CN115474897A - Wearable intelligent monitoring and identification system for audio and non-audio vibration signs - Google Patents
Wearable intelligent monitoring and identification system for audio and non-audio vibration signs Download PDFInfo
- Publication number
- CN115474897A CN115474897A CN202110660885.8A CN202110660885A CN115474897A CN 115474897 A CN115474897 A CN 115474897A CN 202110660885 A CN202110660885 A CN 202110660885A CN 115474897 A CN115474897 A CN 115474897A
- Authority
- CN
- China
- Prior art keywords
- audio
- signals
- signal
- data
- vibration signals
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 21
- 201000010099 disease Diseases 0.000 claims abstract description 36
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 36
- 238000010801 machine learning Methods 0.000 claims abstract description 31
- 210000000056 organ Anatomy 0.000 claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 27
- 230000036541 health Effects 0.000 claims abstract description 14
- 230000002159 abnormal effect Effects 0.000 claims abstract description 8
- 230000036760 body temperature Effects 0.000 claims description 32
- 238000001914 filtration Methods 0.000 claims description 27
- 238000004458 analytical method Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 17
- 238000004891 communication Methods 0.000 claims description 15
- 230000001133 acceleration Effects 0.000 claims description 14
- 238000006243 chemical reaction Methods 0.000 claims description 13
- 230000005856 abnormality Effects 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 12
- 230000009467 reduction Effects 0.000 claims description 11
- 238000005520 cutting process Methods 0.000 claims description 9
- 230000008855 peristalsis Effects 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000007635 classification algorithm Methods 0.000 claims description 7
- 238000000513 principal component analysis Methods 0.000 claims description 5
- 230000007488 abnormal function Effects 0.000 claims description 2
- 239000011230 binding agent Substances 0.000 claims description 2
- 238000010224 classification analysis Methods 0.000 claims description 2
- 238000010230 functional analysis Methods 0.000 claims description 2
- 238000010009 beating Methods 0.000 claims 2
- 230000006870 function Effects 0.000 abstract description 20
- 238000003745 diagnosis Methods 0.000 description 34
- 238000012545 processing Methods 0.000 description 15
- 230000005236 sound signal Effects 0.000 description 15
- 238000000034 method Methods 0.000 description 13
- 210000001519 tissue Anatomy 0.000 description 13
- 239000000523 sample Substances 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 9
- 238000009826 distribution Methods 0.000 description 9
- 230000033001 locomotion Effects 0.000 description 8
- 238000005259 measurement Methods 0.000 description 8
- 230000035945 sensitivity Effects 0.000 description 8
- 239000013598 vector Substances 0.000 description 8
- 230000005540 biological transmission Effects 0.000 description 7
- 238000000354 decomposition reaction Methods 0.000 description 7
- 208000037656 Respiratory Sounds Diseases 0.000 description 6
- 238000002555 auscultation Methods 0.000 description 5
- 210000000038 chest Anatomy 0.000 description 5
- 230000001605 fetal effect Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000011160 research Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000013500 data storage Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000005315 distribution function Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 210000000683 abdominal cavity Anatomy 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 210000004204 blood vessel Anatomy 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 230000002496 gastric effect Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 208000031636 Body Temperature Changes Diseases 0.000 description 1
- 101100441424 Dictyostelium discoideum cupA gene Proteins 0.000 description 1
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 1
- 206010035664 Pneumonia Diseases 0.000 description 1
- 206010037660 Pyrexia Diseases 0.000 description 1
- 240000007651 Rubus glaucus Species 0.000 description 1
- 235000011034 Rubus glaucus Nutrition 0.000 description 1
- 235000009122 Rubus idaeus Nutrition 0.000 description 1
- 230000001154 acute effect Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 210000002458 fetal heart Anatomy 0.000 description 1
- 230000028709 inflammatory response Effects 0.000 description 1
- 230000000968 intestinal effect Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 229910052744 lithium Inorganic materials 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004118 muscle contraction Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 231100000915 pathological change Toxicity 0.000 description 1
- 230000036285 pathological change Effects 0.000 description 1
- 239000013610 patient sample Substances 0.000 description 1
- 230000035790 physiological processes and functions Effects 0.000 description 1
- 230000010349 pulsation Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 210000004888 thoracic abdominal cavity Anatomy 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000009385 viral infection Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/01—Measuring temperature of body parts ; Diagnostic temperature sensing, e.g. for malignant or inflamed tissue
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1101—Detecting tremor
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B7/00—Instruments for auscultation
- A61B7/02—Stethoscopes
- A61B7/04—Electric stethoscopes
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Physiology (AREA)
- Artificial Intelligence (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Dentistry (AREA)
- Acoustics & Sound (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The application discloses wearable audio frequency and non-audio frequency vibration sign's intelligent monitoring and identification system includes: the wearable device comprises a wearable device body, a detection device and a preset terminal, wherein the detection device and the preset terminal are arranged on the wearable device body, and the detection device is used for acquiring audio vibration signals in human tissues and organs or body cavities and low-frequency vibration signals in a non-audio range; the preset terminal analyzes the acquired audio vibration signals and low-frequency vibration signals in a non-audio range of the human tissue organ or the body cavity to obtain a characteristic data sequence, establishes a health and disease comprehensive characteristic database according to a multi-classification machine learning algorithm, and judges whether the audio vibration signals and the non-audio vibration signals in the human body cavity are abnormal or not so as to determine whether the related tissue organ has function and structure abnormity or not. Therefore, the problems of low efficiency, low accuracy, incapability of implementing long-time real-time monitoring and the like of a manual mode in the related art are solved.
Description
Technical Field
The application relates to the technical field of medical equipment, in particular to wearable health examination and clinical physics auxiliary diagnosis equipment.
Background
The biological vibration signals of human tissues, organs and body cavities contain rich medical information, and people mainly use audio vibration signals in the biological vibration signals to diagnose diseases at present, for example, the measurement of lung sound and heart sound signals is an important means commonly used by doctors for clinical medical disease diagnosis.
In the related art, a stethoscope is generally used as a diagnostic device, and an experienced doctor can determine some abnormal conditions of the lung through the stethoscope. However, the traditional manual operation method is often inefficient in signal acquisition, too much depends on the diagnosis experience of doctors during diagnosis, is prone to misdiagnosis, is poor in diagnosis accuracy, requires high time cost and medical cost for diagnosis, and cannot implement long-time real-time monitoring.
Content of application
The application provides a wearable intelligent monitoring and recognition system for audio and non-audio vibration (or local motion) signs, and aims to solve the problems that in the related technology, the manual operation mode is low in acquisition efficiency, poor in diagnosis accuracy, high in time cost and medical cost required by diagnosis and the like.
The embodiment of the application provides a wearable audio frequency and non-audio frequency vibration sign's intelligent monitoring and identification system, includes: the system comprises wearable equipment and a preset terminal; wherein, wearable equipment includes wearable equipment body and sets up the detection device on the wearable body, wherein, detection device includes: the signal acquisition module is used for acquiring audio vibration signals in tissues, organs and body cavities, low-frequency vibration (motion) signals in a non-audio frequency range and body temperature signals of a human body, and respectively converting the signals into first to third electric signals; the control module is used for controlling the signal acquisition module to acquire signals according to a control instruction; the communication module is used for sending the first electric signal, the second electric signal, the third electric signal and the fourth electric signal to a preset terminal; the preset terminal is used for sending the control instruction, carrying out intelligent diagnosis analysis according to the first to third electric signals to obtain a characteristic data sequence, and judging whether the audio and non-audio vibration/pressure signals in the human body cavity are abnormal or not according to a multi-classification machine learning algorithm so as to determine whether the related tissues and organs have abnormal functions and structures or not.
Optionally, the preset terminal is a server or a mobile terminal, so as to determine whether an organ in the human body cavity has a beat or a peristalsis abnormality, or a structural and functional abnormality according to the audio and non-audio vibration signals in the human body cavity.
Optionally, the preset terminal is a diagnostic device disposed on the body, so as to determine whether an organ in the body cavity has a beat or a peristalsis abnormality, or a structural and functional abnormality according to the audio and non-audio vibration signals in the body cavity.
Further, the method also comprises the following steps: and the database is used for storing the data detected by the detection device and the analysis result data of the preset terminal.
Further, the signal acquisition module comprises: the first acquisition circuit is used for acquiring audio vibration signals including human tissue organs or body cavities and converting the audio vibration signals into first electric signals; a second acquisition circuit for acquiring low frequency vibration signals in a non-audio frequency range including within a human tissue organ or a body cavity, and converting the non-audio frequency vibration signals into second electrical signals; and the body temperature acquisition circuit is used for acquiring body temperature signals of a human body and converting the body temperature signals into third electric signals.
Further, the first acquisition circuit comprises: the sound sensor is used for collecting audio vibration signals of human tissue organs or body cavities and converting the audio vibration signals into first current signals; a first amplifying circuit for amplifying the first current signal into a first voltage signal; the first filter circuit is used for performing direct current filtering and high-frequency filtering on the first voltage signal to obtain the first electric signal.
Further, the second acquisition circuit comprises: the acceleration sensor is used for acquiring non-audio vibration signals of human tissue organs or body cavities and converting the non-audio vibration signals into second current signals; the second amplifying circuit is used for amplifying the second current signal into a second voltage signal; and the second filter circuit is used for carrying out direct current filtering and high-frequency filtering on the second voltage signal so as to obtain the second electric signal.
Further, the body temperature acquisition circuit includes: the thermistor is used for acquiring a body temperature signal of a human body and converting the body temperature signal into a third current signal; a third amplifying circuit for amplifying the third current signal into a third voltage signal; the third filter circuit is used for carrying out direct current filtering and high-frequency filtering on the third voltage signal so as to obtain a third electric signal; and the preset terminal is also used for correcting the detection result according to the third electric signal.
Further, the detection device further comprises: and the analog-to-digital conversion circuit is used for converting the first electric signal, the second electric signal and the third electric signal into digital signals and sending the digital signals to the preset terminal through the communication module.
Further, the wearable device body comprises at least one wearable patch,Suction cupAt least one camisole, harness, binder, or a universal strap configuration.
Further, the intelligent diagnosis analysis technology used by the preset terminal comprises: the data preprocessing algorithm is used for extracting, sorting, classifying, filtering, cutting and normalizing the collected data; the machine learning algorithm training model is used for completing the feature extraction of the collected data by using a feature extraction algorithm, respectively establishing a health and comprehensive disease feature database and completing the training of sample database data; (ii) a And the machine learning inference classification algorithm is used for carrying out classification result analysis on the data to be detected by using a machine learning model.
Further, the data preprocessing algorithm comprises: the data preprocessing and cutting algorithm is used for filtering, cutting and standardizing the preprocessed data by using functional analysis; the machine learning algorithm training model comprises: performing feature extraction on data to be processed by using orthogonal basis fitting, PCA (principal component analysis) and nonlinear dimension reduction (such as UMAP); respectively establishing health and comprehensive disease characteristic databases; performing model training on the extracted sample library data feature set by using statistical learning and machine learning; the machine learning inference classification algorithm comprises: and (4) carrying out result classification analysis on the data to be detected by using classification algorithms such as KNN, SVM, naive Bayes and softmax neural networks.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a block diagram of a wearable intelligent monitoring and recognition system for audio and non-audio vibration (or local motion) signs according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a basic structure of a personal vest provided according to an embodiment of the application;
FIG. 3 is a flowchart of processing signals of a body cavity and local organs and tissues according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a wireless receiving circuit according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an audio signal acquisition circuit provided in an embodiment of the present application;
FIG. 6 is a schematic diagram of a frequency detection simulation of an audio signal acquisition circuit according to an embodiment of the present application;
FIG. 7 is a schematic diagram of a non-audio signal acquisition circuit according to an embodiment of the present disclosure;
FIG. 8 is a schematic diagram of a signal frequency detection simulation of a non-audio signal acquisition circuit according to an embodiment of the present application
Fig. 9 is a schematic structural diagram of a body temperature signal measuring circuit provided in an embodiment of the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative and intended to explain the present application and should not be construed as limiting the present application.
The present application is based on the recognition and discovery by the inventors of the following problems:
in the background art, the traditional manual auscultation method has the defects of low signal acquisition efficiency, limited acquisition method, incapability of continuous real-time monitoring, high dependence on doctor diagnosis experience and few diseases to be identified. Moreover, the time cost and medical cost of manual diagnosis are high, and it is difficult to prevent some diseases with rapid onset and long latency, which is easy to cause the loss of life health of patients and the waste of medical resources due to misdiagnosis.
In addition, changes in the internal acoustic and/or non-acoustic vibration signals may often precede the body's own discomfort, as studies have shown that patients with partially asymptomatic new coronary pneumonia virus infections have varying degrees of lung sound abnormalities. In young, old and weak populations, changes in lung sounds are often predictive of acute structural pathological changes (e.g., inflammatory responses) and concomitant increases in body temperature (fever). The effective use of non-audio vibration or pressure/tension signals originating from the body cavity is an unsolved problem in addition to the sound signals, and therefore, it is necessary to develop a more portable chest auscultation and chest vibration signal measuring apparatus for more accurate extraction of body cavity signals, and with the development of the current artificial intelligence technology, signal extraction and disease diagnosis and development of auxiliary treatment by means of more advanced data processing modes and machine learning models based on the artificial intelligence technology are undoubtedly developing in the future.
At present, a plurality of auscultation measuring devices exist in professional diagnosis and treatment places such as hospitals, however, the auscultation measuring devices only play a role in measuring body cavity sound signals to assist doctors in diagnosis, and do not have functions of efficient analysis and automatic diagnosis. In addition, all auscultation devices so far can only detect audio signals in a certain frequency range, and cannot recognize and collect vibration or internal tension/pressure signals generated by low-frequency vibration outside various audio frequency ranges and slow movement of local tissues and organs. Various low-frequency vibrations or motion signals of local tissues and organs (such as pulse, muscle contraction, gastrointestinal peristalsis, fetal movement and the like) are closely related to the health and disease states of the body. Therefore, with the innovation and popularization of wearable equipment, an intelligent medical system integrating high-precision measurement, intelligent diagnosis and disease management is researched and developed, daily early warning and diagnosis of multiple chest diseases of a user are realized through intelligent data acquisition, real-time monitoring and data analysis, auxiliary diagnosis service is provided for professional medical places, and therefore the medical cost, time and labor cost of medical treatment are greatly reduced, and the development trend of future intelligent medical treatment is achieved.
At present, some groups in the scientific research community develop a series of researches based on deep learning technology aiming at body cavity signals, and obtain some preliminary results aiming at intelligent diagnosis of specific diseases. However, compared with the statistical learning model, the deep learning technology has lower interpretability, is difficult to study the cross-correlation study of various diseases, and has lower adaptability to the auxiliary diagnosis function expected by doctors. The statistical learning model can deeply research the correlation between specific medical characteristic features and diseases, effectively assist a doctor in diagnosing and improve the understanding of the doctor on the medical characteristics.
The wearable intelligent audio and non-audio vibration sign monitoring and identification system of the embodiment of the application is described below with reference to the accompanying drawings. Aiming at the problems that the collection efficiency of a manual operation mode number in the related technology mentioned in the background technology center is low, the diagnosis accuracy is poor, the time cost and the medical cost required by diagnosis are high, and the like, the method provides the wearable intelligent monitoring and recognition system for the audio and non-audio vibration signs. Therefore, the problems that manual operation mode numbers in the related technology are low in collection efficiency and poor in diagnosis accuracy, time cost and medical cost required by diagnosis are high and the like are solved.
Specifically, fig. 1 is a block schematic diagram of a wearable intelligent monitoring and recognition system for audio and non-audio vibration signs according to an embodiment of the present application.
As shown in fig. 1, the wearable intelligent monitoring and recognition system 10 for audio and non-audio vibration signs includes: the wearable device 100 and the preset terminal 200; the wearable device 100 includes a wearable device body 110 and a detection apparatus 120.
In this embodiment, the wearable device body 110 may include at least one wearable patch,Suction cupA harness or at least one camisole, which can be in close proximity to the body and which can carry multiple test devices 120 simultaneously. The wearable device body 110 may be provided with any type and number of detecting devices 120 according to actual applications, which is not limited in particular.
A wearable patch,Suction cupThe arrangement of the straps and the vest can be carried out by those skilled in the art according to the actual situation, and is not limited in particular. Taking the vest as an example, as shown in fig. 2, the vest comprises a front chest and a back chest, and each part can be respectively provided with 1 to N detection devices according to specific medical requirements120,N is a positive integer greater than 1.
It should be noted that the camisole of the embodiment of the present application has the characteristics of flexibility, elasticity, and being capable of being closely attached to a human body, and the setting positions of the camisole and the detection device 120 may be specifically set according to the characteristics of the user and the medical requirements, and are not specifically limited.
In the present embodiment, the detection device 120 includes: the device comprises a signal acquisition module, a control module and a communication module. The signal acquisition module is used for acquiring an audio vibration signal in a human body cavity, a low-frequency vibration/local motion signal in a non-audio frequency range and a human body temperature signal and respectively converting the signals into first to third electric signals; the control module is used for controlling the signal acquisition module to acquire signals according to the control instruction; the communication module is used for sending the first electric signal, the second electric signal, the third electric signal and the fourth electric signal to a preset terminal.
Wherein the body cavity may include thoracic cavity, abdominal cavity, and uterine cavity of pregnant woman. The audio vibration signal may be a sound signal generated by vibration of human breath, heartbeat, bowel sounds in abdominal cavity, fetal heartbeat in a pregnant woman, and the like, the non-audio vibration signal may be a biological vibration signal generated by low-frequency vibration, and the vibration frequency generated by the audio vibration signal is generally higher than the vibration frequency generated by the non-audio vibration signal, so in the embodiment of the present application, the frequency acquisition range of the non-audio vibration signal sensor may be 0.1Hz to 100Hz, and the frequency acquisition range of the audio vibration signal sensor may be 20Hz to 50kHz.
It is understood that the communication and central control module: a lithium battery is used as power to provide power for the signal acquisition and communication module; the central control module comprises ports connected with the plurality of signal acquisition modules in parallel, a central processing unit for processing multichannel input signals in parallel, wireless connection ports for sending signals to preset terminals in a Bluetooth/WIFI (wireless fidelity) mode and wired connection ports such as a USB (universal serial bus).
In the following embodiments, by taking a non-integrated configuration as an example, the communication module may be a module having a wireless transmission function, such as a WIFI wireless transmission module, a bluetooth wireless transmission module, and the preset terminal may also correspondingly have a corresponding wireless receiving device, or be connected to a wireless receiving device having a wireless receiving function, as shown in fig. 3, the wireless receiving device may include a wireless receiving circuit, a micro data storage and processing unit, and may be a micro data storage and processing chip having a wireless receiving capability, a digital signal processing and temporary storage capability, a digital signal output capability, and a digital signal display output capability, where as shown in fig. 4, the wireless receiving circuit may send a digital signal to the preset terminal in a wired or wireless manner, or may be directly connected to a display for display output; the part in the black frame in fig. 3 can be replaced by a single chip with a bluetooth function, such as raspberry pi 4B, a mobile phone, a computer, etc.; in fig. 4, P0_0, P0_1, P2_2 are output terminals of the digital signal, which can be accessed to the micro data storage and processing unit.
Therefore, the audio vibration signals and the non-audio vibration signals at different positions can be respectively measured in real time, and the signals are wirelessly sent to the preset terminal through the communication module so as to perform data preprocessing and machine learning model calculation and then output diagnosis results, wherein the data preprocessing and the machine learning model are elaborated in detail below and are not set forth more than necessary.
When the communication module is a wireless transmission circuit, the same circuit configuration as that of the wireless reception circuit shown in fig. 4 may be adopted, and of course, other types of circuit configurations may be adopted, which is not particularly limited.
In this embodiment, the preset terminal is configured to send a control instruction, analyze the first to third electrical signals to obtain a feature data sequence, calculate the feature data sequence, and determine whether the audio and non-audio vibration signals in the body cavity of the human body are abnormal according to a multi-classification machine learning algorithm, so as to determine whether the related tissue and organ have functional and structural abnormalities.
The preset terminal 200 is used for controlling the operation of the whole intelligent system, selecting functions according to requirements, transmitting collected data information to a cloud system, and returning and displaying an intelligent analysis result. The artificial intelligence algorithm software system is used for respectively extracting characteristic data sequences of various signals (such as a first electric signal and a second electric signal) acquired by the signal acquisition module, calculating the data sequences, judging whether audio and non-audio vibration signals from the body cavity of a human body are abnormal or not according to a multi-classification machine learning algorithm (such as neural network softmax judgment, k-neighbor method, naive Bayes, decision trees and the like), and further judging whether the function and the structure of related tissues and organs are abnormal or not, so that a basis is provided for daily health monitoring and clinical disease diagnosis.
In some embodiments, the preset terminal 200 may be integrally disposed on the detecting device 120, for example, the preset terminal 200 may be a diagnostic device disposed on the body to determine whether there is a beat or a peristalsis abnormality, or a structural and functional abnormality of an organ in the body cavity according to the audio and non-audio vibration signals in the body cavity. The diagnostor may be a processor with an arithmetic processing function, and may be selected according to actual design requirements without specific limitations.
In some embodiments, the default terminal 200 may also be disposed on the detection device 120 without being integrated, so as to simplify the structure of the detection device 120 and improve the convenience thereof, for example, the default terminal 200 may be a server or a mobile terminal, so as to determine whether there is a beat or a peristalsis abnormality, or a structural and functional abnormality in an organ in the body cavity according to the audio and non-audio vibration signals in the body cavity.
The preset terminal is a display (oscilloscope), a mobile terminal (a smart phone, a mobile computer terminal), a personal computer, a server and the like, and the server runs faster and has higher load than a common computer, is used for providing calculation or service, and can be a cloud server and the like. The mobile terminal may be a mobile terminal of an IOS operating system (IOS is a handheld device operating system developed by apple, inc.), an Android operating system (the Android system is an operating system based on a Linux free and open source code), and a Windows Phone operating system (Windows Phone is a mobile Phone operating system issued by microsoft), and certainly may also be a personal computer and other intelligent mobile terminals, which is not limited in the present application. It should be understood that, in the embodiment of the present application, the mobile terminal may be a hardware device such as a mobile phone, a tablet computer, or the like, which has various operating systems.
In some embodiments, the system 10 of embodiments of the present application further comprises: and the database is used for storing the data detected by the detection device and the analysis result data of the preset terminal. The database can be a block chain distributed cloud database system and is used for storing related data information; the cloud computing system running the intelligent analysis algorithm module realizes analysis, recognition, analysis reporting, early warning output, information storage, processing and the like of various audios and non-audios from body cavities of the human body.
In some embodiments, the signal acquisition module comprises: the temperature acquisition device comprises a first acquisition circuit, a second acquisition circuit and a body temperature acquisition circuit. The first acquisition circuit is used for acquiring audio vibration signals in a human body cavity comprising a body surface blood vessel cavity and converting the audio vibration signals into first electric signals; the second acquisition circuit is used for acquiring low-frequency vibration signals in a non-audio frequency range in a human body cavity comprising a body surface blood vessel cavity and converting the non-audio frequency vibration signals into second electric signals; the body temperature acquisition circuit is used for acquiring a body temperature signal of a human body and converting the body temperature signal into a third electric signal.
In this embodiment, the first collecting circuit may be a circuit of a high-precision wireless electronic stethoscope, and the first collecting circuit includes: the sound sensor, first amplifier circuit and first filter circuit. The acceleration sensor is used for collecting audio vibration signals in a human body cavity and converting the audio vibration signals into first current signals; the first amplifying circuit is used for amplifying the first current signal into a first voltage signal; the first filter circuit is used for carrying out direct current filtering and high-frequency filtering on the first voltage signal so as to obtain a first electric signal.
It is understood that the wireless electronic stethoscope may include: the body cavity audio vibration signal processing circuit comprises a sound sensor, an amplifying circuit and a control circuit, wherein the sound sensor is used for measuring a body cavity audio vibration signal and converting the body cavity audio vibration signal into an electric signal; and the filter circuit is connected with the amplifying circuit and filters the direct current and high frequency signals of the current signal. Wherein the sound measuring sensor may include: various acoustic capacitive sensors for converting acoustic signals with sampling frequency range of 10-100kHz into current signals
For example, as shown in fig. 5, the acoustic sensor may use a high-precision acoustic capacitive sensor 7800725P (or other similar sensors) manufactured by european RS Components, and the specific parameters are: sensitivity: -44dB; frequency range: 50-16000Hz; standard operating voltage: 2V; maximum power supply current: 0.5mA; size: 4X 1.5mm; working temperature range: -20-50 ℃. The sound sensor can collect sound signals of a human body cavity at multiple points, convert the sound signals into current signals and input the current signals into a circuit.
The current signal generated by the conversion of the sound sensor can enter an amplifying circuit for signal amplification, and the embodiment of the application can use a low noise operational amplifier AD8014 (or other similar sensors) manufactured by Analog Devices, so that after passing through a double amplifying circuit as shown in FIG. 3, the induced current of 1 muA-100 muA generated by the sound sensor can be amplified into a voltage signal of 10m-1000 mV.
The wireless electronic stethoscope circuit has extremely strong stability and accuracy, voltage signals generated by the amplifying circuit need to be filtered by the filter circuit to remove direct current and high-frequency signals, interference of high-frequency noise in an external environment can be eliminated, the detectable analog frequency range is shown in figure 6, and the circuit can stably acquire and amplify 10Hz-50kHz audio vibration signals. The vibration frequency range of the known heart sound is
The frequency range of intestinal sounds is
And the frequency range of the lung sounds is
Therefore, compared with the traditional lung sound signal analysis category of 100Hz-2000Hz, the embodiment of the application canThe system is effectively suitable for signal samples required by medical analysis of the body cavity, and has wider signal acquisition range and stronger applicability.
In this embodiment, the second acquisition circuit may be a circuit of a high-precision biomechanical vibration signal collector, the second acquisition circuit including: acceleration sensor, second amplifier circuit and second filter circuit. The acceleration sensor is used for collecting non-audio vibration signals in a human body cavity and converting the non-audio vibration signals into second current signals; the second amplifying circuit is used for amplifying the second current signal into a second voltage signal; and the second filter circuit is used for carrying out direct current filtering and high-frequency filtering on the second voltage signal so as to obtain a second electric signal.
It is understood that the biomechanical vibration signal collector may comprise: the acceleration sensor is used for measuring the biological vibration signals of the tissue organ and the body cavity and converting the signals into electric signals; the amplifying circuit is connected with the acceleration sensor and can sample and amplify the current signal measured by the acceleration sensor for multiple times; and the filter circuit is connected with the amplifying circuit and filters the direct current and high frequency signals of the current signal. Wherein, the acceleration sensor may include: and various acceleration sensors or pressure-sensitive sensors for converting the biological vibration signals with the sampling frequency range of 0.1-100Hz into current signals.
For example, as shown in fig. 7, the acceleration Sensor used in the biomechanical Vibration signal detector circuit is a minisensor 100Vibration Sensor microsensor (or other similar Sensor) manufactured by TE Connectivity's (TE) Measurement Specialties, and its specific parameters are: voltage sensitivity: 1.1V/g; resonance frequency: 75Hz; resonance frequency voltage sensitivity: 6V/g; weight: 0.6g; working temperature range: -20 to-60 ℃. The acceleration sensor can carry out multi-point acquisition on biological vibration signals of a human body cavity, convert the biological vibration signals into current signals and input the current signals into the circuit.
The current signal generated by the acceleration sensor conversion can enter an amplifying circuit for signal amplification, and the embodiment of the application also uses a low-noise operational amplifier AD8014 produced by Analog Devices company, and after the direct current and high-frequency signals are filtered by a double amplifying circuit and a filter circuit shown in FIG. 3, the induced current of 1-100 muA generated by the sensor can be amplified into a voltage signal of 10-2000 mV.
The biological vibration signal device measuring circuit has extremely strong stability and accuracy, the detectable analog frequency range is shown in fig. 8, the circuit can stably acquire and amplify non-audio vibration signals of 0.1-100Hz, and the biological vibration signal device measuring circuit is suitable for analyzing mechanical signals generated in a human body cavity.
In this embodiment, the body temperature acquisition circuit is configured to acquire a body temperature signal of a human body and convert the body temperature signal into a third electrical signal; the preset terminal is also used for correcting the detection result according to the third electric signal. Wherein, body temperature acquisition circuit includes: the temperature sensor comprises a thermistor, a third amplifying circuit and a third filtering circuit, wherein the thermistor is used for acquiring a body temperature signal of a human body and converting the body temperature signal into a third current signal, and the sensitivity of the thermistor is less than 0.01 ℃; a third amplifying circuit for amplifying the third current signal into a third voltage signal; and the third filter circuit is used for carrying out direct current filtering and high-frequency filtering on the third voltage signal so as to obtain a third electric signal.
It is understood that the body temperature acquisition circuit may be a circuit of a wireless body temperature detector, including: a thermistor for measuring body temperature and converting into an electric signal; the amplifying circuit is connected with the thermistor and can sample and amplify the current signal measured by the thermistor; and the filter circuit is connected with the amplifying circuit. Wherein, the thermistor can include: the sensitivity of the thermistor, which can efficiently convert the body temperature change into a circuit signal, needs to be lower than 0.01 ℃.
For example, as shown in fig. 9, the wireless body temperature detector circuit can perform multi-point acquisition on the temperature change of the body cavity of the human body, convert the temperature change into a digital signal, and transmit the digital signal together with the vibration digital signal and the sound digital signal to a subsequent receiving device. In addition, the embodiment of the application can also be provided with a temperature-sensitive sensor for measuring the surface temperature signal of the skin outside the body cavity, so that the real-time digital signal of the surface temperature can be sent to a subsequent receiving device.
It should be noted that the amplifying circuits in the above embodiments may include: various amplifying circuits for amplifying the current signals of 1-100 muA into voltage signals of 50-5000 mV.
In some embodiments, the detection device 120 further comprises: an analog-to-digital conversion circuit. The analog-to-digital conversion circuit is configured to convert the first electrical signal, the second electrical signal, and the third electrical signal into digital signals, and send the digital signals to the preset terminal 200 through the communication module.
The analog-to-digital conversion circuit may be integrated with the communication module, or may not be integrated with the communication module, which is not limited specifically. Taking the integrated configuration as an example, the module for analog-to-digital conversion and wireless transmission is a microchip with integrated analog-to-digital conversion capability, temporary digital signal storage capability and wireless transmission capability, and the microchip can be further provided with a network communication module directly connected with a network to transmit the digital signal to a preset terminal.
For example, the analog-to-digital conversion circuit is connected to the output end of the filter circuit, and a series of chips with analog-to-digital conversion function and wireless transmission function can be used in the embodiment of the present application to convert the circuit signal into a digital signal and transmit the digital signal to the preset terminal. As shown in FIG. 7, an ADCMIC10 chip provided in Cadence software is used, and the chip has a 10-bit A/D, which can realize the digital conversion of the body cavity sound signal. The chip that can be adopted in the embodiment of the present application can include Dialog DA1469x, apollo3 Blue, TI MSP430F5528, TI CC2540F256, etc., and the like. The measured data is transmitted to the receiving device at the frequencies f1, f2, a.
The wearable audio frequency and non-audio frequency vibration sign's intelligent monitoring and identification system of embodiment of this application, collect data signal through the circuit and can propagate with wireless form, can realize high-efficient accurate measurement to the body cavity signal, control and medical diagnosis, wherein, biomechanics vibration signal detector that equipment includes and wireless electron stethoscope can realize to the low frequency vibration signal and the breathing that originate from human body cavity, the heartbeat, the internal bowel sound of abdominal cavity and the internal foetus heartbeat digitization of pregnant woman gather, wherein, mechanics vibration signal measurement circuit keeps high accuracy at 0.1Hz-100Hz to measure, sound signal measurement circuit keeps high accuracy at 100-50kHz to measure, almost all biological vibration signal's frequency range has been covered, the data analysis who provides the data sample of adaptation for follow-up full frequency channel. From this, the equipment of this application embodiment has advantages such as high practicality, high accuracy, wide, the suitability is strong of collection scope.
In the embodiment of the application, the preset terminal can realize efficient and accurate measurement, monitoring and medical diagnosis of the body cavity signals through a data processing algorithm and an intelligent diagnosis technology based on a machine learning algorithm, wherein the intelligent diagnosis technology comprises three parts, namely data preprocessing, machine learning algorithm training and machine learning inference classification; the data preprocessing algorithm finishes extraction, sorting, classification, filtering and cutting of collected data, the machine learning algorithm training finishes feature extraction of the data and training of sample library data, and the machine learning reasoning classification finishes result analysis of the data to be detected, and the method is as follows:
1. data preprocessing algorithm
(1) Data file extraction: reading in a digital signal file, decoding the digital signal file in a corresponding format and converting the digital signal file into an easily-read format data file, such as csv or txt;
(2) Data file arrangement: classifying the data files according to the corresponding labels (time, ID, parts, symptoms) and the sensor serial numbers;
(3) And (3) filtering treatment: since the biomechanical vibration signal and the acoustic signal are non-stationary and non-linear weak signals and can be influenced by noise such as baseline drift, power frequency, electromyographic interference and the like in the acquisition process, the acceleration sensor is insensitive to the signal larger than 200Hz and the acoustic sensor has a large sensitivity range, so that the mechanical vibration signal and the acoustic signal need to be subjected to low-frequency filtering and high-frequency filtering respectively.
The embodiment of the application adopts a wavelet transform algorithm capable of carrying out local analysis on the time sequence signal to filter the collected vibration signal. For a continuous time sequence signal Φ (t), the wavelet transform thereof is:
wherein, a and b are translation and scale parameters respectively.
Discrete timing signal phi (t) for analog acquisition of continuous timing signal phi (t) k ) (k =1,2.. N), which is reconstructed according to Mallat multiscale analysis algorithm expressed as:
the orthogonal wavelet transform coefficients are:
where i is the number of layers of decomposition, H, G are wavelet decomposition high-pass and low-pass filters in the time domain, respectively, A i 、D i Are wavelet coefficients.
The wavelet decomposition is carried out on the acquired data signals, the decomposition scales of baseline drift, power frequency and electromyographic noise centralized distribution are filtered out through threshold values, and filtered data are obtained through reduction and reconstruction. The data is sorted into a readable format, such as csv or txt, according to the m-type data tag.
(4) Data preprocessing and cutting: respectively extracting the time series signals in each data file and carrying out normalized processing, and the steps are as follows:
(4.1) setting the length of the time series data to be L, calculating second-order difference of the series signals with the sequence numbers between 0.1 and 0.9 to determine the minimum value, defining a threshold (example 70%), and taking out all fragments with the second-order difference smaller than the minimum value multiplied by the threshold;
(4.2) aiming at the fragments obtained in the step (4.1), calculating the median of the corresponding sequence signals, and returning to a median corresponding unit, namely a maximum value unit with representative significance;
(4.3) cutting the maximum unit with representative meaning followed by 0.1 × l units as normalized time series (vector of 0.1 × l × 1);
and (4.4) respectively storing all the data files into various types of general files according to the labels, and inputting the general files into a subsequent machine learning model.
2. Further, the machine learning model training includes the following steps of feature extraction and model training: wherein the feature extraction is realized by the following steps: performing data dimensionality reduction on the time series data input in the step 1, and performing three dimensionality reduction processes on the 0.1 × l × 1 dimensionality vector obtained in the step 1:
(1) Fitting and extracting coefficients of the orthonormal basis: for one-dimensional time series signals, fourier bases, wavelet bases, haar bases and the like can be selected;
(1.1) the method in which coefficient extraction is performed using a Fourier base or a Haar base is: for the input time series data, a linear combination of Fourier bases or Haar bases closest to the normalized data is calculated by the least squares method, and the coefficients of the combination are output as a feature representation of the functional data (a vector of M1, where M is the number of defined bases), and the formula is as follows (taking 300 Fourier bases as an example):
the functional data processing process comprises the following specific steps (taking Fourier basis as an example):
(a) And (3) extracting the features of the data: respectively carrying out functional data conversion on the normalized time sequence corresponding to each label: a fixed time interval [0, N ] is set, and p orthogonal canonical Fourier bases over the interval are determined.
(b) Calculating a linear combination of Fourier bases closest to the normalized data by a least square method, and outputting coefficients of the combination as a feature representation (vector of p 1) of the functional data;
(c) Storing the coefficients of the functional data as a feature matrix as a patient feature database (matrix of p m), wherein m is the total amount of patient sample data;
(d) Analysis of new enrollment data: and repeating the steps to obtain the characteristic data sequence of the new recorded data.
(1.2) the method in which the coefficient extraction is performed using wavelet basis is: the acquired discrete timing signal Φ (tk) (k =1, 2.. Was, n) was simulated for the continuous timing signal Φ (t), which was reconstructed according to the Mallat multiscale analysis algorithm as:
the orthogonal wavelet transform coefficients are:
wherein i is the number of layers of decomposition, H and G are wavelet decomposition high-pass and low-pass filters in the time domain, and Ai and Di are wavelet transformation coefficients. Extracting and storing the wavelet transform coefficients as characteristic wavelet coefficients of the input sequence data;
(2) Principal factor analysis PCA: and (3) calculating a data covariance matrix in the S4, performing singular value decomposition, taking unit orthogonal eigenvectors corresponding to the largest m largest eigenvalues, and performing inner product on the data and the eigenvectors to obtain a main factor, wherein the formula is as follows:
α i (x)＝<x，η i >
wherein x is original data and is a unit characteristic vector corresponding to the ith maximum characteristic value;
(3) Nonlinear dimension reduction: such as UMAP dimension reduction, etc.;
3. model training, including the following two algorithms: statistical learning and machine learning;
(1) Wherein, the statistical learning aims at providing prior probability distribution of the data subjected to dimension reduction of step 2 for each label, wherein x is the data, and y is the label. Under the naive bayes framework, each component is assumed to be independently distributed. At each fixed label, a distribution function is given using a non-parametric kernel estimate for each data one-dimensional component, here a gaussian kernel estimate, with the formula:
wherein h represents the kernel function preset standard deviation (the smaller the obtained data is, the closer the obtained data is to the real distribution under the condition of more data quantity), K is a Gaussian function, and x i＝1,2,…,n Known as a one-dimensional component. The prior joint distribution under the label is given by the product of independent distribution of each component, and then similar operation is carried out on the other labels to obtain the prior distribution under all the labels;
(2) Where machine learning aims to give a prediction function f (x) from the dimension reduced data to the label, where x is the data. The models to be selected are XGboost based on decision trees, random forests and LSTM algorithm based on neural networks.
4. Classification algorithms such as KNN, SVM, naive bayes, softmax neural networks, and the like.
(1) KNN: calculating the average of the minimum K Euclidean distances from the new sequence data to be detected to each label sample library, and returning the sample library serial number corresponding to the minimum value;
(2) SVM (for two classes only): linear separation of any two sample pools was performed using the cost function of Soft-margin, with the optimization formula as follows (labeling two diseases numbered 0 and 1-1 and sample pool size m0 and m1, respectively):
wherein x i 、y i Respectively, the feature vectors and labels in the library, and b and lambda are adjustable parameters. New data calculation wx-b, if it is greater than or equal to 1, it is disease No. 1, and if it is less than or equal to-1, it is disease No. 1Disease No. 0.
(3) Naive bayes: the posterior probability based on the Bayesian formula can be given according to the prior probability distribution of the tag data. And inquiring regional or national statistical data to obtain the distribution probability p (z) of each specific population z in the total population. The above dimension reduction and feature extraction are performed on the sequence data signals to be diagnosed to obtain dimension reduction data x _ obs, and then the probability of y disease can be given by Bayes formula:
more specifically, the foregoing bayesian formula can be implemented by the following steps: and if the database matrix corresponding to the disease label i is extracted, giving a distribution function of each row through a kernel function. The newly measured data is x, each j component of x is found out, and the nearest quantity y of j rows in the disease sample library Mi is found out ij And obtaining the corresponding distribution function value as the prior distribution P (y) ij I). If the total number of diseases is L (including health), the probability that a person with the sequence signal corresponding to the coefficient vector x has disease i can be finally estimated as:
wherein P (i) is the incidence of disease i, and is generally given in the literature or statistically.
If one considers the division of the sequence data of a person into a plurality of medical features w, i.e. where m is the length of the coefficient vector corresponding to each medical feature. According to the calculation of the disease probability of the single medical characteristic, a probability correlation matrix of the multiple medical characteristics and the multiple disease probabilities can be obtained.
(4) Softmax neural network: and training at least three layers of hidden neural networks by using the characteristics as input and the corresponding labels as output. The first layer is a rigid transformation layer, the second layer is a ReLu activation or logistic regression activation layer, the third layer is a softmax function layer, and finally the third layer is a probability output layer. Wherein the weight of the rigid transformation has the mostA small MSE loss results. The MSE loss function of a general neural network is defined as follows, where x j 、y j M is the total amount of data of all libraries, N is the unit number of a hidden layer, and w, a and b are parameters to be optimized;
performing the above dimension reduction feature extraction on the sequence data signals to be diagnosed to obtain dimension reduction data x _ obs, calculating fz (x _ obs) by using a prediction function, and then giving the posterior probability by a softmax formula:
in summary, the embodiment of the present application can measure the fixed-point design of the body cavity signal for different specific diseases, and can use the corresponding sensor fixed-point measurement design according to the disease study object; hardware equipment for measuring body cavity sound and vibration signals with high precision and high sensitivity is designed, so that the body cavity low-frequency vibration signals and medium-high frequency sound signals can be measured with high sensitivity, and noise interference can be effectively shielded; based on the signal processing of various machine learning models, various body cavity signals can be extracted and classified efficiently, specific diseases can be diagnosed and analyzed, and cross correlation research can be carried out on various diseases and signal characteristics.
Therefore, the embodiment of the application can provide a brand-new and effective solution for the defects of low efficiency, low accuracy, incapability of implementing long-time real-time monitoring and the like of the traditional manual mode. The intelligent monitoring and identifying system can intelligently analyze the in-cavity audio frequency and non-audio frequency vibration signals of people with different ages, sexes and physical characteristics, including breathing sounds, heart sounds, heartbeat, gastrointestinal peristalsis, bowel sounds, fetal heart sounds (fetal sounds), fetal peristalsis (fetal movement), body surface arterial pulsation (pulse) and body surface temperature, thereby realizing real-time discrimination and early warning of physiological function abnormal changes of relevant organs of organisms, providing timely, quick and accurate health warning information of relevant users for users, nurses or clinical medical staff, and simultaneously providing direct support for clinical diagnosis.
According to the wearable intelligent monitoring and identification system of audio frequency and non-audio frequency vibration sign that this application embodiment provided, can utilize real-time accurate measuring audio frequency and non-audio frequency vibration signal to carry out automated diagnosis, the efficient of automatic acquisition signal, and automatic diagnosis need not too rely on doctor's experience, can effectively improve diagnostic efficiency and accuracy, and effectively reduce diagnostic time and medical cost, also can improve the understanding of doctor to medical characteristics when effectively assisting doctor's diagnosis, it is very convenient when wearable equipment uses simultaneously, effectively improve doctor's use and experience.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Moreover, various embodiments or examples and features of various embodiments or examples described in this specification can be combined and combined by one skilled in the art without being mutually inconsistent.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "N" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.
Claims (10)
1. The utility model provides a wearable audio frequency and intelligent monitoring and identification system of non-audio frequency vibration sign which characterized in that includes: the system comprises wearable equipment and a preset terminal; wherein, the first and the second end of the pipe are connected with each other,
wearable equipment includes wearable equipment body and sets up the detection device on wearable body, wherein, detection device includes:
the signal acquisition module is used for acquiring audio vibration signals, low-frequency vibration signals or pressure signals in a non-audio frequency range and body temperature signals of a human body in a human tissue organ or a body cavity, and respectively converting the audio vibration signals, the low-frequency vibration signals or the pressure signals and the body temperature signals into first to third electric signals;
the control module is used for controlling the signal acquisition module to acquire signals according to a control instruction;
the communication module is used for sending the first to third electric signals to a preset terminal;
and the preset terminal is used for sending the control instruction, analyzing according to the first to third electric signals to obtain a characteristic data sequence, establishing a comprehensive health and disease characteristic database according to a multi-classification machine learning algorithm, and judging whether the audio and non-audio vibration signals in the tested human tissue organ or the body cavity are abnormal or not so as to determine whether the related tissue organ has abnormal functions and structures or not.
2. The apparatus of claim 1,
the preset terminal is a server or a mobile terminal, and is used for determining whether the tissue organ has beating or peristalsis abnormity or structural and functional abnormity according to the audio and non-audio vibration signals in the human tissue organ or the body cavity;
the preset terminal is a diagnotor arranged on the body and is used for determining whether the related tissue organ has beating or peristalsis abnormality or structural and functional abnormality according to the audio and non-audio vibration signals in the human tissue organ or the body cavity.
3. The apparatus of claim 1, wherein the signal acquisition module comprises:
the first acquisition circuit is used for acquiring audio vibration signals including various tissues, organs and body cavities and converting the audio vibration signals into first electric signals;
the second acquisition circuit is used for acquiring low-frequency vibration and pressure signals in a non-audio frequency range including tissue organs and body cavities, and converting the non-audio frequency vibration signals into second electric signals;
the body temperature acquisition circuit is used for acquiring a body temperature signal of a human body and converting the body temperature signal into a third electric signal.
4. The apparatus of claim 3, wherein the first acquisition circuit comprises:
the sound sensor is used for collecting audio vibration signals in the human body cavity and converting the audio vibration signals into first current signals;
a first amplifying circuit for amplifying the first current signal into a first voltage signal;
the first filter circuit is used for carrying out direct current filtering and high-frequency filtering on the first voltage signal to obtain a first electric signal.
5. The apparatus of claim 4, wherein the second acquisition circuit comprises:
the acceleration sensor is used for collecting non-audio vibration signals in tissues, organs and body cavities and converting the non-audio vibration signals into second current signals;
the second amplifying circuit is used for amplifying the second current signal into a second voltage signal;
and the second filter circuit is used for carrying out direct current filtering and high-frequency filtering on the second voltage signal so as to obtain the second electric signal.
6. The device of claim 4, wherein the body temperature acquisition circuit comprises:
the thermistor is used for acquiring a body temperature signal of a human body and converting the body temperature signal into a third current signal;
a third amplifying circuit for amplifying the third current signal into a third voltage signal;
a third filter circuit, configured to perform dc filtering and high-frequency filtering on the third voltage signal to obtain the third electrical signal;
and the preset terminal is also used for correcting the detection result according to the third electric signal.
7. The apparatus of claim 1, further comprising:
the database is used for storing the data detected by the detection device and the analysis result data of the preset terminal;
the detection device further comprises:
and the analog-to-digital conversion circuit is used for converting the first electric signal, the second electric signal and the third electric signal into digital signals and sending the digital signals to the preset terminal through the communication module.
8. The device of claim 1, wherein the wearable device body comprises at least one wearable patch, suction cups, at least one camisole, a harness, a binder, or a universal strap.
9. The apparatus of claim 1, wherein the machine learning algorithm used by the predetermined terminal comprises:
the data preprocessing algorithm is used for extracting, sorting, classifying, filtering and cutting the collected data; completing the feature extraction of data, respectively establishing a health and disease comprehensive disease feature database, and completing the training of sample database data; completing result analysis of the data to be detected;
the machine learning algorithm training model is used for extracting the characteristics of the collected data, respectively establishing a health and comprehensive disease characteristic database and training the data of the sample database;
and the machine learning inference classification algorithm is used for analyzing the classification result of the data to be detected by using a machine learning model.
10. The apparatus of claim 9, wherein,
the data preprocessing algorithm comprises: the data preprocessing and cutting algorithm is used for filtering, cutting and standardizing the preprocessed data by using functional analysis; completing the feature extraction of the data by using a feature extraction algorithm, respectively establishing a health and comprehensive disease feature database, and completing the training of the data of the sample database;
the machine learning algorithm training model comprises: performing feature extraction on data to be processed by using orthogonal basis fitting, PCA (principal component analysis) and nonlinear dimension reduction; respectively establishing health and comprehensive disease characteristic databases; performing model training on the extracted sample library data feature set by using statistical learning and machine learning;
the machine learning inference classification algorithm comprises the following steps: KNN, SVM, naive Bayes and softmax neural network classification algorithms are used for carrying out result classification analysis on data to be detected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110660885.8A CN115474897A (en) | 2021-06-15 | 2021-06-15 | Wearable intelligent monitoring and identification system for audio and non-audio vibration signs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110660885.8A CN115474897A (en) | 2021-06-15 | 2021-06-15 | Wearable intelligent monitoring and identification system for audio and non-audio vibration signs |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115474897A true CN115474897A (en) | 2022-12-16 |
Family
ID=84418929
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110660885.8A Pending CN115474897A (en) | 2021-06-15 | 2021-06-15 | Wearable intelligent monitoring and identification system for audio and non-audio vibration signs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115474897A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115856628A (en) * | 2023-02-28 | 2023-03-28 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105592781A (en) * | 2014-11-27 | 2016-05-18 | 英特尔公司 | Wearable personal computer and medical device |
| CN110477863A (en) * | 2019-08-07 | 2019-11-22 | 王满 | A kind of intelligent algorithm model system and method based on cardiac function dynamic monitoring |
| CN110477864A (en) * | 2019-08-07 | 2019-11-22 | 王满 | A kind of management service system and method based on cardiac function dynamic monitoring and analysis |
| WO2021004345A1 (en) * | 2019-07-10 | 2021-01-14 | 东南大学 | Heart sound acquisition and analysis system and method employing cloud architecture |
| CN112472042A (en) * | 2020-11-06 | 2021-03-12 | 无锡闻心电子科技有限责任公司 | Wearable human body characteristic acquisition device, detection device and detection underwear |
-
2021
- 2021-06-15 CN CN202110660885.8A patent/CN115474897A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105592781A (en) * | 2014-11-27 | 2016-05-18 | 英特尔公司 | Wearable personal computer and medical device |
| WO2021004345A1 (en) * | 2019-07-10 | 2021-01-14 | 东南大学 | Heart sound acquisition and analysis system and method employing cloud architecture |
| CN110477863A (en) * | 2019-08-07 | 2019-11-22 | 王满 | A kind of intelligent algorithm model system and method based on cardiac function dynamic monitoring |
| CN110477864A (en) * | 2019-08-07 | 2019-11-22 | 王满 | A kind of management service system and method based on cardiac function dynamic monitoring and analysis |
| CN112472042A (en) * | 2020-11-06 | 2021-03-12 | 无锡闻心电子科技有限责任公司 | Wearable human body characteristic acquisition device, detection device and detection underwear |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115856628A (en) * | 2023-02-28 | 2023-03-28 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
| CN115856628B (en) * | 2023-02-28 | 2023-06-27 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9168018B2 (en) | System and method for classifying a heart sound | |
| CN108392211B (en) | Fatigue detection method based on multi-information fusion | |
| Fan et al. | Homecare-oriented intelligent long-term monitoring of blood pressure using electrocardiogram signals | |
| Chen et al. | Intelligent heartsound diagnostics on a cellphone using a hands-free kit | |
| CN111973151B (en) | Infectious disease monitoring system and method based on wearable intelligent bandage | |
| EP1855594A1 (en) | Analysis of auscultatory sounds using voice recognition | |
| Chowdhury et al. | Machine learning in wearable biomedical systems | |
| CN111184521B (en) | Pressure identification bracelet | |
| Zhao et al. | An IoT-based wearable system using accelerometers and machine learning for fetal movement monitoring | |
| Singh et al. | Short unsegmented PCG classification based on ensemble classifier | |
| CN111370124A (en) | Health analysis system and method based on facial recognition and big data | |
| CN110731764A (en) | pulse detection system | |
| Zarrabi et al. | A system for accurately predicting the risk of myocardial infarction using PCG, ECG and clinical features | |
| CN115474897A (en) | Wearable intelligent monitoring and identification system for audio and non-audio vibration signs | |
| Li et al. | Electret-based flexible pressure sensor for respiratory diseases auxiliary diagnosis system using machine learning technique | |
| CN110638482B (en) | Real-time monitoring system for bowel sound and abdominal pressure | |
| Omarov et al. | Electronic stethoscope for heartbeat abnormality detection | |
| Sobhan et al. | Data analysis methods for health monitoring sensors: a survey | |
| Lavanya et al. | [Retracted] Wearable Sensor‐Based Edge Computing Framework for Cardiac Arrhythmia Detection and Acute Stroke Prediction | |
| CN114176519A (en) | Non-contact electrocardiosignal quality classification method | |
| Martins et al. | Be-sys: Big data e-health system for analysis and detection of risk of septic shock in adult patients | |
| Wang | The real-time monitoring system for in-patient based on zigbee | |
| Konnova et al. | Application of machine learning algorithms for SCG signal classification | |
| Wadhwani et al. | IOT based biomedical wireless sensor networks and machine learning algorithms for detection of diseased conditions | |
| Yatnalli et al. | Development of Personalized Health Detection Device for Covid-19 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |