CN201683910U - Intelligent cardiopulmonary analyzing instrument - Google Patents
Intelligent cardiopulmonary analyzing instrument Download PDFInfo
- Publication number
- CN201683910U CN201683910U CN2010202155841U CN201020215584U CN201683910U CN 201683910 U CN201683910 U CN 201683910U CN 2010202155841 U CN2010202155841 U CN 2010202155841U CN 201020215584 U CN201020215584 U CN 201020215584U CN 201683910 U CN201683910 U CN 201683910U
- Authority
- CN
- China
- Prior art keywords
- unit
- cardiopulmonary
- heart
- signal
- sound
- 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 - Lifetime
Links
Images
Landscapes
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The utility model discloses an intelligent cardiopulmonary analyzing instrument comprising a cardiopulmonary sound transducer and a host machine wherein the host machine includes a signal processing unit, a sound outputting unit, an AD conversion unit, a DSP processing unit, a liquid crystal display unit as well as a sound saving unit. The sound processing unit receives signals transmitted by the cardiopulmonary sound transducer and sends out the signals respectively to the sound outputting unit and the AD conversion unit; the DSP processing unit receives and processes the signals from the AD conversion unit before transmitting the processed signals respectively to the liquid crystal display unit and the sound saving unit. The instrument acquires the cardiopulmonary signal of a person by adopting a transducer and the introduction of a high-speed DSP processor and a mould recognizing procedure to recognize acquired cardiopulmonary signal makes possible the detection of unusual cardiopulmonary actions, therefore, providing with doctors valuable information for diagnose. Further, convenient, practical to be carried around and having a more accurate detection rate, the instrument can be widely used at home or in community clinics.
Description
Technical field
This utility model relates to a kind of medical diagnostic equipment, and specifically a kind of family, community clinic etc. of being applicable to are used for the analysis and the identification of heart and lung sounds, can identify the unusual intelligent cardiopulmonary analyser of common heart and lung sounds.
Background technology
Auscultation is the sound that sends according to parts of body, judges whether normal a kind of diagnostic method of health with audition.By auscultation, the doctor can be according to the characteristic and the variation of sound, as the frequency height of sound, power, blanking time, noise etc., diagnoses relevant internal organs to have or not pathological changes.Because the heart and lung sounds of human body is fainter, and requires doctor's a lot of content of auscultation in a short period of time, have only through clinical experience accumulation all the year round, the doctor could tell the abnormal sound of cardiopulmonary more exactly.Though existing electronic stethoscope can amplify heart and lung sounds and handle, and makes heart and lung sounds more clear, auscultation still identifies the unusual of heart and lung sounds by clinical experience.
Existing auscultation apparatus is based on mostly that general-purpose computing system analyzes, this system constitutes huge, complicated, volume is excessive, cost is too high, be not easy to carry, as Chinese patent application number is 200680009502.4, and it number is 200910199630.5 that name is called " visual stethoscope " and Chinese patent application, is called " Bluetooth electronic cardiophone ".
The utility model content
For solving auscultation analytical system complexity, bulky weak point in the prior art, this utility model provides a kind of volume little easy to carry, and the recognition accuracy height is adapted at the intelligent cardiopulmonary analyser that uses in family and the community clinic.
For solving the problems of the technologies described above, the technical scheme of this novel employing is:
This intelligence cardiopulmonary analyser, have heart and lung sounds pick off and main frame, wherein main frame comprises signal processing unit, voice output unit, AD converting unit, dsp processor unit, liquid crystal display and recording memory element, wherein signal processing unit receives the heart and lung sounds signal of sensor, exports voice output unit and AD converting unit respectively to; After handling, the signal of dsp processor unit reception AD converting unit exports liquid crystal display and recording memory element respectively to.
Described signal processing unit is the structure connected in series that is formed by elementary amplifying circuit, low-pass filter circuit, high-pass filtering circuit and second amplifying circuit.
The dsp processor unit comprises crystal oscillator, the chip that resets, dsp processor and memorizer, and its dsp processor one termination is received crystal oscillator, chip signal output resets; Operation in the other end reception memorizer and start-up routine.
Said dsp processor adopts the TMS320C5000 series processors.
Said AD converting unit adopts the AD7705 chip.
The voice output unit is made up of sound driver chip TPA311 and speaker, and this speaker is a boombox.
The heart and lung sounds pick off adopts contact-type piezoelectric membrane pickup sheet.
This instrument compared with prior art has following beneficial effect:
1. information processing rate is fast, and the instrument volume is little, is fit to family and community and uses.This utility model utilizes sensor special to extract the heart and lung sounds signal of human body, adopts the high-speed dsp processor, has improved information processing rate.This instrument application mode recognition technology is carried out analyzing and processing to the heart and lung sounds signal that extracts, the identification of comparing after the feature of extraction signal, thus it is unusual to judge common cardiopulmonary, for the doctor provides valuable diagnostic message; And this instrument is easy to carry, uses simply, and the recognition accuracy height is highly suitable in family and the community clinic and uses.
2. has the intellectual analysis function.This utility model utilizes mode identification technology that the heart and lung sounds signal is analyzed, and automatically identifies the unusual kind of common cardiopulmonary, has improved the accuracy rate of medical diagnosis on disease effectively.
3. can amplify output to the heart and lung sounds signal.This utility model is provided with the voice output unit of being made up of sound driver chip TPA311 and speaker, is that acoustical signal is exported by speaker with the heart and lung sounds conversion of signals after handling.
4. has sound-recording function.This utility model is provided with the recording memory element and is used for the heart and lung sounds data that storage of collected arrives.The heart and lung sounds of gathering is stored with the form of wav, can transfer on the computer easily and preserve, and carries out the auscultation analysis for the doctor.
Description of drawings
Fig. 1 is this utility model system structure sketch map;
Fig. 2 is Fig. 1 signal processing unit structural representation;
Fig. 3 is Fig. 1 dsp processor cellular construction sketch map;
Fig. 4 is system's main program flow chart;
Fig. 5 is Fig. 4 pattern recognition program structure chart.
The specific embodiment
Below in conjunction with accompanying drawing this novel embodiment is described in detail:
With reference to accompanying drawing 1, this instrument is made up of heart and lung sounds pick off and main frame two parts, and its pick off is converted to the signal of telecommunication by audio connecting cord with the human body heart and lung sounds that collects and transfers to main frame.Main frame comprises signal processing unit, voice output unit, AD converting unit, dsp processor unit, liquid crystal display, recording memory element; Wherein signal processing unit receives the heart and lung sounds signal of sensor, exports voice output unit and AD converting unit respectively to; After handling, the signal of dsp processor unit reception AD converting unit exports liquid crystal display and recording memory element respectively to.
Said heart and lung sounds pick off adopts contact-type piezoelectric membrane pickup sheet, can more clearly obtain human body heart and lung sounds signal.Owing to directly do not carry out pickup, therefore can reduce the friction noise of skin and medicated clothing effectively from human body surface by the intermediate air layer.
With reference to accompanying drawing 2, signal processing unit carries out filtering and amplification with the heart and lung sounds sensor output signal, divides two-way to deliver to AD converting unit and sound output unit respectively then.It is made up of elementary amplifying circuit, low-pass filter circuit, high-pass filtering circuit, second amplifying circuit, and is connected in series in order between it.Elementary amplifying circuit amplifies chip OPA336 by the high input impedance instrument of special use and constitutes, and the heart and lung sounds signal that is used for its pick off is obtained tentatively amplifies; The low-pass filtering amplifying circuit is formed voltage controlled voltage source second-order low-pass filter amplifying circuit by the AD8544 chip, and it is 1KHz by frequency configuration.High-pass filtering circuit is formed voltage controlled voltage source second order high-pass filtering amplifying circuit by the AD8544 chip, and it is 20Hz by frequency configuration.Second amplifying circuit is formed the in-phase proportion operational amplification circuit by the AD8544 chip, and the heart and lung sounds signal is amplified.The heart and lung sounds signal is by behind the band filter of being made up of low-pass filtering amplifying circuit and high-pass filtering circuit, filtering the component environment noise, only kept the signal of 20Hz to 1KHz.
The voice output unit is made up of sound driver chip TPA311 and speaker, and its speaker is a boombox, can be that acoustical signal is exported by speaker with the heart and lung sounds conversion of signals after handling.The TPA311 chip connects in the BTL mode, so that output is more high-power, is used to drive speaker.
The AD converting unit adopts 16 AD7705 chips of the dual pathways of high accuracy, low-power consumption, this chip links to each other with the dsp processor unit by the SPI interface, sample rate with 16KHz is carried out analog digital conversion to mimic heart and lung sounds signal, and the data signal that generates is sent to the dsp processor unit by the SPI interface.Higher sample rate can guarantee that sampled signal is undistorted, in order to avoid influence postorder signal processing effect.
Liquid crystal display adopts 65K color TFT liquid crystal display screen, and liquid crystal display screen carries the driving chip, is connected by the I mouth with dsp processor, and be used for signals collecting and show and identifying information, be the man-machine interface part of instrument.
The recording memory element is provided with memorizer, adopts Nand FLASH memorizer, is used to preserve the heart and lung sounds digital signal after the analog digital conversion.Data storage adopts wav form, 16KHz sample rate and 16 sampling precisions.Wav formatted file after the storage can be transferred to easily and preserve on the computer and play.
With reference to accompanying drawing 3, the dsp processor unit adopts the TMS320C5000 series DSP processor of TI company.The main feature of this series DSP processor is low-power consumption, is fit to very much portable equipment and uses.The preferred TMS320C5416 dsp processor of the dsp processor of TMS320C5000 series has synchronous serial interface, HPI parallel interface, intervalometer, DMA peripheral hardware, has fast operation and becomes original low characteristics.
The dsp processor unit comprises crystal oscillator, the chip that resets, dsp processor and memorizer, and its dsp processor one termination is received crystal oscillator, the chip output end signal resets; Operation in the other end reception memorizer and start-up routine.Its memorizer has Nor Flash memorizer and eeprom memory, and memorizer Nor Flash is used for storing the program that DSP moves, and storage code can directly move in Nor Flash, has improved program operation speed widely.Eeprom memory adopts the storage chip of SPI interface, the startup boot that is used to store DSP.
With reference to accompanying drawing 4, the dsp processor unit is the key control unit of whole instrument, controls by memorizer intermediate range ordered pair heart and lung sounds signals sampling, front-end processing, feature extraction, model training, analysis identification, the storage that shows, records.
Front-end processing: mainly the heart and lung sounds signal is carried out preemphasis, divides frame, windowing, end-point detection, adopt double threshold detection method, remove the quiet and of short duration noise of heart and lung sounds in conjunction with short-time energy and short-time zero-crossing rate.
Feature extraction: adopt Mel frequency cepstral coefficient method (being called for short MFCC), after the heart and lung sounds signal being carried out preemphasis, dividing the processing of frame, windowing, end-point detection, extract the typical characteristic information of heart and lung sounds.
Model training: adopt the Baum-Welch algorithm that various typical heart and lung sounds characteristic informations are set up Hidden Markov (being called for short HMM) model respectively.
Analyze identification: adopt the Viterbi algorithm, calculate the matching probability of heart and lung sounds to be identified and each model, the heart and lung sounds model of choosing the probability maximum is as recognition result.
System program operation step is as follows:
During this analyzer system work: program operation beginning at first requires the user to input user name, preserves the heart and lung sounds that each user collects separately.If do not select, program is with the user name storage data of acquiescence.
Next program requires the user to select cardiac auscultation or auscultation of lung, and program shows and gather the position of auscultatory sound according to user's selection result.Because cardiac auscultation and auscultation of lung all have separately typical auscultation position, so the figure of program display also has nothing in common with each other, and the parameter setting of each wave filter is also inequality when follow-up signal processing.
At this moment the user is according to prompts displayed information on the screen, and the heart and lung sounds pick off is placed into the auscultation position that the user selects, and when hearing clear no friction sound from speaker, definite key of pressing on the instrument is gathered its auscultatory sound.According to transient copy, all auscultation positions are carried out the collection of auscultatory sound.
Next program is saved in the NandFLAS memory chip after the auscultatory sound that collects is numbered according to the auscultation position, then each section auscultatory sound is handled and analyzed, utilize pattern recognition program that heart and lung sounds is classified at last, recognition result is shown on the TFT LCD screen.
With reference to accompanying drawing 5, pattern recognition program is to move on the basis that the heart and lung sounds characteristic model is set up.Its processing procedure: the sample of at first gathering normal heart and lung sounds and various typical unusual heart and lung sounds, its feature will be extracted after the front-end processing, then every type heart and lung sounds is set up the HMM of self, will be set up good characteristic model at last and be saved in the Nor Flash memory chip of instrument.
When storage auscultatory sound when entering pattern recognition program, initial parameter at first is set, i.e. quantity, the counter O reset of the quantity of HMM model, the heart and lung sounds file that reads.Then the HMM model is loaded in the dsp processor, and reads the heart and lung sounds signal of storage.Then the auscultatory sound file that reads is carried out front-end processing: preemphasis, branch frame, windowing, end-point detection.Use the preemphasis main purpose to be exactly the power frequency interference of filtering low-frequency disturbance, especially 50Hz or 60Hz, less owing to the heart and lung sounds energy mainly concentrates between 10Hz～1kHz simultaneously at high-frequency energy, be subject to noise jamming.The adding preemphasis increases the weight of the HFS of heart and lung sounds, carries out frequency spectrum and promotes, and makes the frequency spectrum of signal become smooth, remains on low frequency in the whole frequency of high frequency, can ask frequency spectrum with same signal to noise ratio, so that spectrum analysis or channel parameters analysis; Because the heart and lung sounds signal has stationarity in short-term, thereby can handle with the analytical method of stationary random process.Based on this principle, when handling the heart and lung sounds signal, the heart and lung sounds signal be carried out the branch frame, calculate the short-time characteristic of each frame; That windowing is partly adopted is Hanning window (Hanning window).The characteristics of this window are that main lobe is widened, and secondary lobe significantly reduces, and side lobe attenuation speed is also very fast, have suppressed spectrum leakage effectively; End-point detection adopts the double threshold detection method in conjunction with short-time energy and short-time zero-crossing rate, removes the quiet and of short duration noise of heart and lung sounds.
Heart and lung sounds signal after front-end processing carries out feature extraction.This method adopts Mel frequency cepstral coefficient method (being called for short MFCC), at first time-domain signal is changed into frequency domain with fast Fourier transform FFT, use the triangular filter group that distributes according to the Mel scale to carry out convolution algorithm to its logarithm energy spectrum afterwards, the vector that each wave filter output is constituted carries out discrete cosine transform at last, and then extracts the characteristic information of heart and lung sounds.
The heart and lung sounds characteristic information that extracts is utilized the matching probability of Viterbi algorithm computation and the HMM model that is written into.The Viterbi algorithm is a kind of dynamic programming algorithm, is used for seeking the latent status switch of the most probable that is produced by observation information, is used in the hidden Markov model usually.Calculate the matching probability of heart and lung sounds MFCC feature to be identified and each model by this method, the heart and lung sounds model of choosing the probability maximum is as recognition result.
Claims (7)
1. intelligent cardiopulmonary analyser, it is characterized in that: have heart and lung sounds pick off and main frame, wherein main frame comprises signal processing unit, voice output unit, AD converting unit, dsp processor unit, liquid crystal display and recording memory element, wherein signal processing unit receives the heart and lung sounds signal of sensor, exports voice output unit and AD converting unit respectively to; After handling, the signal of dsp processor unit reception AD converting unit exports liquid crystal display and recording memory element respectively to.
2. by the described intelligent cardiopulmonary analyser of claim 1, it is characterized in that: the connected in series structure of described signal processing unit for forming by elementary amplifying circuit, low-pass filter circuit, high-pass filtering circuit and second amplifying circuit.
3. by the described intelligent cardiopulmonary analyser of claim 1, it is characterized in that: the dsp processor unit comprises crystal oscillator, the chip that resets, dsp processor and memorizer, and its dsp processor one termination is received crystal oscillator, chip signal output resets; Operation in the other end reception memorizer and start-up routine.
4. according to claim 1 or 3 described intelligent cardiopulmonary analysers, it is characterized in that: said dsp processor adopts the TMS320C5000 series processors.
5. intelligent cardiopulmonary analyser according to claim 1 is characterized in that: said AD converting unit adopts the AD7705 chip.
6. intelligent cardiopulmonary analyser according to claim 1, it is characterized in that: the voice output unit is made up of sound driver chip TPA311 and speaker, and this speaker is a boombox.
7. intelligent cardiopulmonary analyser according to claim 1 is characterized in that: the heart and lung sounds pick off adopts contact-type piezoelectric membrane pickup sheet.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010202155841U CN201683910U (en) | 2010-06-04 | 2010-06-04 | Intelligent cardiopulmonary analyzing instrument |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010202155841U CN201683910U (en) | 2010-06-04 | 2010-06-04 | Intelligent cardiopulmonary analyzing instrument |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN201683910U true CN201683910U (en) | 2010-12-29 |
Family
ID=43372201
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010202155841U Expired - Lifetime CN201683910U (en) | 2010-06-04 | 2010-06-04 | Intelligent cardiopulmonary analyzing instrument |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN201683910U (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103417241A (en) * | 2012-05-17 | 2013-12-04 | 辽宁中医药大学 | Automatic analyzer of lung sounds |
| CN103932733A (en) * | 2014-04-11 | 2014-07-23 | 中国人民解放军第三军医大学第三附属医院 | Digitalized detecting analysis method for pulmonary interstitial fibrosis based on lung sound |
| CN105662459A (en) * | 2016-04-12 | 2016-06-15 | 中国人民解放军第三军医大学第三附属医院 | Wearable lung sound detection device |
| CN107468275A (en) * | 2017-07-14 | 2017-12-15 | 舒林华 | Storage medium and electronic auscultation system for the processing of auscultation voice data |
| CN111150422A (en) * | 2020-03-09 | 2020-05-15 | 国微集团(深圳)有限公司 | Intelligent auscultation system and auscultation method thereof |
-
2010
- 2010-06-04 CN CN2010202155841U patent/CN201683910U/en not_active Expired - Lifetime
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103417241A (en) * | 2012-05-17 | 2013-12-04 | 辽宁中医药大学 | Automatic analyzer of lung sounds |
| CN103417241B (en) * | 2012-05-17 | 2017-03-22 | 辽宁中医药大学 | Automatic analyzer of lung sounds |
| CN103932733A (en) * | 2014-04-11 | 2014-07-23 | 中国人民解放军第三军医大学第三附属医院 | Digitalized detecting analysis method for pulmonary interstitial fibrosis based on lung sound |
| CN105662459A (en) * | 2016-04-12 | 2016-06-15 | 中国人民解放军第三军医大学第三附属医院 | Wearable lung sound detection device |
| CN107468275A (en) * | 2017-07-14 | 2017-12-15 | 舒林华 | Storage medium and electronic auscultation system for the processing of auscultation voice data |
| CN111150422A (en) * | 2020-03-09 | 2020-05-15 | 国微集团(深圳)有限公司 | Intelligent auscultation system and auscultation method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102697520B (en) | Electronic stethoscope based on intelligent distinguishing function | |
| US20210030390A1 (en) | Electronic stethoscope | |
| CN110353725A (en) | A kind of heart sound acquisition and analysis system and method based on cloud framework | |
| CN102144916A (en) | Multi-channel pulse signal detecting method and device capable of automatically regulating pressure | |
| CN201683910U (en) | Intelligent cardiopulmonary analyzing instrument | |
| CN101919704B (en) | Heart sound signal positioning and segmenting method | |
| CN203555751U (en) | Abdomen bowel sound analyzer system based on time-frequency analysis | |
| CN102715915A (en) | Portable heart sound automatic sorting assistant diagnostic apparatus | |
| CN202801659U (en) | Electronic stethoscope based on intelligent identification function | |
| CN113436726B (en) | Automatic lung pathological sound analysis method based on multi-task classification | |
| CN102512138A (en) | Cardiac sound monitoring and early warning method | |
| CN105496447B (en) | Electronic auscultation device with active noise reduction and auxiliary diagnosis function | |
| WO2020151169A1 (en) | Method for automatic removal of frictional sound interference of electronic stethoscope | |
| CN105943080A (en) | Intelligent stethophone | |
| Omarov et al. | Artificial Intelligence in Medicine: Real Time Electronic Stethoscope for Heart Diseases Detection. | |
| CN106037790A (en) | Wireless auscultation device | |
| CN202198616U (en) | Electronic stethoscope | |
| CN101862201B (en) | Method for representing cardiophonogram in intelligent cardiophone of mobile phone | |
| CN111588370A (en) | Heart signal integrated intelligent acquisition sensing system terminal, method and cloud control platform | |
| CN203988121U (en) | Electronics bluetooth stethoscope | |
| CN105997134A (en) | Multichannel lung sound signal collection system and method | |
| CN215349053U (en) | Congenital heart disease intelligent screening robot | |
| Behbahani | A hybrid algorithm for heart sounds segmentation based on phonocardiogram | |
| Wang et al. | Feature extraction based on mel-scaled wavelet transform for heart sound analysis | |
| CN104510492A (en) | Method and device for auscultating inaudible signals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CX01 | Expiry of patent term |
Granted publication date: 20101229 |
|
| CX01 | Expiry of patent term |