithmetic speed, reduce power consumption.PPE starts the work of DMA request coprocessor by memory map I/O (MMIO) control register, can also support to communicate by letter with coprocessor; Allow the multidiameter delay operating system simultaneously, and by total line traffic control support.Thereby system can make other coprocessor be in sleep state by software control when carrying out simple data acquisition, air pump control; When carrying out wavelet transformation, pattern recognition, Neural Network Optimization, can pass through one or several coprocessor of software startup.
Because the inside of dsp chip adopts Harvard structure that program and data separate, have special hardware multiplier, adopt pile line operation, special DSP instruction is provided, can realize various digital signal processing algorithms fast, thus in each coprocessor an integrated DSP core devices.
As shown in Figure 5, coprocessor communicates with PPE by the inner bus EIB (Element Interconnect Bus) that connects in unit.The EIB bus is made up of the data channel of 4 128bit bandwidth, and for the signal that reduces between data channel disturbs, the data channel in the EIB bus is parallel or vertical, and makes in these 4 passages that data transfer direction is different between the adjacency channel.
Need owing to drive different internal memories, PPE has two different cache memories: L1 to be used for the standard L2 of instruction and data and 512kB with the EIB seam.The L1macro of 32kB follows depositor to constitute by two inputs, as seen from the figure, L1 is made of 3 layers of circulation, and the ground floor circulation is total address decoder, array access is carried out in second layer circulation, and the 3rd layer of circulation carried out last even-odd check, data format composing, reached the final circuit selection of L1; L2cache is by 48 road combined 512kB cache register of constituting of 1024 * 140macros independently, in order to reduce power consumption, is chosen in decoding before the array access.
The present invention adopts microbridge pick off standard block, and diaphragm-operated being shaped as has highly sensitive square, improves the certainty of measurement of system; The sensor array distribution design is the structure that intermediate density is big, marginal density is little, has enlarged the scope of measuring, and has improved the precision of measuring, and has reduced the number of sampled point; For improving the arithmetic speed of system, system proposes to adopt the method for using coprocessor in integrated 5 methods of handling core of CPU.Speed is fast, precision is high, can realize in real time, accurately that the traditional Chinese medical science feels the pulse.
The above; only for the preferable specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.
Claims (7)
1. intelligent pulse wave spectrum system, it is characterized in that, comprise the microbridge sensor array that is arranged to by a plurality of microbridge pick off standard blocks, in the described microbridge sensor array, the distance between adjacent two microbridge pick off standard blocks is exponential type and distributes from middle mind-set edge;
Described exponential type distributes and comprises:
Mind-set number of edges from described microbridge sensor array, the distance at first microbridge pick off standard block matrix row center is 2
0D, second microbridge pick off standard block is 2 apart from the distance of first microbridge pick off standard block
1D, the distance of second microbridge pick off standard block of the 3rd microbridge pick off standard block distance is 2
2D, the distance of the 3rd microbridge pick off standard block of the 4th microbridge pick off standard block distance is 2
3D, and the like;
d＝0.3mm。
2. intelligent pulse wave spectrum according to claim 1 system is characterized in that described microbridge sensor array is classified square array or circular array as.
3. intelligent pulse wave spectrum according to claim 2 system is characterized in that described square array is 8 * 8 square array.
4. according to each described intelligent pulse wave spectrum system of claim 1 to 3, it is characterized in that, described microbridge sensor array is connected with reading circuit, and described reading circuit comprises integrating amplification circuit, sampling hold circuit, output buffer, multi-channel transmission channel, the A/D change-over circuit that connects successively.
5. intelligent pulse wave spectrum according to claim 4 system is characterized in that comprise main control chip, described main control chip comprises a core processor and many coprocessors, and described coprocessor is by internal bus and core processor interactive information.
6. intelligent pulse wave spectrum according to claim 5 system is characterized in that described coprocessor has 4, is respectively applied for and realizes following date processing:
Wavelet transformation, mould are known identification, Neural Network Optimization, specialist system.
7. intelligent pulse wave spectrum according to claim 6 system is characterized in that described internal bus comprises many data channel that be arranged in parallel, and the data transfer direction in many passages between the adjacency channel is opposite.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010100345244A CN101773385B (en) | 2010-01-19 | 2010-01-19 | Intelligent chinese medicine pulse-taking system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010100345244A CN101773385B (en) | 2010-01-19 | 2010-01-19 | Intelligent chinese medicine pulse-taking system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101773385A CN101773385A (en) | 2010-07-14 |
| CN101773385B true CN101773385B (en) | 2011-08-31 |
Family
ID=42510071
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010100345244A Expired - Fee Related CN101773385B (en) | 2010-01-19 | 2010-01-19 | Intelligent chinese medicine pulse-taking system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101773385B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102988031A (en) * | 2012-12-30 | 2013-03-27 | 张锐 | Remote pulse diagnosis apparatus |
| CN104083157B (en) * | 2014-07-07 | 2016-04-20 | 北京印刷学院 | A kind of pulse signal recognition methods |
| CN104983406B (en) * | 2015-06-08 | 2018-04-27 | 之医（上海）健康科技有限公司 | A kind of network TCM pulse diagnosis system |
| CN109124595A (en) * | 2018-07-20 | 2019-01-04 | 南开大学 | Intelligent sphygmus diagnostic method based on dynamic 3 D pulse wave image |
| CN111449652B (en) * | 2020-05-06 | 2022-11-29 | 北方工业大学 | Construction safety monitoring method and device based on brain wave analysis |
-
2010
- 2010-01-19 CN CN2010100345244A patent/CN101773385B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN101773385A (en) | 2010-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101773385B (en) | Intelligent chinese medicine pulse-taking system | |
| CN104771163B (en) | EEG feature extraction method based on CSP and R CSP algorithms | |
| CN106295799B (en) | A kind of implementation method of deep learning multilayer neural network | |
| CN100415159C (en) | Dynamic characteristic analysis method of real-time tendency of heart state | |
| CN107730497A (en) | A kind of plaque within blood vessels property analysis method based on depth migration study | |
| CN101828917B (en) | Method and system for extracting electrocardiosignal characteristic | |
| CN106901723A (en) | A kind of electrocardiographic abnormality automatic diagnosis method | |
| CN104523264B (en) | Electrocardiosignal processing method | |
| Hao et al. | Breast cancer histopathological images classification based on deep semantic features and gray level co-occurrence matrix | |
| CN100518631C (en) | Three-dimensional dynamic pulse manifestation testing system | |
| CN108720831A (en) | A kind of automatic arrhythmia analysis method based on lead deep neural network | |
| CN110163360A (en) | A kind of computing device and method | |
| CN109584283A (en) | A kind of Medical Image Registration Algorithm based on convolutional neural networks | |
| CN108764457A (en) | The automatic arrhythmia analysis method of deep neural network is merged based on sampling channel | |
| CN110969614B (en) | Brain age prediction method and system based on three-dimensional convolutional neural network | |
| CN109949304B (en) | Training and acquiring method of image detection learning network, image detection device and medium | |
| CN108937912A (en) | A kind of automatic arrhythmia analysis method based on deep neural network | |
| CN109948646A (en) | A kind of time series data method for measuring similarity and gauging system | |
| Wang et al. | Lung CT image aided detection COVID-19 based on Alexnet network | |
| CN109620203A (en) | A kind of electrocardiosignal characteristic automatic extraction method based on one-dimensional convolutional neural networks | |
| CN205015889U (en) | Definite system of traditional chinese medical science lingual diagnosis model based on convolution neuroid | |
| CN110299207A (en) | For chronic disease detection in based on computer prognosis model data processing method | |
| CN107045591A (en) | A kind of electronic medical record system of female pelvic floor dysfunctional disease | |
| CN111047073B (en) | Aquaculture water quality prediction method and system based on neural network | |
| CN105943022B (en) | A kind of cardioelectric monitor system that there are three leads to reconstruct 12 lead function |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110831 Termination date: 20150119 |
|
| EXPY | Termination of patent right or utility model |