WO2020029255A1 - Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time - Google Patents

Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time Download PDF

Info

Publication number
WO2020029255A1
WO2020029255A1 PCT/CN2018/099985 CN2018099985W WO2020029255A1 WO 2020029255 A1 WO2020029255 A1 WO 2020029255A1 CN 2018099985 W CN2018099985 W CN 2018099985W WO 2020029255 A1 WO2020029255 A1 WO 2020029255A1
Authority
WO
WIPO (PCT)
Prior art keywords
vibration
detection
parameters
data
point
Prior art date
Application number
PCT/CN2018/099985
Other languages
French (fr)
Chinese (zh)
Inventor
韦岗
曹燕
王一歌
赵明剑
Original Assignee
广州丰谱信息技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 广州丰谱信息技术有限公司 filed Critical 广州丰谱信息技术有限公司
Priority to PCT/CN2018/099985 priority Critical patent/WO2020029255A1/en
Publication of WO2020029255A1 publication Critical patent/WO2020029255A1/en

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/02—Measuring pulse or heart rate

Definitions

  • the present invention mainly relates to the technical field of human body vibration characteristic detection, and particularly relates to a human body multi-point mechanical vibration wave parameter, a biological tissue sound attenuation parameter, and a method and device for detecting characteristics that evolve with time.
  • the normal human body has many vibrations, the biggest source of which is the heart.
  • the normal heart is a powerful muscle organ, just like a water pump. It beats day and night, rhythmically, and is the power device of the human blood circulation system.
  • the human blood circulation system contains blood vessels and blood, which are responsible for the transportation of oxygen, carbon dioxide, nutrients and waste in the human body.
  • Blood vessels are divided into arteries, veins and capillaries according to their structural functions. Blood contracts through the heart's left ventricle, squeezes into the aorta, and then passes to the systemic arteries.
  • Arteries are flexible ducts formed by connective tissue and muscle. When a large amount of blood enters the artery, it will increase the pressure of the artery and cause the diameter of the tube to expand. The artery can feel the expansion at a shallower surface and touch this pulse, which is called the pulse.
  • the number of times the heart beats (vibrates) in one minute is called heart rate, and the heart rate of normal adults is 60 to 100 times per minute.
  • the points on the vibration path can be regarded as the next-level vibration source, and the frequency of vibration is the same as the frequency of the vibration source. Therefore, the number of pulses and the number of heartbeats in a normal person Are consistent.
  • human blood is sticky or blood vessels become blocked, the blood vessels become harder, which affects the speed of vibration, which makes the pulse vibration and the shape of the heartbeat slightly different from normal conditions.
  • the principle of the electrocardiogram is to detect the weak current generated during the activation of the heart muscle before and after the heart beats (vibrates), and this current is transmitted to various parts of the body surface through the human tissue. Due to the different tissues of the body and the distance from the heart, the electrical potential of the ECG signal in different parts of the body is also different. For a normal heart, the direction, frequency, and amplitude of ECG signal changes are regular. If the electrical signals of different parts of the body surface are detected by electrodes, then amplified by an amplifier, and recorded by a recorder, an ECG pattern can be obtained.
  • the doctor can diagnose such things as arrhythmia, myocardial infarction, preconstriction, hypertension, and Heart diseases such as ectopic beats.
  • Pulse diagnosis is a disease diagnosis method unique to TCM.
  • TCM doctors diagnose by cutting the arterial pulse on the wrist.
  • the wrist arterial pulse is formed by the spread of heart beats along the arterial blood vessels and blood flow to the periphery.
  • the comprehensive information on the shape, strength, speed, and rhythm of the wrist reflects many physiological and pathological characteristics of the human cardiovascular system.
  • TCM doctors perceive the depth, speed, intensity, rhythm, and morphology of arterial pulsation, distinguish the cause of the disease based on the obtained pulses, infer the changes in the disease, identify the true and false conditions of the disease, and judge the prognosis of the disease.
  • TCM practitioners make subjective judgments based on finger perception and relying entirely on personal experience. Therefore, it is often the case that different TCM physicians interpret the pulse of the same patient differently.
  • detecting the vibration of the human body is a means of diagnosing and early warning of body diseases.
  • the actual situation is often: they have not felt well or have detected some symptoms before they go to the hospital for testing.
  • the first investigation is the late stage of the serious illness, which loses the precious time window for initial cure. .
  • freezing three feet is not a day of cold.
  • the physiological characteristics of the human body can be frequently detected and long-term comparative analysis can be performed, disease early warning can be issued as soon as possible and timely treatment can be performed.
  • the diagnosis experience of traditional Chinese medicine pulse diagnosis and western medicine electrocardiogram fully proves that by detecting the vibration of different positions of the human body, the relevant vibration vector parameters are extracted, the long-term data accumulation is performed, and the change of the vibration vector parameters with time can be compared and analyzed to determine It can detect the cardiovascular condition of the human body, so as to achieve health assessment and disease early warning.
  • the vibration of the human body generates a bioelectrical signal, so the vibration of the human body can be recorded by measuring the bioelectrical signal of the human body.
  • the electrical measurement method is mainly suitable for recording relatively strong vibrations, such as the beating of the heart. It must be pointed out that even the electrical signals measured near the human heart tissue are not just bioelectrical signals generated by the activities of the heart organs, but are the comprehensive result of various types of bioelectrical signals in the human body transmitted through multiple paths. Therefore, when performing ECG measurements, the human body needs to lie flat, do not speak, and it is best to close your eyes to avoid bioelectrical signals generated by other non-cardiac organ activities.
  • the human body's mechanical vibration waves can also be directly measured to obtain the human body's multi-point vibration characteristics, including passive, active and active-passive mixed measurement methods.
  • the so-called passive detection method which uses a vibration sensor to directly record mechanical vibration, has the advantages of simplicity and low cost, and the disadvantage is that the intensity of vibration measurement is easily blocked by muscle, fat and skin, resulting in weak useful signals.
  • the so-called active detection method is to send a detection signal to the inside of the human body at some positions on the human body, and then receive the reflected signal, to judge the characteristics of the vibration according to the difference between the detected signal and the reflected signal.
  • the active detection method is also subject to interference, mainly due to multipath transmission interference of ultrasonic signals.
  • Active detection can detect vibration sources by transmitting a frequency-modulated signal, and the frequency-modulated detection signal has a strong ability to resist amplitude attenuation. Therefore, combining the results of passive detection and active detection, in addition to detecting multi-dimensional wave parameters such as frequency, phase, amplitude, and harmonic composition of the vibration source, it can also extract muscle, fat, and vibration between the vibration source and the detection point.
  • the sound attenuation characteristics of tissues, such as skin allow a more comprehensive assessment of the health of the human body and more accurate early warning of diseases.
  • this patent proposes to use a combination of passive and active detection methods to perform multi-point and long-term monitoring of human vibration, extract multi-dimensional wave parameters of the vibration source and muscle, fat, skin, etc. between the vibration source and the detection point. Sound attenuation parameters of biological tissues, and methods and devices for measuring the characteristics of evolution over time.
  • the present invention provides a hybrid measurement method for the evolution of human body's multi-point mechanical vibration parameters over time. Combined with the device, it adopts a combination of passive and active detection methods to detect mechanical vibration waves at different detection points in the human body for a long time, and extract the multi-dimensional wave parameters of the vibration source and the muscle, fat, skin, etc. between the vibration source and the detection point.
  • the acoustic attenuation parameters of biological tissues are continuously accumulated to form a vector time series of vibration parameters, and the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning, and a single collection of data is tested to obtain a judgment result.
  • the hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time is characterized by performing a hybrid detection method combining passive and active detection at different spatial detection points of the human body to detect mechanical vibration waves.
  • Obtain the detection data sequence with controllable correlation in time and strong correlation in space extract the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point to form vibration parameters
  • the vector characteristic of the vector time series is obtained by deep learning.
  • the above-mentioned mixed detection method of passive and active means that: passive detection and active detection are adopted at the same space detection point; passive detection uses a vibration sensor to directly record mechanical vibration; active detection is emitted to the inside of the human body at this space detection point Detect the signal and then receive the reflected signal to determine the characteristics of the vibration based on the difference between the detected signal and the reflected signal.
  • Each detection data contains two parts, one is the passive detection data obtained by passive detection, and the other is the active detection data obtained by active detection.
  • a detection data matrix Y [y (n), y (n + 1), y (n + 2), ...] is formed, and each row of the matrix is called a detection Data sequence, which corresponds to the long-term accumulation of mechanical vibration waves collected at a certain space detection point.
  • the above-mentioned time-controllable correlation means that the detection data sequence can obtain the time of day, log, week, and month of the collected data according to different extraction frequencies, so that data of different degrees of correlation can be obtained on the time axis.
  • the detection data sequence is spatially related, for example, the detection data y 1 (n) collected at the space detection point 1 and the detection data collected at the space detection point 2 y 2 (n), the correlation between y 1 (n) and y 2 (n) reflects the state of the pipeline between the space detection point 1 and the space detection point 2. Because the space detection point 1 to the space detection point 2 pass through a variety of media, such as muscle, skin, blood vessels, and other special pipes, the state of the pipes is not limited to blood vessels.
  • the position determination of the same spatial detection point is performed by passively collecting mechanical vibration waves in the nearby area for comparison, and the position at which the maximum mechanical vibration wave is obtained is located as a spatial detection point in this area and marked.
  • the position at which the maximum mechanical vibration wave is obtained is located as a spatial detection point in this area and marked.
  • the formation of the above-mentioned vector time series of vibration parameters refers to: for a certain space detection point, the detection data y 1 (n) collected at a certain time n, assuming its length is L, and the detection data of the previous L1 length is passive detection data, The following L-L1 length detection data is active detection data. Based on the passive detection data, the active detection data, and the transmission of the detection signal, the sound attenuation parameters of muscle, fat, skin and other biological tissues between the space detection point and the vibration source are inverted.
  • M spatial detection points [c 1 (n + 1), c 2 (n + 1), c 3 ( n + 1), ..., c M (n + 1)] T.
  • a vector time matrix C [c (n), c (n + 1), c (n + 2), ...] is formed according to the evolution of time.
  • the vector time matrix of the above-mentioned vibration parameters is calculated and formed from the accumulated collection data of the M space detection points at different times.
  • the evolution characteristic model of the vibration parameters is established in the space-time dimension to obtain the nonlinear space-time of the human vibration. characteristic.
  • the evolution characteristic model indicates that the location of the space detection point may be abnormal.
  • the vibration parameters of some two space detection points do not match the originally learned vibration
  • the parameter evolution characteristic model indicates that an abnormality occurs between the two spatial detection points.
  • the general-standard vibration wave data can be used for training.
  • the multi-dimensional wave parameters and the sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point form a vector time series of vibration parameters, and then a deep learning model of the vector time series of vibration parameters is obtained.
  • the model trained from standard vibration wave data is a standard universal vibration parameter evolution characteristic model.
  • the vector time series of the vibration parameters obtained from the currently collected detection data can be tested based on the evolution characteristic model of the vibration parameters in front of itself, or with the standard evolution model of the vibration parameters.
  • the background processor stores a standard universal vibration parameter evolution characteristic model database.
  • a hybrid measuring device for realizing the human body's multi-point mechanical vibration parameters over time which realizes the above method, is characterized by including multiple vibration wave data acquisition modules, human-computer interaction modules, and background processors;
  • the vibration wave data acquisition module is mainly responsible for collecting vibration wave data , Made into flexible bands tied to various parts of the human body;
  • the human-machine interaction module is responsible for the collaborative work and switch control of multiple vibration wave data acquisition modules, storing and transmitting the collected data of multiple vibration wave data acquisition modules, and displaying Monitoring results:
  • the background processor extracts the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the muscle, fat, skin and other biological tissues between the vibration source and the detection point for the detection data of multiple vibration wave data acquisition modules to form the vibration parameter.
  • the vector time series, and the evolution characteristic model of the vector time series of the vibration parameters are obtained through deep learning, and the collected data is evaluated each time to give an evaluation result.
  • the vibration wave data acquisition module of the above-mentioned hybrid measuring device of human body's multi-point mechanical vibration parameters evolves with time includes a variety of acquisition sensors, multiple ultrasonic transceiver arrays, acquisition control units, buffers, and wireless transmission units; they are implanted in a flexible band When working, the acquisition sensor and the ultrasonic transceiver array are close to the skin, and the ultrasonic transceiver array is also coated with a coupling agent to isolate the air; the acquisition sensor includes a variety of scalar sensors and vibration passive acquisition sensors; the ultrasonic transceiver array is arranged into a plurality of transceivers in a certain way.
  • each sending and receiving group is arranged by the way that the middle probe sends ultrasonic waves and the surrounding probes receive ultrasonic waves.
  • Each sending and receiving group works in turn, and can only receive and not send, to achieve passive acquisition, and also send and receive, to achieve active collection;
  • the acquisition control unit controls the ultrasonic transceiver array and the vibration passive acquisition sensor to collect the vibration wave data, and the scalar sensor collects the scalar data, then buffers it to the buffer, and then transmits it to the human-computer interaction module through the wireless transmission unit. It is also sent by the interactive module unit over a wireless transmission through the wireless transmitting unit receives, via the collection control unit converts a plurality of ultrasonic array referred to ultrasound emitted.
  • the shape of the vibration wave data acquisition module is various to suit different parts. Note that the passive vibration acquisition sensor does not need to be set, and the passive acquisition can be realized by the ultrasonic transceiver array.
  • the human-machine interaction module of the above-mentioned hybrid measuring device of human body multi-point mechanical vibration parameters evolves with time includes a human-machine interaction interface, a control unit, a memory, and a communication unit; the human-machine interaction interface is responsible for receiving user input instructions and parameters, and the control unit according to the instructions Coordinated work with parameter control remote control multiple vibration wave data acquisition modules, responsible for generating specific transmission data according to user requirements and transmitting at the specific timing for the transmission probe of the ultrasonic transceiver array, and for each vibration wave data acquisition module.
  • the data is buffered after corresponding classification, and the control unit controls the transmitting probe of the ultrasonic transceiver array to send ultrasonic waves according to a specific timing.
  • the transmission timing is strictly controllable and can be sent simultaneously or asynchronously.
  • the communication unit mainly implements the human-computer interaction module and Wireless communication of multiple vibration wave data acquisition modules, two-way transmission of transmission control instructions and data; and communication between the human-computer interaction module and the background processor, which can be wireless, wired, or even a memory card Fang Single data acquisition of multiple vibration wave data acquisition modules.
  • the background processor of the above-mentioned hybrid measuring device of the human body's multi-point mechanical vibration parameters evolves with time to perform simple processing on all data collected by the vibration wave data acquisition module, such as filtering, denoising, and decorrelation, and then according to different spaces. Classification of detection points. For the vibration wave data of each spatial detection point, multi-dimensional wave parameters of the vibration source and sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point are extracted to form a vector time series of vibration parameters. Then, the evolution characteristic model of the vector time series of the vibration parameters is obtained through deep learning. Based on the model, a single data acquisition is tested and intelligently analyzed. The intuitive analysis results are given and transmitted to the human-computer interaction interface for display. The background processor stores the user's vibration wave collection data and each person's unique vibration parameter evolution characteristic model, as well as a standard universal vibration parameter evolution characteristic model database.
  • the hybrid measurement of human body's multi-point mechanical vibration parameters evolves with time, and the method of using a single hybrid measurement includes the following steps:
  • space detection point marking is to facilitate the finding of the same space detection point at different time acquisitions. Determine the position of the space detection point. Collect common mechanical vibration waves of the area near a measured point through common vibration sensors, and then compare. The position where the maximum mechanical vibration wave is obtained is located as the space detection point of this area and marked ( You can mark with a color pen or inkjet pen). When collecting detection data, if the mark is still present, the next time point is directly collected at this mark. If the mark is not clear, the same spatial detection point is re-determined by the same method.
  • vibration wave data acquisition modules have different shapes and are implanted on a flexible tape.
  • the flexible tapes have different shapes in order to adapt to different detection locations and areas, but some have the same shape.
  • a correspondence of a spatial detection point is required. If you wear a vibration wave data acquisition module on the navel and turn on the switch, the wireless transmission unit sends a communication handshake signal to the human-machine interaction module. After receiving the human-machine interaction module, it prompts the user to enter the detection location and sends “Settings” "Complete” feedback information to the corresponding vibration wave data acquisition module.
  • Each vibration wave data acquisition module starts acquisition at the same time.
  • Each vibration wave data acquisition module uses a combination of passive and active detection methods to collect detection data.
  • the human-machine interaction module sends a control instruction to each vibration wave data acquisition module, instructing it to start the acquisition.
  • Temperature sensor, pressure sensor, etc. collect the temperature and pressure of the measured part; passive vibration acquisition sensor passively collects the vibration wave of the measured part;
  • the receiving probes of each transmitting and receiving group of the ultrasonic transmitting and receiving array first passively collect vibration waves, and then in accordance with the order, the transmitting probes of each transmitting and receiving group send ultrasonic waves, and the receiving probes in the same group receive echoes.
  • the vibration wave data is sorted and stored in the buffer according to the serial number.
  • the ultrasonic wave sent by the transmitting probe of each sending and receiving group is sent through the wireless transmission unit through the human-computer interaction module.
  • the human-machine interaction module summarizes the detection data collected by multiple vibration wave data acquisition modules.
  • Each vibration wave data acquisition module transmits the single acquisition data stored in the buffer with its own space detection point mark to the human-computer interaction module through its own wireless transmission unit.
  • the human-machine interaction module summarizes the data collected by each vibration wave data acquisition module according to the space detection point marks.
  • the human-computer interaction module transmits the collected collection and detection data to the background server.
  • the human-machine interaction module adopts wireless or wired mode, or even a memory card mode, and transmits the collected data plus personal tag information to the background server.
  • the background server extracts vibration parameters from all data collected by the vibration wave data acquisition module, obtains the evolution model of the vector time series of the vibration parameters through deep learning, and gives an evaluation result for the current collected data.
  • the background processor processes all the data collected by the vibration wave data acquisition module in accordance with the space detection points, and passively acquires the active detection data, actively detects the data, and transmits the detection signal for each space detection point, inverting the detection point to the vibration source.
  • the acoustic attenuation parameters of biological tissues such as muscle, fat, skin, and the multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constitute the parameter vector.
  • the parameter vector of each space detection point is integrated and accumulated on the basis of the previously acquired vibration parameters to form a vector time series of vibration parameters, and then the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning.
  • the single vibration parameters are intelligently evaluated.
  • the evolutionary characteristic model indicates that the location of the space detection point may be abnormal.
  • the vibration wave parameters of some two space detection points do not match the originally learned
  • the evolution characteristic model of the vibration parameters indicates that an abnormality occurs between the two points.
  • the background server feeds back the analysis results to the user.
  • the background server gives an intuitive analysis result and transmits it to the human-computer interaction interface for display.
  • the present invention performs multi-point long-term monitoring of human vibration, detects mechanical vibration waves for a long time at different spatial detection points of the human body, and obtains a detection data sequence of controllable correlation in time and strong correlation in space. Limited to a certain test, speak with big data.
  • the present invention adopts a combination of passive and active detection methods, and uses different mechanical vibration wave data obtained from passive and active detection to invert the biological tissues such as muscle, fat, skin, etc. between the vibration source and the detection point.
  • the sound attenuation parameters and the multi-dimensional wave parameters extracted from the vibration source are rich in parameter information, which can not only explain the state of the vibration source, but also analyze multilayer biological characteristics from the vibration source to the detection point.
  • the present invention can not only perform routine detection on some detection quantities of scalars, such as body temperature and blood pressure, etc., but also can send back an ultrasonic wave to detect the echo of a body part in an active way.
  • scalars such as body temperature and blood pressure, etc.
  • multi-point detection it can Obtaining mechanical vibration waves collected at different points in space at different times obviously has the characteristics of spatio-temporal multi-dimensionality, it is no longer a scalar, and contains a wealth of information.
  • the present invention performs multi-point vibration detection on multiple parts of the body, which is equivalent to the observation of vibrations with multiple parallel channels. Since the mechanical vibration wave to be inspected is a sound wave, it is transmitted in blood vessels, muscles, and skin. Slow, with enough delay to capture. Instead of capturing electrical signals in various parts like an electrocardiogram, the electrical signals propagate fast and can hardly capture the delay characteristics, so the ECG will not observe this delay characteristic. However, the time-delay characteristic is closely related to the patency of the pipeline. The invention can make full use of the time-delay characteristic when detecting and collecting mechanical vibration waves and analyzing the spatial characteristics at different space detection points.
  • the mechanical vibration wave data collected by the present invention is accumulated daily to form big data, and combined with artificial intelligence deep learning to establish an evolution model of vector time series of vibration parameters of different spatial detection points unique to each person, according to the difference between today and yesterday , To detect weak differences in the body; in addition, standard universal vibration parameter evolution characteristic models can also be used for testing.
  • the device of the present invention has strong applicability. Except that some special-shaped vibration wave data acquisition modules are used only in special parts, other vibration wave data acquisition modules can be used in different places. Moreover, the device of the present invention can also be used during exercise, because human body movements, such as walking, the vibration caused by human body walking is greatly different from the vibration of heartbeat and pulse, and it is easy to be filtered out.
  • the multi-point detection of the device of the present invention uses a holistic view to monitor the health status and subtle differences, and it is easy to detect subtle differences in physical conditions, such as whether it is pregnant, and when the fetal heartbeat occurs.
  • FIG. 1 is a schematic diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment
  • FIG. 2 is a structural diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment
  • FIG. 3 (a) is a schematic top view of a vibration wave data acquisition module of the ultrasonic array of the device of the embodiment including three transceiver groups;
  • FIG. 3 (b) is a schematic diagram of the vibration wave data acquisition module of the ultrasonic array of the device according to this embodiment, which includes three transceiver groups;
  • FIG. 3 (c) is a schematic top view of a vibration wave data acquisition module in which the ultrasonic array of the device of this embodiment includes two transceiver groups;
  • FIG. 5 is a schematic diagram of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time according to this embodiment is placed on a human body for testing;
  • FIG. 6 is a flowchart of a hybrid measurement method in which human body multi-point mechanical vibration parameters evolve with time in this embodiment.
  • FIG. 1 it is a schematic diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment, and includes multiple vibration wave data acquisition modules 101, a human-machine interaction module 102, and a background processor 103.
  • the vibration wave data acquisition module 101 is mainly responsible for collecting vibration wave data and making it into a flexible band to be tied or attached to various parts of the human body.
  • the human-machine interaction module 102 is responsible for the cooperative work and switch control of multiple vibration wave data acquisition modules, storing and transmitting the collected data of the multiple vibration wave data acquisition modules, and displaying the monitoring results.
  • the background processor 103 accumulates the collected detection data of multiple vibration wave data acquisition modules, extracts the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the biological tissues such as muscle, fat, and skin between the vibration source and the detection point to form vibrations
  • the vector time series of parameters, the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning, and the evaluation results are given for each collected data.
  • the vibration wave data acquisition module includes an acquisition sensor, multiple ultrasonic transceiver arrays, an acquisition control unit, a buffer, and a wireless transmission unit.
  • the human-machine interaction module includes a human-machine interaction interface, a control unit, a memory and a communication unit.
  • Acquisition sensors include a variety of scalar sensors, such as temperature sensors, and passive vibration acquisition sensors, which can be implemented using thin-film vibration sensors; ultrasonic transceiver arrays collect vibration wave data. These data are collected and controlled by the acquisition control unit and passed through a buffer.
  • the wireless transmission unit transmits to the human-computer interaction module; in addition, the transmission data of the ultrasonic transceiver array is also sent by the human-machine interaction module to be received by the wireless transmission unit, and is transferred to multiple ultrasonic transceiver arrays by the acquisition control unit to be converted into ultrasonic transmission.
  • the human-machine interaction module performs corresponding classification on the data collected by each vibration wave data acquisition module and buffers the memory; the communication unit realizes wireless communication between the human-machine interaction module and multiple vibration wave data acquisition modules, that is, the transmission control instructions and Two-way data transmission; In addition, it is also responsible for the communication between the human-computer interaction module and the background processor. This can be wireless, wired, or even memory card. A single transmission of multiple vibration wave data acquisition modules is transmitted. data collection.
  • FIG. 3 it is a schematic diagram of a vibration wave data acquisition module of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment.
  • the embodiment includes a temperature acquisition sensor 301, a pressure acquisition sensor 302, and a passive vibration acquisition sensor 303.
  • the ultrasonic transceiver array is also coated with a coupling agent to isolate the air, as shown in Figure 3 (a).
  • the passive vibration acquisition sensor 303 passively collects the vibration wave detection data. It is assumed that after the space detection points are determined and marked, the middle transmitting probe of the ultrasonic transmitting and receiving array 304 is aligned with the marking position. The passive acquisition sensor 303 is also aimed at the vibration source at its center, which is favorable for passive acquisition of vibration waves.
  • Figure 3 (a) shows that the ultrasonic transceiver array has three transceiver groups. The buffer and the wireless transmission unit are implanted above the flexible band, as shown in Fig. 3 (b).
  • the number of ultrasonic transmitting and receiving array transmitting and receiving groups is set according to the inspection site.
  • Figure 3 (c) shows the vibration data acquisition module of the ultrasonic transmitting and receiving array with only two transmitting and receiving groups, which is used in the lower arm of the hand.
  • the ultrasonic transmitting and receiving array of the hybrid measuring device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment is arranged into a plurality of transmitting and receiving groups in a certain manner.
  • three transmitting and receiving groups are illustrated here. It is arranged in such a manner that the middle probe 401 sends ultrasonic waves, and the eight nearby probes 402 receive ultrasonic waves.
  • Each sending and receiving group works in turn. It can only receive and not send, that is, passive detection, and it can also send and receive, and it is active detection. All probes can collect passive inspection data.
  • FIG. 5 it is a schematic diagram of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment is placed on a human body for testing.
  • Several vibration wave data acquisition modules are placed on the human body.
  • the figure shows the placement of one at the heart o to monitor the mechanical vibration waves of the heart's attachments. From the right shoulder, five are placed on the right arm in order to form the OA path.
  • the OA1, A1A2, A2A3, A3A4, A4A5 sub-channels are placed on the left arm in order from the left shoulder to form an OB channel, which includes the OB1, B1B2, B2B3, B3B4, and B4B5 sub-channels, monitoring the mechanical vibration of the shoulder and upper and lower arms Waves; placing D1, D2, and E1 in the upper abdomen to form the OD pathway and the OE pathway; placing 13 to the lower limbs in the middle of the abdomen to form the OC pathway, including OC1, C1C2, C2C3, and two parallel branches C3C4 after C3 -C4C5-C5C6-C6C7-C7C8 and C3C9-C9C10-C10C11-C11C12-C12C13.
  • vibration wave data acquisition modules are placed at different spatial detection points of the human body, and a combination of passive and active detection methods is used to detect mechanical vibration waves for a long time to obtain detection data sequences with controllable correlation in time and strong correlation in space.
  • Each detection data contains two parts, one is the passive detection data obtained by passive detection, and the other is the active detection data obtained by active detection.
  • a detection data matrix Y [y (n), y (n + 1), y (n + 2), ...] is formed, and each row of the matrix is called a detection Data sequence, which corresponds to the long-term accumulation of mechanical vibration waves collected at a certain space detection point, such as fixedly accumulating the detection data of space detection point 1 (heart o) in the time dimension to form the detection of space detection point 1
  • the data sequence y 1 [y 1 (n), y 1 (n + 1), y 1 (n + 2), ...].
  • the above-mentioned time-controllable correlation means that the detection data sequence can obtain the time of day, log, week, and month of the collected data according to different extraction frequencies, so that data of different degrees of correlation can be obtained on the time axis.
  • the formation of the vector time series of the above-mentioned vibration parameters refers to: assuming that for the spatial detection point 1 (at the heart o), the detection data y 1 (n) collected at a certain time n, assuming its length is L, the detection data of the previous L1 length It is passive detection data, and the following L-L1 length detection data is active detection data.
  • a vector time matrix C [c (n), c (n + 1), c (n + 2), ...] is formed according to the evolution of time.
  • the vector time matrix of the above-mentioned vibration parameters is calculated and formed from the accumulated collection data of the M space detection points at different times.
  • the evolution characteristic model of the vibration parameters is established in the space-time dimension to obtain the nonlinear space-time of the human vibration. characteristic.
  • the evolutionary characteristic model indicates that the location of the space detection point may be abnormal.
  • the evolution characteristic model of vibration parameters indicates that an abnormality occurs between the two spatial detection points. For example, if the measured vibration parameters of the A2 space detection point do not match the evolution characteristic model of the original learned vibration parameters, it may indicate that the space detection point may be abnormal.
  • the vibration wave data collected at the upper arm and the vibration wave data collected at the shoulder and neck pass through certain channels such as muscle, skin, blood vessels, etc., and these two spatial detection points calculated from the long-term collected detection data The vector time series of vibration wave parameters.
  • the conduction model parameters of this pipeline are trained by artificial intelligence deep learning. When the measured vibration wave parameters of the two spatial detection points do not match the evolution characteristics of the original learned vibration parameters, the model prompts The pipe and its parts may be abnormal.
  • FIG. 6 it is a flowchart of a hybrid measurement method in which human body multi-point mechanical vibration parameters evolve with time in this embodiment.
  • the hybrid measuring device Before the multi-point mechanical vibration parameters of the human body evolve with time, the hybrid measuring device becomes a product, and before serving human bodies with unknown health status, it is used for a period of time on healthy, sub-healthy, and different diseased human bodies to collect vibration waves of various parts.
  • the data is used to form chronometer, diary, Zhou Zhi, Yue Zhi, etc., and use these data to train a standard universal vibration parameter evolution characteristic model.
  • When serving a human body of unknown health status due to the limited data collected in the previous period, there is no unique vibration parameter evolution characteristic model of its own, referring to this standard universal vibration parameter evolution characteristic model to evaluate the user's health status.
  • the hybrid measurement of human body's multi-point mechanical vibration parameters evolves with time, and the method of using a single hybrid measurement includes the following steps:
  • space detection point marking is to facilitate the finding of the same space detection point at different time acquisitions. Determine the position of the space detection point. Collect common mechanical vibration waves of the area near a measured point through common vibration sensors, and then compare. The position where the maximum mechanical vibration wave is obtained is located as the space detection point of this area and marked ( You can mark with a color pen or inkjet pen). When collecting detection data, if the mark is still present, the next time point is directly collected at this mark. If the mark is not clear, the same spatial detection point is re-determined by the same method.
  • vibration wave data acquisition modules have different shapes and are implanted on a flexible tape.
  • the flexible tapes have different shapes in order to adapt to different detection locations and areas, but some have the same shape.
  • a correspondence of a spatial detection point is required. If you wear a vibration wave data acquisition module on the navel and turn on the switch, the wireless transmission unit sends a communication handshake signal to the human-machine interaction module. After receiving the human-machine interaction module, it prompts the user to enter the detection location and sends “Settings” "Complete” feedback information to the corresponding vibration wave data acquisition module.
  • Each vibration wave data acquisition module starts acquisition at the same time.
  • Each vibration wave data acquisition module uses a combination of passive and active detection methods to collect detection data.
  • the human-machine interaction module sends a control instruction to each vibration wave data acquisition module, instructing it to start the acquisition.
  • Temperature sensor, pressure sensor, etc. collect the temperature and pressure of the measured part; passive vibration acquisition sensor passively collects the vibration wave of the measured part;
  • the receiving probes of each transmitting and receiving group of the ultrasonic transmitting and receiving array first passively collect vibration waves, and then in accordance with the order, the transmitting probes of each transmitting and receiving group send ultrasonic waves, and the receiving probes in the same group receive echoes.
  • the vibration wave data is sorted and stored in the buffer according to the serial number.
  • the ultrasonic wave sent by the transmitting probe of each sending and receiving group is sent through the wireless transmission unit through the human-computer interaction module.
  • the human-machine interaction module summarizes the detection data collected by multiple vibration wave data acquisition modules.
  • Each vibration wave data acquisition module transmits its single acquisition data stored in the buffer to its human-computer interaction module through its own wireless transmission unit.
  • the human-machine interaction module summarizes the data collected by each vibration wave data acquisition module according to the space detection point marks.
  • the human-computer interaction module transmits the collected collection and detection data to the background server.
  • the human-machine interaction module adopts wireless or wired mode, or even a memory card mode, and transmits the collected data plus personal tag information to the background server.
  • the background server extracts vibration parameters from all data collected by the vibration wave data acquisition module, obtains the evolution model of the vector time series of the vibration parameters through deep learning, and gives an evaluation result for the current collected data.
  • the background processor processes all the data collected by the vibration wave data acquisition module in accordance with the space detection points, and passively acquires the active detection data, actively detects the data, and transmits the detection signal for each space detection point, inverting the detection point to the vibration source.
  • the acoustic attenuation parameters of biological tissues such as muscle, fat, skin, and the multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constitute the parameter vector.
  • the parameter vector of each space detection point is integrated and accumulated on the basis of the previously acquired vibration parameters to form a vector time series of vibration parameters, and then the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning.
  • the single vibration parameters are intelligently evaluated.
  • the evolutionary characteristic model indicates that the location of the space detection point may be abnormal.
  • the vibration wave parameters of some two space detection points do not match the originally learned
  • the evolution characteristic model of the vibration parameters indicates that an abnormality occurs between the two points.
  • the background server feeds back the analysis results to the user.
  • the background server gives an intuitive analysis result and transmits it to the human-computer interaction interface for display.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Cardiology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Measuring And Recording Apparatus For Diagnosis (AREA)

Abstract

Disclosed are a method and a device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time, wherein a combined detection mode combining passive detection and active detection is employed to detect mechanical vibration waves at different spatial detection points on a human body over an extended period of time, extract multi-dimensional wave parameters of a vibration source and acoustic attenuation parameters of a biological tissue such as muscle, fat, or skin between the vibration source and the detection points, build up and form a vector time series of vibration parameters, and obtain an evolution characteristics model of the vector time series of vibration parameters through deep learning, with the data obtained in a single collection undergoing testing and determination. The method and the device overcome the limitations of a single detection which is subject to a chance result, and are useful for monitoring sub-optimal-health states, pregnancy, monitoring for a first fetal heartbeat, tracking chronic diseases, etc.

Description

人体多点机械振动参数随时间演进的混合测量方法与装置Hybrid measurement method and device of human body multi-point mechanical vibration parameters evolved with time 技术领域Technical field
本发明主要涉及人体振动特性检测技术领域，具体涉及人体多点机械振动波参数、生物组织声衰减参数，及其随时间演进特性的检测方法与装置。The present invention mainly relates to the technical field of human body vibration characteristic detection, and particularly relates to a human body multi-point mechanical vibration wave parameter, a biological tissue sound attenuation parameter, and a method and device for detecting characteristics that evolve with time.
背景技术Background technique
正常的人体有很多振动，其中最大的振动源是心脏。正常的心脏是一个强有力的肌肉器官，就像一个水泵，它日夜不停地、有节律地搏动着，是人体血液循环***的动力装置。人体血液循环***除了心脏，还包含血管和血液，负责人体氧气、二氧化碳、养分及废物的运送。血管按构造功能不同，分为动脉、静脉和毛细血管三种。血液经由心脏的左心室收缩而挤压流入主动脉，随即传递到全身动脉。动脉为富有弹性的***与肌肉所形成的管路。当大量血液进入动脉将使动脉压力变大而使管径扩张，在体表较浅处动脉即可感受到此扩张，触摸到此搏动，即所谓的脉搏。The normal human body has many vibrations, the biggest source of which is the heart. The normal heart is a powerful muscle organ, just like a water pump. It beats day and night, rhythmically, and is the power device of the human blood circulation system. In addition to the heart, the human blood circulation system contains blood vessels and blood, which are responsible for the transportation of oxygen, carbon dioxide, nutrients and waste in the human body. Blood vessels are divided into arteries, veins and capillaries according to their structural functions. Blood contracts through the heart's left ventricle, squeezes into the aorta, and then passes to the systemic arteries. Arteries are flexible ducts formed by connective tissue and muscle. When a large amount of blood enters the artery, it will increase the pressure of the artery and cause the diameter of the tube to expand. The artery can feel the expansion at a shallower surface and touch this pulse, which is called the pulse.
心脏在一分钟内跳动(振动)的次数被称为心率，正常成年人的心率每分钟为60～100次。根据物理学的惠更斯原理，振动路径(管路)上的点都可以看成是下一级振动源，振动频率和振动源的频率一样，因此，正常人的脉搏的次数与心跳的次数是一致的。当人体血液粘稠或者血管发生堵塞，血管***就会影响振动的快慢，从而使得脉搏的振动和心跳的形态与正常情况相比有些不同。为此，在医学上可以通过监测心脏和脉搏的振动形态，如振动的频率、相位、幅度、谐波成分等多维波参数，来判断人体是否健康。西医的心电图、中医的脉诊等就属此类检查。The number of times the heart beats (vibrates) in one minute is called heart rate, and the heart rate of normal adults is 60 to 100 times per minute. According to the Huygens principle of physics, the points on the vibration path (pipeline) can be regarded as the next-level vibration source, and the frequency of vibration is the same as the frequency of the vibration source. Therefore, the number of pulses and the number of heartbeats in a normal person Are consistent. When human blood is sticky or blood vessels become blocked, the blood vessels become harder, which affects the speed of vibration, which makes the pulse vibration and the shape of the heartbeat slightly different from normal conditions. For this reason, in medicine, it is possible to judge whether the human body is healthy by monitoring the vibration patterns of the heart and pulse, such as multi-dimensional wave parameters such as frequency, phase, amplitude, and harmonic components of the vibration. The electrocardiogram of western medicine and the pulse diagnosis of traditional Chinese medicine are examples of such examinations.
心电图的原理是检测心脏在搏动(振动)前后，心肌发生激动过程中产生的微弱电流，该电流经人体组织向身体表面的各个部位传导。由于身体各部分组织不同，距心脏的距离不同，心电信号在身体不同的部位所表现出的电位也不同。对正常心脏来说，心电信号变化的方向、频率、幅度等是有规律的。若通过电极将体表不同部位的电信号检测出来，再用放大器加以放大，并用记录器描记下来，就可得到心电图形。医生根据所记录的心电图波形的形态、波幅大小以及各波之间的相对时间关系，再与正常心电图相比较，便能诊断出诸如心电节律不齐、心肌梗塞、期前收缩、高血压、心脏异位搏动等心脏疾病。The principle of the electrocardiogram is to detect the weak current generated during the activation of the heart muscle before and after the heart beats (vibrates), and this current is transmitted to various parts of the body surface through the human tissue. Due to the different tissues of the body and the distance from the heart, the electrical potential of the ECG signal in different parts of the body is also different. For a normal heart, the direction, frequency, and amplitude of ECG signal changes are regular. If the electrical signals of different parts of the body surface are detected by electrodes, then amplified by an amplifier, and recorded by a recorder, an ECG pattern can be obtained. Based on the recorded shape of the ECG waveform, the amplitude of the waveform, and the relative time relationship between the waves, the doctor can diagnose such things as arrhythmia, myocardial infarction, preconstriction, hypertension, and Heart diseases such as ectopic beats.
脉诊是中医独有的一种疾病诊断方法，中医师通过切按手腕部的动脉脉搏来诊断。手腕部动脉脉搏是心脏搏动沿着动脉血管和血流向外周传播而形成，它所表现出来的形态、强度、速度与节律等方面的综合信息反映出人体心血管***的许多生理和病理特征。中医师感知动脉搏动显现部位的深浅、速率的快慢及其强度、节律和形态等方面，根据获得的脉象分辨疾病成因、推断疾病变化、识别病情真假和判断疾病预后等，但这些脉象要靠中医师凭手指感知，且完全依赖于个人的经验做出主观判断。因此，常常出现不同的中医师对同一就诊者的 脉象做出不同解析的情况。Pulse diagnosis is a disease diagnosis method unique to TCM. TCM doctors diagnose by cutting the arterial pulse on the wrist. The wrist arterial pulse is formed by the spread of heart beats along the arterial blood vessels and blood flow to the periphery. The comprehensive information on the shape, strength, speed, and rhythm of the wrist reflects many physiological and pathological characteristics of the human cardiovascular system. TCM doctors perceive the depth, speed, intensity, rhythm, and morphology of arterial pulsation, distinguish the cause of the disease based on the obtained pulses, infer the changes in the disease, identify the true and false conditions of the disease, and judge the prognosis of the disease. TCM practitioners make subjective judgments based on finger perception and relying entirely on personal experience. Therefore, it is often the case that different TCM physicians interpret the pulse of the same patient differently.
由此可见，对人体的振动进行检测，是一种诊断与预警身体疾病的手段。但是对于个人来说，实际情形经常是：已经感觉到身体很不舒服或发现某些疾病征兆后，才去医院进行检测的，往往一查就是大病的后期，失去了宝贵的初期治愈的时间窗口。其实，冰冻三尺非一日之寒，除非是突发的外部事故，人体脏器器官、血管、血液等若出现问题，必定是长期日积月累的结果。因此，如果能对人体的生理特性经常进行检测，并进行长期的对比分析，就可以尽早发出疾病预警并及时医治。目前，可以用家用血压计、智能体温计等进行人体血压、体温的日常记录及长时间观测。但是血压、体温这类标量参数太过单一，且随时间的变化非常缓慢，携带的信息量太小，很难反映人体心血管特性等生理参数。因此，要实现对人体健康状况的评估、疾病的预警，还必须有更多包含人体器官及心血管特性的信息，并且最好是这些信息能够用简便、对人体无损、无创的方式获得。中医脉诊、西医心电图的诊断经验充分证明，通过检测人体不同位置的振动，提取出相关的振动矢量参数，进行长时间的数据积累，并对振动矢量参数随时间的变化进行对比分析，可以判别出人体心血管的状况，从而实现健康评估及疾病预警。It can be seen that detecting the vibration of the human body is a means of diagnosing and early warning of body diseases. However, for individuals, the actual situation is often: they have not felt well or have detected some symptoms before they go to the hospital for testing. Often, the first investigation is the late stage of the serious illness, which loses the precious time window for initial cure. . In fact, freezing three feet is not a day of cold. Unless there is a sudden external accident, if there is a problem with human organs, blood vessels, blood, etc., it must be the result of long-term accumulation. Therefore, if the physiological characteristics of the human body can be frequently detected and long-term comparative analysis can be performed, disease early warning can be issued as soon as possible and timely treatment can be performed. At present, daily monitoring and long-term observation of human blood pressure and body temperature can be performed with a household sphygmomanometer and a smart thermometer. However, scalar parameters such as blood pressure and body temperature are too single, and change very slowly over time. The amount of information carried is too small to reflect physiological parameters such as cardiovascular characteristics of the human body. Therefore, in order to realize the assessment of human health and early warning of disease, there must be more information containing human organs and cardiovascular characteristics, and it is best that this information can be obtained in a simple, non-destructive and non-invasive way. The diagnosis experience of traditional Chinese medicine pulse diagnosis and western medicine electrocardiogram fully proves that by detecting the vibration of different positions of the human body, the relevant vibration vector parameters are extracted, the long-term data accumulation is performed, and the change of the vibration vector parameters with time can be compared and analyzed to determine It can detect the cardiovascular condition of the human body, so as to achieve health assessment and disease early warning.
为了更多地获取人体器官活动的信息，以及更好地鉴别不同器官的活动特性，很有必要同时测量人体多点的振动特性。比如，当母体怀孕有胎儿的时候，会多一个胎儿心跳的振动源，要将母体的心跳振动信息与胎儿的心跳振动信息加以区分，就需要进行多点测试和孕前孕后的长时监测。这对于尽早捕捉到胎儿的心跳、分析其是否有先天性心脏病十分重要。In order to obtain more information about the activities of human organs and to better distinguish the characteristics of the activities of different organs, it is necessary to measure the vibration characteristics of the human body at multiple points at the same time. For example, when the mother is pregnant with a fetus, there will be one more source of fetal heartbeat vibration. To distinguish the heartbeat vibration information of the mother from the fetus heartbeat vibration information, multipoint testing and long-term monitoring before and after pregnancy are needed. This is important to capture the fetal heartbeat as soon as possible and analyze whether it has congenital heart disease.
人体的振动会产生生物电信号，因此可以通过测量人体生物电信号的方式记录人体的振动。但由于人有许多器官，不同器官的活动均会激发生物电，而人体是导电体，所有的生物电流会瞬时传遍全身。因此电测量的方法主要适合于记录比较强的振动，如心脏的跳动。必须指出，即使在人体心脏组织附近测量到的电信号，也并非仅仅是心脏器官活动产生的生物电信号，而是人体中各类生物电信号经过多条路径传播后的综合结果。因此，在进行心电图测量时，人体需要平躺、不说话、最好闭上眼睛，以尽量避免其他非心脏器官活动产生的生物电信号。The vibration of the human body generates a bioelectrical signal, so the vibration of the human body can be recorded by measuring the bioelectrical signal of the human body. However, because humans have many organs, the activities of different organs will stimulate bioelectricity, and the human body is a conductive body, and all biological currents will be transmitted instantly throughout the body. Therefore, the electrical measurement method is mainly suitable for recording relatively strong vibrations, such as the beating of the heart. It must be pointed out that even the electrical signals measured near the human heart tissue are not just bioelectrical signals generated by the activities of the heart organs, but are the comprehensive result of various types of bioelectrical signals in the human body transmitted through multiple paths. Therefore, when performing ECG measurements, the human body needs to lie flat, do not speak, and it is best to close your eyes to avoid bioelectrical signals generated by other non-cardiac organ activities.
为了避免生物电信号瞬间在人体全身中的混合，也可以通过直接测量人体机械振动波，获取人体多点振动特性，包括被动、主动以及主被动混合测量的方式。所谓被动检测方式，即采用振动传感器直接记录机械振动，其优点是简单、成本低，缺点则是振动测量强度易受肌肉、脂肪及皮肤等的阻隔，导致有用信号微弱。所谓主动检测方式，是在人体某些位置向人体内部发射探测信号，然后接收反射信号，根据探测信号与反射信号的差异来判断振动的特性。为了对人体无害，常采用一定功率限制的超声信号(其频率要比振动源的振动频率高很多)作为发射探测信号。前期我们申请的专利“基于心脏点波动传导特性的血管状态检测 方法与装置”就是采用超声有源主动探测来采集心脏的波动和不同血管处的脉搏波，然后根据不同点采集到的信号，对血管的声波传输特性进行分析。In order to avoid the instantaneous mixing of bioelectric signals in the whole body of the human body, the human body's mechanical vibration waves can also be directly measured to obtain the human body's multi-point vibration characteristics, including passive, active and active-passive mixed measurement methods. The so-called passive detection method, which uses a vibration sensor to directly record mechanical vibration, has the advantages of simplicity and low cost, and the disadvantage is that the intensity of vibration measurement is easily blocked by muscle, fat and skin, resulting in weak useful signals. The so-called active detection method is to send a detection signal to the inside of the human body at some positions on the human body, and then receive the reflected signal, to judge the characteristics of the vibration according to the difference between the detected signal and the reflected signal. In order to be harmless to the human body, a certain power-limited ultrasonic signal (the frequency of which is much higher than the vibration frequency of the vibration source) is often used as the transmission detection signal. In the previous period, we applied for a patent "Method and Device for Detecting Blood Vessel Status Based on the Conduction Characteristics of Cardiac Point Waves", which uses ultrasonic active active detection to collect heart waves and pulse waves at different blood vessels, and then according to the signals collected at different points, The sound wave transmission characteristics of blood vessels were analyzed.
与被动检测方式一样，主动检测方式同样会受到干扰，主要是超声信号的多径传输干扰。主动探测可以通过发射频率调制信号的方式，对振动源进行探测，而频率调制探测信号抗幅度衰减的能力很强。因此，将被动检测与主动检测的结果相结合，除可以在检测振动源的频率、相位、幅度、谐波构成等多维波参数外，还可提取振动源到探测点之间的肌肉、脂肪及皮肤等组织的声衰减特性，从而更全面地评估人体的健康状况，更准确地进行疾病预警。Like the passive detection method, the active detection method is also subject to interference, mainly due to multipath transmission interference of ultrasonic signals. Active detection can detect vibration sources by transmitting a frequency-modulated signal, and the frequency-modulated detection signal has a strong ability to resist amplitude attenuation. Therefore, combining the results of passive detection and active detection, in addition to detecting multi-dimensional wave parameters such as frequency, phase, amplitude, and harmonic composition of the vibration source, it can also extract muscle, fat, and vibration between the vibration source and the detection point. The sound attenuation characteristics of tissues, such as skin, allow a more comprehensive assessment of the health of the human body and more accurate early warning of diseases.
故本专利提出，采用被动及主动相结合的混合检测方式，对人体振动进行空间多点、长时监测，提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，及其随时间演进特性的测量方法与装置。Therefore, this patent proposes to use a combination of passive and active detection methods to perform multi-point and long-term monitoring of human vibration, extract multi-dimensional wave parameters of the vibration source and muscle, fat, skin, etc. between the vibration source and the detection point. Sound attenuation parameters of biological tissues, and methods and devices for measuring the characteristics of evolution over time.
发明内容Summary of the invention
针对人体振动可以反映人体生命的健康状态，而目前多是单次采集心跳和脉搏波，以及采集信号以电为主等，本发明提供一种人体多点机械振动参数随时间演进的混合测量方法与装置，采用被动及主动相结合的混合检测方式，在人体不同的空间检测点长时检测机械振动波，提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，不断积累形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，并对于单次的采集数据进行测试，得到评判结果。In view of the fact that human vibration can reflect the health status of human life, most of the current methods are single-time acquisition of heartbeat and pulse waves, and the main signal is electricity. The present invention provides a hybrid measurement method for the evolution of human body's multi-point mechanical vibration parameters over time. Combined with the device, it adopts a combination of passive and active detection methods to detect mechanical vibration waves at different detection points in the human body for a long time, and extract the multi-dimensional wave parameters of the vibration source and the muscle, fat, skin, etc. between the vibration source and the detection point. The acoustic attenuation parameters of biological tissues are continuously accumulated to form a vector time series of vibration parameters, and the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning, and a single collection of data is tested to obtain a judgment result.
为了实现上述目的，本发明给出的人体多点机械振动参数随时间演进的混合测量方法，其特征在于在人体不同的空间检测点进行被动及主动相结合的混合检测方式，检测机械振动波，获取时间上可控相关性、空间上强相关性的检测数据序列，提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型。In order to achieve the above purpose, the hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time provided by the present invention is characterized by performing a hybrid detection method combining passive and active detection at different spatial detection points of the human body to detect mechanical vibration waves. Obtain the detection data sequence with controllable correlation in time and strong correlation in space, extract the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point to form vibration parameters The vector characteristic of the vector time series is obtained by deep learning.
上述被动及主动相结合的混合检测方式是指：在同一空间检测点既采取被动检测也采取主动检测；被动检测是采用振动传感器直接记录机械振动；主动检测是在此空间检测点向人体内部发射探测信号，然后接收反射信号，根据探测信号与反射信号的差异来判断振动的特性。The above-mentioned mixed detection method of passive and active means that: passive detection and active detection are adopted at the same space detection point; passive detection uses a vibration sensor to directly record mechanical vibration; active detection is emitted to the inside of the human body at this space detection point Detect the signal and then receive the reflected signal to determine the characteristics of the vibration based on the difference between the detected signal and the reflected signal.
上述检测数据序列是指：在某时刻n上采集的检测数据为y(n)＝[y 1(n),y 2(n),y 3(n),...,y M(n)] T，这里假设有M个空间检测点，y 1(n)是空间检测点1的检测数据，y 2(n)是空间检测点2的检测数据，.......，y M(n)是空间检测点M的检测数据。每个检测数据都包含两部分，一部分是被动检测获取的被动检测数据，一部分是主动检测获取的主动检测数据。在下一时刻n+1上采集的检测数据为y(n+1)＝[y 1(n+1),y 2(n+1),y 3(n+1),...,y M(n+1)] T。依次根据时间的演进，形成检测数据矩阵 Y＝[y(n),y(n+1),y(n+2),......]，该矩阵的每一行则称为一个检测数据序列，它对应于某一空间检测点采集的机械振动波的长时累积，如对空间检测点1的检测数据进行时间维度上的积累，形成空间检测点1的检测数据序列y 1＝[y 1(n),y 1(n+1),y 1(n+2),......]。 The above detection data sequence means that the detection data collected at a certain time n is y (n) = [y 1 (n), y 2 (n), y 3 (n), ..., y M (n) ] T , here it is assumed that there are M space detection points, y 1 (n) is the detection data of space detection point 1, y 2 (n) is the detection data of space detection point 2, ..., y M (n) is detection data of the space detection point M. Each detection data contains two parts, one is the passive detection data obtained by passive detection, and the other is the active detection data obtained by active detection. The detection data collected at the next moment n + 1 is y (n + 1) = [y 1 (n + 1), y 2 (n + 1), y 3 (n + 1), ..., y M (n + 1)] T. According to the evolution of time in turn, a detection data matrix Y = [y (n), y (n + 1), y (n + 2), ...] is formed, and each row of the matrix is called a detection Data sequence, which corresponds to the long-term accumulation of mechanical vibration waves collected at a certain space detection point. For example, the detection data of space detection point 1 is accumulated in the time dimension to form a detection data sequence of space detection point 1 y 1 = [ y 1 (n), y 1 (n + 1), y 1 (n + 2), ...].
上述时间上可控相关性是指检测数据序列根据不同的抽取频率，可以得到采集数据的时辰志、日志、周志和月志等，这样在时间轴上可以得到不同相关程度的数据。The above-mentioned time-controllable correlation means that the detection data sequence can obtain the time of day, log, week, and month of the collected data according to different extraction frequencies, so that data of different degrees of correlation can be obtained on the time axis.
上述人体不同的空间检测点采集的机械振动波，即检测数据序列是有空间关联的，如，在空间检测点1采集的检测数据y 1(n)，在空间检测点2处采集的检测数据y 2(n)，则y 1(n)和y 2(n)的关联反映了空间检测点1到空间检测点2之间管道的状态。由于空间检测点1到空间检测点2经过多种介质，如经过肌肉、皮肤、血管等特殊的管道，故此管道的状态不局限于血管。 The above-mentioned mechanical vibration waves collected at different space detection points of the human body, that is, the detection data sequence is spatially related, for example, the detection data y 1 (n) collected at the space detection point 1 and the detection data collected at the space detection point 2 y 2 (n), the correlation between y 1 (n) and y 2 (n) reflects the state of the pipeline between the space detection point 1 and the space detection point 2. Because the space detection point 1 to the space detection point 2 pass through a variety of media, such as muscle, skin, blood vessels, and other special pipes, the state of the pipes is not limited to blood vessels.
上述同一空间检测点的位置确定是通过被动采集附近区域的机械振动波来比较，获得最大机械振动波处的位置则定位为这块区域的空间检测点，并且做好标记。采集检测数据时，若标记在，则直接在此标记处采集，若标记不清晰，则用同样的方法重新确定此同一个空间检测点。The position determination of the same spatial detection point is performed by passively collecting mechanical vibration waves in the nearby area for comparison, and the position at which the maximum mechanical vibration wave is obtained is located as a spatial detection point in this area and marked. When collecting detection data, if the mark is on, it is directly collected at the mark. If the mark is not clear, the same spatial detection point is re-determined by the same method.
上述振动参数的矢量时间序列的形成是指：对于某一个空间检测点，某时刻n上采集的检测数据y 1(n)，假设其长度为L，前面L1长度的检测数据为被动检测数据，后面L-L1长度的检测数据为主动检测数据，根据被动检测数据、主动检测数据以及发射探测信号，反演出该空间检测点到振动源之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分等，组成参数矢量c 1(n)，M个空间检测点，则形成c(n)＝[c 1(n),c 2(n),c 3(n),...,c M(n)] T。同理，根据下一时刻n+1上采集的检测数据，M个空间检测点，形成c(n+1)＝[c 1(n+1),c 2(n+1),c 3(n+1),...,c M(n+1)] T。依次根据时间的演进，形成振动参数的矢量时间矩阵C＝[c(n),c(n+1),c(n+2),......]，该矩阵的每一行为某个空间检测点的振动参数的矢量时间序列。 The formation of the above-mentioned vector time series of vibration parameters refers to: for a certain space detection point, the detection data y 1 (n) collected at a certain time n, assuming its length is L, and the detection data of the previous L1 length is passive detection data, The following L-L1 length detection data is active detection data. Based on the passive detection data, the active detection data, and the transmission of the detection signal, the sound attenuation parameters of muscle, fat, skin and other biological tissues between the space detection point and the vibration source are inverted. , And the multi-dimensional wave parameters of the vibration source are calculated, including the frequency, phase, amplitude, and harmonic components of the vibration source, and the parameter vector c 1 (n) is composed, and M space detection points form c (n) = [c 1 (n), c 2 (n), c 3 (n), ..., c M (n)] T. Similarly, according to the detection data collected at the next moment n + 1, M spatial detection points form c (n + 1) = [c 1 (n + 1), c 2 (n + 1), c 3 ( n + 1), ..., c M (n + 1)] T. According to the evolution of time, a vector time matrix C = [c (n), c (n + 1), c (n + 2), ...] is formed according to the evolution of time. Vector time series of vibration parameters of three spatial detection points.
上述振动参数的矢量时间矩阵，是通过M个空间检测点不同时间上的累积采集数据计算和形成的，通过深度学习，在时空维度上建立振动参数的演进特性模型，获取人体振动的非线性时空特性。The vector time matrix of the above-mentioned vibration parameters is calculated and formed from the accumulated collection data of the M space detection points at different times. Through the deep learning, the evolution characteristic model of the vibration parameters is established in the space-time dimension to obtain the nonlinear space-time of the human vibration. characteristic.
当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动参数不匹配原来学习的振动参数的演进特性模型则提示该两个空间检测点之间出现异常。When the measured vibration parameters of a certain space detection point do not match the originally learned vibration parameters, the evolution characteristic model indicates that the location of the space detection point may be abnormal. When the vibration parameters of some two space detection points do not match the originally learned vibration The parameter evolution characteristic model indicates that an abnormality occurs between the two spatial detection points.
上述采集的振动波数据在某个个体上还比较少时，可以用通用标准的振动波数据来训练。先在健康、亚健康、不同病症的的人体身上使用佩戴一段时间，采集各个部位的机械振动波，即检测数据，形成时辰志、日志、周志和月志等，通过这些检测数据提取振动源的多维波参 数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型。通过标准的振动波数据训练出来的模型为标准通用的振动参数的演进特性模型。即是说，从当前采集的检测数据获取的振动参数的矢量时间序列既可以基于自己前面的振动参数的演进特性模型做测试，也可以和标准通用的振动参数的演进特性模型做测试。后台处理器上存有标准通用的振动参数演进特性模型数据库。When the above-mentioned collected vibration wave data is relatively small on a certain individual, the general-standard vibration wave data can be used for training. First use it on a healthy, sub-healthy, human body with different conditions for a period of time, collect mechanical vibration waves of various parts, that is, test data, and form time, log, week, and month records, and use these test data to extract the vibration source. The multi-dimensional wave parameters and the sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point form a vector time series of vibration parameters, and then a deep learning model of the vector time series of vibration parameters is obtained. The model trained from standard vibration wave data is a standard universal vibration parameter evolution characteristic model. That is to say, the vector time series of the vibration parameters obtained from the currently collected detection data can be tested based on the evolution characteristic model of the vibration parameters in front of itself, or with the standard evolution model of the vibration parameters. The background processor stores a standard universal vibration parameter evolution characteristic model database.
实现上述方法的人体多点机械振动参数随时间演进的混合测量装置，其特征在于包括多个振动波数据采集模块、人机交互模块和后台处理器；振动波数据采集模块主要负责采集振动波数据，做成柔性的带状式绑在人体的各个部位；人机交互模块负责多个振动波数据采集模块的协同工作和开关控制，存储和传输多个振动波数据采集模块的采集数据，以及显示监测结果；后台处理器对多个振动波数据采集模块的检测数据提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，以及对每次的采集数据进行测评，给出评价结果。A hybrid measuring device for realizing the human body's multi-point mechanical vibration parameters over time, which realizes the above method, is characterized by including multiple vibration wave data acquisition modules, human-computer interaction modules, and background processors; the vibration wave data acquisition module is mainly responsible for collecting vibration wave data , Made into flexible bands tied to various parts of the human body; the human-machine interaction module is responsible for the collaborative work and switch control of multiple vibration wave data acquisition modules, storing and transmitting the collected data of multiple vibration wave data acquisition modules, and displaying Monitoring results: The background processor extracts the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the muscle, fat, skin and other biological tissues between the vibration source and the detection point for the detection data of multiple vibration wave data acquisition modules to form the vibration parameter. The vector time series, and the evolution characteristic model of the vector time series of the vibration parameters are obtained through deep learning, and the collected data is evaluated each time to give an evaluation result.
上述人体多点机械振动参数随时间演进的混合测量装置的振动波数据采集模块包括多种采集传感器，多个超声波收发阵列，采集控制单元，缓存器，无线传输单元；它们植入于一个柔性带子上，工作时采集传感器和超声波收发阵列紧贴皮肤，超声波收发阵列还涂抹耦合剂以隔绝空气；采集传感器包括多种标量传感器和振动被动采集传感器；超声波收发阵列按照一定的方式排列成多个收发组，每个收发组采取中间探头发送超声波、周围的探头接收超声波的方式来排列，每个收发组轮流工作，可以只收不发，实现被动采集，也可以也发也收，实现主动采集；采集控制单元控制超声波收发阵列和振动被动采集传感器采集振动波数据，以及标量传感器采集标量数据，然后缓存到缓存器，随后通过无线传输单元传给人机交互模块；另外超声波收发阵列的发送数据也由人机交互模块通过无线传输单元发送过来通过无线传输单元接收，经由采集控制单元交由多个超声波收发阵列转换成超声发射出去。振动波数据采集模块的形状多样以适用不同的部位。注意振动被动采集传感器也可以不设置，而由超声波收发阵列来实现被动采集即可。The vibration wave data acquisition module of the above-mentioned hybrid measuring device of human body's multi-point mechanical vibration parameters evolves with time includes a variety of acquisition sensors, multiple ultrasonic transceiver arrays, acquisition control units, buffers, and wireless transmission units; they are implanted in a flexible band When working, the acquisition sensor and the ultrasonic transceiver array are close to the skin, and the ultrasonic transceiver array is also coated with a coupling agent to isolate the air; the acquisition sensor includes a variety of scalar sensors and vibration passive acquisition sensors; the ultrasonic transceiver array is arranged into a plurality of transceivers in a certain way. Group, each sending and receiving group is arranged by the way that the middle probe sends ultrasonic waves and the surrounding probes receive ultrasonic waves. Each sending and receiving group works in turn, and can only receive and not send, to achieve passive acquisition, and also send and receive, to achieve active collection; The acquisition control unit controls the ultrasonic transceiver array and the vibration passive acquisition sensor to collect the vibration wave data, and the scalar sensor collects the scalar data, then buffers it to the buffer, and then transmits it to the human-computer interaction module through the wireless transmission unit. It is also sent by the interactive module unit over a wireless transmission through the wireless transmitting unit receives, via the collection control unit converts a plurality of ultrasonic array referred to ultrasound emitted. The shape of the vibration wave data acquisition module is various to suit different parts. Note that the passive vibration acquisition sensor does not need to be set, and the passive acquisition can be realized by the ultrasonic transceiver array.
上述人体多点机械振动参数随时间演进的混合测量装置的人机交互模块包括人机交互界面、控制单元、存储器和通信单元；人机交互界面负责接收用户的输入指令和参数，控制单元根据指令和参数控制遥控多个振动波数据采集模块的协同工作，负责按照用户的要求产生特定的发送数据和按照特定的时序供超声波收发阵列的发送探头发送，以及对每个振动波数据采集模块采集的数据进行对应分类后给存储器进行缓存，控制单元控制超声波收发阵列的发送探头按照特定的时序发送超声波，发送时序严格可控，可以同时发送，也可以异步发送； 通信单元主要实现人机交互模块和多个振动波数据采集模块的无线通信，传输控制指令和数据的双向传输；以及实现人机交互模块和后台处理器的通信，这可以是无线方式，也可以是有线方式，甚至是用存储卡的方式，每次存储传输多个振动波数据采集模块的单次数据采集。The human-machine interaction module of the above-mentioned hybrid measuring device of human body multi-point mechanical vibration parameters evolves with time includes a human-machine interaction interface, a control unit, a memory, and a communication unit; the human-machine interaction interface is responsible for receiving user input instructions and parameters, and the control unit according to the instructions Coordinated work with parameter control remote control multiple vibration wave data acquisition modules, responsible for generating specific transmission data according to user requirements and transmitting at the specific timing for the transmission probe of the ultrasonic transceiver array, and for each vibration wave data acquisition module. The data is buffered after corresponding classification, and the control unit controls the transmitting probe of the ultrasonic transceiver array to send ultrasonic waves according to a specific timing. The transmission timing is strictly controllable and can be sent simultaneously or asynchronously. The communication unit mainly implements the human-computer interaction module and Wireless communication of multiple vibration wave data acquisition modules, two-way transmission of transmission control instructions and data; and communication between the human-computer interaction module and the background processor, which can be wireless, wired, or even a memory card Fang Single data acquisition of multiple vibration wave data acquisition modules.
上述人体多点机械振动参数随时间演进的混合测量装置的后台处理器对振动波数据采集模块采集的所有数据先进行简单的处理，如滤波去噪，去相关等处理后，然后按照不同的空间检测点分类，对每个空间检测点的振动波数据提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，在有模型的基础上对单次数据采集进行测试、智能分析，给出直观分析结果，同时传给人机交互界面进行显示。后台处理器存有用户的振动波采集数据和每个人特有的振动参数演进特性模型，以及标准通用的振动参数演进特性模型数据库。The background processor of the above-mentioned hybrid measuring device of the human body's multi-point mechanical vibration parameters evolves with time to perform simple processing on all data collected by the vibration wave data acquisition module, such as filtering, denoising, and decorrelation, and then according to different spaces. Classification of detection points. For the vibration wave data of each spatial detection point, multi-dimensional wave parameters of the vibration source and sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point are extracted to form a vector time series of vibration parameters. Then, the evolution characteristic model of the vector time series of the vibration parameters is obtained through deep learning. Based on the model, a single data acquisition is tested and intelligently analyzed. The intuitive analysis results are given and transmitted to the human-computer interaction interface for display. The background processor stores the user's vibration wave collection data and each person's unique vibration parameter evolution characteristic model, as well as a standard universal vibration parameter evolution characteristic model database.
在后台处理器上已经存有标准通用的振动参数演进特性模型数据库的情形下，人体多点机械振动参数随时间演进的混合测量，进行单次混合测量的使用方法包括如下步骤：In the case where a standard universal vibration parameter evolution characteristic model database is already stored on the background processor, the hybrid measurement of human body's multi-point mechanical vibration parameters evolves with time, and the method of using a single hybrid measurement includes the following steps:
(1)多个振动波数据采集模块佩戴在身体的各个部位，做好空间检测点的标记，以及空间检测点的对应，即是和人机交互模块做好空间检测点的对应关系。(1) Multiple vibration wave data acquisition modules are worn on various parts of the body to mark the space detection points and correspond to the space detection points, that is, to correspond to the space detection points with the human-computer interaction module.
空间检测点标记的目的是便于在不同时间采集时找到同一空间检测点。空间检测点位置确定通过常见的振动传感器来采集某一被测点附近区域的机械振动波，然后比较，获得最大机械振动波处的位置则定位为这块区域的空间检测点，做好标记(可以用彩色笔或者喷墨笔做标记)。采集检测数据时，若标记还在，则下次时间点直接在此标记处采集，若标记不清晰，则用同样的方法重新确定此同一个空间检测点。The purpose of space detection point marking is to facilitate the finding of the same space detection point at different time acquisitions. Determine the position of the space detection point. Collect common mechanical vibration waves of the area near a measured point through common vibration sensors, and then compare. The position where the maximum mechanical vibration wave is obtained is located as the space detection point of this area and marked ( You can mark with a color pen or inkjet pen). When collecting detection data, if the mark is still present, the next time point is directly collected at this mark. If the mark is not clear, the same spatial detection point is re-determined by the same method.
多个振动波数据采集模块有不同的外形，植入于一个柔性带子上，柔性的带子为了适应不同的检测部位和区域有不同的外形，但是有的外形是一样的。为了区别是放置在哪个空间检测点的，即是采集的检测数据是对应哪个位置的，需要一个空间检测点的对应。如佩戴一个振动波数据采集模块于肚脐上，打开其开关，则其无线传输单元发一个通信握手信号给人机交互模块，人机交互模块收到后则提示用户输入检测部位，同时发“设置完毕”的反馈信息给对应的振动波数据采集模块。Multiple vibration wave data acquisition modules have different shapes and are implanted on a flexible tape. The flexible tapes have different shapes in order to adapt to different detection locations and areas, but some have the same shape. In order to distinguish which spatial detection point is placed, that is, which position the collected detection data corresponds to, a correspondence of a spatial detection point is required. If you wear a vibration wave data acquisition module on the navel and turn on the switch, the wireless transmission unit sends a communication handshake signal to the human-machine interaction module. After receiving the human-machine interaction module, it prompts the user to enter the detection location and sends “Settings” "Complete" feedback information to the corresponding vibration wave data acquisition module.
多个振动波数据采集模块佩戴的时候还要涂抹一定的耦合液，使超声波收发阵列和皮肤紧贴在一起。When multiple vibration wave data acquisition modules are worn, a certain coupling fluid must be applied to make the ultrasonic transceiver array and the skin close together.
(2)多个振动波数据采集模块同时开始采集工作，每个振动波数据采集模块采用被动及主动相结合的混合检测方式采集检测数据。(2) Multiple vibration wave data acquisition modules start acquisition at the same time. Each vibration wave data acquisition module uses a combination of passive and active detection methods to collect detection data.
人机交互模块发出控制指令给每个振动波数据采集模块，命令其开始采集工作。The human-machine interaction module sends a control instruction to each vibration wave data acquisition module, instructing it to start the acquisition.
温度传感器、压力传感器等采集所测部位的温度和压力；振动被动采集传感器被动采集所测部位的振动波；Temperature sensor, pressure sensor, etc. collect the temperature and pressure of the measured part; passive vibration acquisition sensor passively collects the vibration wave of the measured part;
超声波收发阵列每个收发组的接收探头先进行被动采集振动波，然后按照次序，每个收发组的发送探头发超声波，同组内的接收探头接收回波。振动波数据按照序号分类存储到缓存器。The receiving probes of each transmitting and receiving group of the ultrasonic transmitting and receiving array first passively collect vibration waves, and then in accordance with the order, the transmitting probes of each transmitting and receiving group send ultrasonic waves, and the receiving probes in the same group receive echoes. The vibration wave data is sorted and stored in the buffer according to the serial number.
每个收发组的发送探头发的超声波是经由人机交互模块通过无线传输单元发送过来的。The ultrasonic wave sent by the transmitting probe of each sending and receiving group is sent through the wireless transmission unit through the human-computer interaction module.
(3)人机交互模块汇总多个振动波数据采集模块采集的检测数据。(3) The human-machine interaction module summarizes the detection data collected by multiple vibration wave data acquisition modules.
每个振动波数据采集模块都通过自身的无线传输单元把缓存器里面存储的单次采集数据加上自己的空间检测点标记传给人机交互模块。人机交互模块根据空间检测点标记来汇总每个振动波数据采集模块采集的数据。Each vibration wave data acquisition module transmits the single acquisition data stored in the buffer with its own space detection point mark to the human-computer interaction module through its own wireless transmission unit. The human-machine interaction module summarizes the data collected by each vibration wave data acquisition module according to the space detection point marks.
(4)人机交互模块把汇总的采集检测数据传给后台服务器。(4) The human-computer interaction module transmits the collected collection and detection data to the background server.
人机交互模块采用无线或者有线的方式，甚至是用存储卡的方式，把汇总的采集数据加上个人标签信息传给后台服务器。The human-machine interaction module adopts wireless or wired mode, or even a memory card mode, and transmits the collected data plus personal tag information to the background server.
(5)后台服务器对振动波数据采集模块采集的所有数据提取振动参数，通过深度学习得到振动参数的矢量时间序列的演进模型，以及对当次的采集数据给出评价结果。(5) The background server extracts vibration parameters from all data collected by the vibration wave data acquisition module, obtains the evolution model of the vector time series of the vibration parameters through deep learning, and gives an evaluation result for the current collected data.
后台处理器对振动波数据采集模块采集的所有数据按照空间检测点来进行处理，对每个空间检测点获取的被动检测数据、主动检测数据以及发射探测信号，反演出该检测点到振动源之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分等，组成参数矢量。综合每个空间检测点的参数矢量，以及在原来获取的振动参数基础之上累积，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型。The background processor processes all the data collected by the vibration wave data acquisition module in accordance with the space detection points, and passively acquires the active detection data, actively detects the data, and transmits the detection signal for each space detection point, inverting the detection point to the vibration source. The acoustic attenuation parameters of biological tissues such as muscle, fat, skin, and the multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constitute the parameter vector. The parameter vector of each space detection point is integrated and accumulated on the basis of the previously acquired vibration parameters to form a vector time series of vibration parameters, and then the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning.
然后在有模型的基础上对单次振动参数进行智能评价。当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动波参数不匹配原来学习的振动参数的演进特性模型则提示该两点之间出现异常。Then based on the model, the single vibration parameters are intelligently evaluated. When the measured vibration parameters of a certain space detection point do not match the previously learned vibration parameters, the evolutionary characteristic model indicates that the location of the space detection point may be abnormal. When the vibration wave parameters of some two space detection points do not match the originally learned The evolution characteristic model of the vibration parameters indicates that an abnormality occurs between the two points.
(6)后台服务器把分析结果反馈给用户。(6) The background server feeds back the analysis results to the user.
后台服务器给出直观分析结果，同时传给人机交互界面进行显示。The background server gives an intuitive analysis result and transmits it to the human-computer interaction interface for display.
本发明所提出的人体多点机械振动参数随时间演进的混合测量方法与装置，具有以下优点：The hybrid measurement method and device of human body multi-point mechanical vibration parameters evolved over time provided by the present invention have the following advantages:
(1)本发明对人体的振动进行多点长时监测，在人体不同的空间检测点长时检测机械振动波，获取时间上可控相关性、空间上强相关性的检测数据序列，不再局限于某一次的检测，用大数据来说话。(1) The present invention performs multi-point long-term monitoring of human vibration, detects mechanical vibration waves for a long time at different spatial detection points of the human body, and obtains a detection data sequence of controllable correlation in time and strong correlation in space. Limited to a certain test, speak with big data.
(2)本发明采用被动及主动相结合的混合检测方式，利用被动及主动检测获取的不同的机械振动波数据的来反演振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，以及提取振动源的多维波参数，参数信息丰富，不仅能说明该振动源的状态，也能分析振动源到探测点之间多层生物特征。(2) The present invention adopts a combination of passive and active detection methods, and uses different mechanical vibration wave data obtained from passive and active detection to invert the biological tissues such as muscle, fat, skin, etc. between the vibration source and the detection point. The sound attenuation parameters and the multi-dimensional wave parameters extracted from the vibration source are rich in parameter information, which can not only explain the state of the vibration source, but also analyze multilayer biological characteristics from the vibration source to the detection point.
(3)本发明不仅可以对标量的一些检测量，如体温和血压等做常规检测，还能通过主动有源的方式对身体部位发一个超声波后检测其回波，根据多点的检测，能够获取不同空间检测点上不同时间上采集的机械振动波，显然具有时空多维特性，不再是标量，含有丰富的信息量。(3) The present invention can not only perform routine detection on some detection quantities of scalars, such as body temperature and blood pressure, etc., but also can send back an ultrasonic wave to detect the echo of a body part in an active way. According to multi-point detection, it can Obtaining mechanical vibration waves collected at different points in space at different times obviously has the characteristics of spatio-temporal multi-dimensionality, it is no longer a scalar, and contains a wealth of information.
(4)本发明在身体的多个部位进行多点振动检测，相当于对振动的观察有多条并行的通道，由于检查的机械振动波是一种声波，在血管、肌肉、皮肤传播都比较慢，有足够的延时可以捕捉。而非像心电图一样捕捉各个部位的电信号，电信号传播快，几乎不能捕捉到延时特性，故心电图不会观测此延时特征。但是延时特征跟管道的通畅程度密切相关，本发明在不同空间检测点检测采集机械振动波分析空间特性时能充分利用此延时特征。(4) The present invention performs multi-point vibration detection on multiple parts of the body, which is equivalent to the observation of vibrations with multiple parallel channels. Since the mechanical vibration wave to be inspected is a sound wave, it is transmitted in blood vessels, muscles, and skin. Slow, with enough delay to capture. Instead of capturing electrical signals in various parts like an electrocardiogram, the electrical signals propagate fast and can hardly capture the delay characteristics, so the ECG will not observe this delay characteristic. However, the time-delay characteristic is closely related to the patency of the pipeline. The invention can make full use of the time-delay characteristic when detecting and collecting mechanical vibration waves and analyzing the spatial characteristics at different space detection points.
(5)本发明采集的机械振动波数据每天积累，形成大数据，再结合人工智能深度学习建立每个人特有的不同空间检测点振动参数的矢量时间序列的演进模型，可以根据今天和昨天的不同，查出身体微弱的不同；另外也可以采用标准通用的振动参数演进特性模型进行测试。(5) The mechanical vibration wave data collected by the present invention is accumulated daily to form big data, and combined with artificial intelligence deep learning to establish an evolution model of vector time series of vibration parameters of different spatial detection points unique to each person, according to the difference between today and yesterday , To detect weak differences in the body; in addition, standard universal vibration parameter evolution characteristic models can also be used for testing.
(6)本发明装置适用性强，除一些特殊形状的振动波数据采集模块仅在特殊部位使用外，另外的振动波数据采集模块可以在不同的地方使用。而且，本发明装置在运动时也可以使用，因为人体运动，如走路，人体走路带来的振动跟心跳和脉搏的振动区别大，容易被滤波筛除。(6) The device of the present invention has strong applicability. Except that some special-shaped vibration wave data acquisition modules are used only in special parts, other vibration wave data acquisition modules can be used in different places. Moreover, the device of the present invention can also be used during exercise, because human body movements, such as walking, the vibration caused by human body walking is greatly different from the vibration of heartbeat and pulse, and it is easy to be filtered out.
(7)本发明装置的多点检测，是用整体观对健康的状态和细微区别进行监测，容易查出细微的身体状况区别，如是否怀孕，以及胎儿的心跳是什么时候出现等。(7) The multi-point detection of the device of the present invention uses a holistic view to monitor the health status and subtle differences, and it is easy to detect subtle differences in physical conditions, such as whether it is pregnant, and when the fetal heartbeat occurs.
(8)本方法整体观的思想与中医把人作为整个***的思想类似，以及机械振动波的多点检测可以作为中医望闻问切四诊中切---“脉诊”的具体实施和拓展。(8) The idea of the overall view of this method is similar to that of traditional Chinese medicine in regard to people as a whole system, and the multi-point detection of mechanical vibration waves can be used as a specific implementation and expansion of the "pulse diagnosis" of traditional Chinese medicine.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
图1是本实施例人体多点机械振动参数随时间演进的混合测量装置示意图；FIG. 1 is a schematic diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment; FIG.
图2是本实施例人体多点机械振动参数随时间演进的混合测量装置结构图；FIG. 2 is a structural diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment; FIG.
图3(a)是本实施例装置超声波阵列含有3个收发组的振动波数据采集模块的上面示意图；FIG. 3 (a) is a schematic top view of a vibration wave data acquisition module of the ultrasonic array of the device of the embodiment including three transceiver groups; FIG.
图3(b)是本实施例装置超声波阵列含有3个收发组的振动波数据采集模块的下面示意图；FIG. 3 (b) is a schematic diagram of the vibration wave data acquisition module of the ultrasonic array of the device according to this embodiment, which includes three transceiver groups;
图3(c)是本实施例装置超声波阵列含有2个收发组的振动波数据采集模块的上面示意图；FIG. 3 (c) is a schematic top view of a vibration wave data acquisition module in which the ultrasonic array of the device of this embodiment includes two transceiver groups;
图4是本实施例人体多点机械振动参数随时间演进的混合测量装置的超声波收发阵列；4 is an ultrasonic transmitting and receiving array of a hybrid measuring device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment;
图5是本实施例人体多点机械振动参数随时间演进的混合测量装置放置在人体上进行测试的示意图；FIG. 5 is a schematic diagram of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time according to this embodiment is placed on a human body for testing;
图6是本实施例人体多点机械振动参数随时间演进的混合测量方法的流程图。FIG. 6 is a flowchart of a hybrid measurement method in which human body multi-point mechanical vibration parameters evolve with time in this embodiment.
具体实施方式detailed description
下面结合附图对本发明的具体实施方式作进一步说明，但本发明的实施不限于此。The specific embodiments of the present invention will be further described below with reference to the accompanying drawings, but the implementation of the present invention is not limited thereto.
如图1所示，是本实施例人体多点机械振动参数随时间演进的混合测量装置示意图，包括多个振动波数据采集模块101、人机交互模块102和后台处理器103。振动波数据采集模块101主要负责采集振动波数据，做成柔性的带状式绑在或者贴在人体的各个部位。人机交互模块102负责多个振动波数据采集模块的协同工作和开关控制，存储和传输多个振动波数据采集模块的采集数据，以及显示监测结果。后台处理器103对多个振动波数据采集模块的采集检测数据进行积累，提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，以及对每次的采集数据给出评价结果。As shown in FIG. 1, it is a schematic diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment, and includes multiple vibration wave data acquisition modules 101, a human-machine interaction module 102, and a background processor 103. The vibration wave data acquisition module 101 is mainly responsible for collecting vibration wave data and making it into a flexible band to be tied or attached to various parts of the human body. The human-machine interaction module 102 is responsible for the cooperative work and switch control of multiple vibration wave data acquisition modules, storing and transmitting the collected data of the multiple vibration wave data acquisition modules, and displaying the monitoring results. The background processor 103 accumulates the collected detection data of multiple vibration wave data acquisition modules, extracts the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the biological tissues such as muscle, fat, and skin between the vibration source and the detection point to form vibrations The vector time series of parameters, the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning, and the evaluation results are given for each collected data.
如图2所示，是本实施例人体多点机械振动参数随时间演进的混合测量装置结构图。振动波数据采集模块包括采集传感器，多个超声波收发阵列，采集控制单元，缓存器，无线传输单元。人机交互模块包括人机交互界面、控制单元、存储器和通信单元。采集传感器包括多种标量传感器，如温度传感器，以及振动被动采集传感器，可采用薄膜振动传感器来实现；超声波收发阵列采集振动波数据，这些数据由采集控制单元统一控制采集，经过缓存器缓存后通过无线传输单元传给人机交互模块；另外超声波收发阵列的发送数据也由人机交互模块发送过来通过无线传输单元接收，经由采集控制单元交由多个超声波收发阵列转换成超声发射出去。人机交互模块对每个振动波数据采集模块采集的数据进行对应分类后给存储器进行缓存；通信单元实现人机交互模块和多个振动波数据采集模块的无线通信，即是实现传输控制指令和数据的双向传输；另外还负责人机交互模块和后台处理器的通信，这可以是无线方式，也可以是有线方式，甚至是用存储卡的方式，传输多个振动波数据采集模块的单次数据采集。As shown in FIG. 2, it is a structural diagram of a hybrid measurement device in which human body multi-point mechanical vibration parameters evolve with time in this embodiment. The vibration wave data acquisition module includes an acquisition sensor, multiple ultrasonic transceiver arrays, an acquisition control unit, a buffer, and a wireless transmission unit. The human-machine interaction module includes a human-machine interaction interface, a control unit, a memory and a communication unit. Acquisition sensors include a variety of scalar sensors, such as temperature sensors, and passive vibration acquisition sensors, which can be implemented using thin-film vibration sensors; ultrasonic transceiver arrays collect vibration wave data. These data are collected and controlled by the acquisition control unit and passed through a buffer. The wireless transmission unit transmits to the human-computer interaction module; in addition, the transmission data of the ultrasonic transceiver array is also sent by the human-machine interaction module to be received by the wireless transmission unit, and is transferred to multiple ultrasonic transceiver arrays by the acquisition control unit to be converted into ultrasonic transmission. The human-machine interaction module performs corresponding classification on the data collected by each vibration wave data acquisition module and buffers the memory; the communication unit realizes wireless communication between the human-machine interaction module and multiple vibration wave data acquisition modules, that is, the transmission control instructions and Two-way data transmission; In addition, it is also responsible for the communication between the human-computer interaction module and the background processor. This can be wireless, wired, or even memory card. A single transmission of multiple vibration wave data acquisition modules is transmitted. data collection.
如图3所示，是本实施例人体多点机械振动参数随时间演进的混合测量装置的振动波数据采集模块示意图，实施例这里包括温度采集传感器301、压力采集传感器302、振动被动采集传感器303，超声波收发阵列304，缓存器305，无线传输单元306；它们植入于一个柔性带子307上，温度采集传感器、压力采集传感器，振动被动采集传感器，超声波收发阵列要对着皮肤采集数据，故嵌入柔性带子的下方，工作时采集传感器和超声波收发阵列紧贴皮肤， 超声波收发阵列还涂抹耦合剂以隔绝空气，如图3(a)所示。实施例中振动被动采集传感器303被动采集振动波检测数据，假设经过空间检测点确定和标记后，超声波收发阵列304的中间发探头对准标记位，中间探头的***的环形状的一圈为振动被动采集传感器303，其中心也是对准振动源，有利于被动采集振动波。图3(a)中示意的是超声波收发阵列有3个收发组。缓存器和无线传输单元植入在柔性带子的上方，如图3(b)所示。超声波收发阵列收发组的个数视***位设定，图3(c)示意了只有2个收发组的超声波收发阵列的振动波数据采集模块，用在手下臂等。As shown in FIG. 3, it is a schematic diagram of a vibration wave data acquisition module of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment. The embodiment includes a temperature acquisition sensor 301, a pressure acquisition sensor 302, and a passive vibration acquisition sensor 303. , Ultrasonic transceiver array 304, buffer 305, wireless transmission unit 306; they are implanted on a flexible band 307, temperature acquisition sensor, pressure acquisition sensor, vibration passive acquisition sensor, the ultrasonic transceiver array needs to collect data against the skin, so it is embedded Below the flexible band, the acquisition sensor and the ultrasonic transceiver array are close to the skin during work. The ultrasonic transceiver array is also coated with a coupling agent to isolate the air, as shown in Figure 3 (a). In the embodiment, the passive vibration acquisition sensor 303 passively collects the vibration wave detection data. It is assumed that after the space detection points are determined and marked, the middle transmitting probe of the ultrasonic transmitting and receiving array 304 is aligned with the marking position. The passive acquisition sensor 303 is also aimed at the vibration source at its center, which is favorable for passive acquisition of vibration waves. Figure 3 (a) shows that the ultrasonic transceiver array has three transceiver groups. The buffer and the wireless transmission unit are implanted above the flexible band, as shown in Fig. 3 (b). The number of ultrasonic transmitting and receiving array transmitting and receiving groups is set according to the inspection site. Figure 3 (c) shows the vibration data acquisition module of the ultrasonic transmitting and receiving array with only two transmitting and receiving groups, which is used in the lower arm of the hand.
如图4所示，是本实施例人体多点机械振动参数随时间演进的混合测量装置的超声波收发阵列按照一定的方式排列成多个收发组，这里示意了3个收发组，每个收发组采取中间探头401发送超声波，旁边的8个探头402接收超声波的方式来排列。每个收发组轮流工作，可以只收不发，即为被动探测，也可以也发也收，则为主动探测。所有的探头都可以采集被动检测数据。As shown in FIG. 4, the ultrasonic transmitting and receiving array of the hybrid measuring device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment is arranged into a plurality of transmitting and receiving groups in a certain manner. Here, three transmitting and receiving groups are illustrated here. It is arranged in such a manner that the middle probe 401 sends ultrasonic waves, and the eight nearby probes 402 receive ultrasonic waves. Each sending and receiving group works in turn. It can only receive and not send, that is, passive detection, and it can also send and receive, and it is active detection. All probes can collect passive inspection data.
如图5所示，是本实施例人体多点机械振动参数随时间演进的混合测量装置放置在人体上进行测试的示意。在人体放置了若干个振动波数据采集模块，图中示意了在心脏o处放置1个，监测心脏附件的机械振动波；从右肩开始依次在右臂上放置5个，组成OA通路，包含OA1、A1A2、A2A3、A3A4、A4A5子通路、从左肩开始依次在左臂上放置5个，组成OB通路，包含OB1、B1B2、B2B3、B3B4、B4B5子通路,监测肩部和上下臂的机械振动波；在上腹部放置D1、D2和E1，组成OD通路和OE通路；在腹部中间开始放置13个至下肢，组成OC通路，包含OC1、C1C2、C2C3、以及C3之后两个并联的支路C3C4-C4C5-C5C6-C6C7-C7C8和C3C9-C9C10-C10C11-C11C12-C12C13。As shown in FIG. 5, it is a schematic diagram of a hybrid measurement device whose human body multi-point mechanical vibration parameters evolve with time in this embodiment is placed on a human body for testing. Several vibration wave data acquisition modules are placed on the human body. The figure shows the placement of one at the heart o to monitor the mechanical vibration waves of the heart's attachments. From the right shoulder, five are placed on the right arm in order to form the OA path. The OA1, A1A2, A2A3, A3A4, A4A5 sub-channels are placed on the left arm in order from the left shoulder to form an OB channel, which includes the OB1, B1B2, B2B3, B3B4, and B4B5 sub-channels, monitoring the mechanical vibration of the shoulder and upper and lower arms Waves; placing D1, D2, and E1 in the upper abdomen to form the OD pathway and the OE pathway; placing 13 to the lower limbs in the middle of the abdomen to form the OC pathway, including OC1, C1C2, C2C3, and two parallel branches C3C4 after C3 -C4C5-C5C6-C6C7-C7C8 and C3C9-C9C10-C10C11-C11C12-C12C13.
图中在人体不同的空间检测点放置振动波数据采集模块，采用被动及主动相结合的混合检测方式长时检测机械振动波，获取时间上可控相关性、空间上强相关性的检测数据序列，提取振动源的多维波参数与振动源到探测点之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进模型。In the figure, vibration wave data acquisition modules are placed at different spatial detection points of the human body, and a combination of passive and active detection methods is used to detect mechanical vibration waves for a long time to obtain detection data sequences with controllable correlation in time and strong correlation in space. , Extract the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of muscle, fat, skin and other biological tissues between the vibration source and the detection point to form a vector time series of the vibration parameters, and then obtain the vector time series of the vibration parameters by deep learning Evolution model.
上述检测数据序列是指：在某时刻n上采集的检测数据为y(n)＝[y 1(n),y 2(n),y 3(n),...,y M(n)] T，这里假设有M个空间检测点，y 1(n)是空间检测点1(心脏o处)的检测数据，y 2(n)是空间检测点2(A1处)的检测数据，.......，y M(n)是空间检测点M(E1处)的检测数据。每个检测数据都包含两部分，一部分是被动检测获取的被动检测数据，一部分是主动检测获取的主动动检测数据。 The above detection data sequence means that the detection data collected at a certain time n is y (n) = [y 1 (n), y 2 (n), y 3 (n), ..., y M (n) ] T , here it is assumed that there are M spatial detection points, y 1 (n) is the detection data of the spatial detection point 1 (at the heart o), and y 2 (n) is the detection data of the spatial detection point 2 (at the A1). ..., y M (n) is the detection data of the spatial detection point M (at E1). Each detection data contains two parts, one is the passive detection data obtained by passive detection, and the other is the active detection data obtained by active detection.
在下一时刻n+1上采集的检测数据为y(n+1)＝[y 1(n+1),y 2(n+1),y 3(n+1),...,y M(n+1)] T。依次根据时间的演进，形成检测数据矩阵Y＝[y(n),y(n+1),y(n+2),......]，该矩阵的每一行则 称为一个检测数据序列，它对应于某一空间检测点采集的机械振动波的长时累积，如固定对空间检测点1(心脏o处)的检测数据进行时间维度上的积累，形成空间检测点1的检测数据序列y 1＝[y 1(n),y 1(n+1),y 1(n+2),......]。 The detection data collected at the next moment n + 1 is y (n + 1) = [y 1 (n + 1), y 2 (n + 1), y 3 (n + 1), ..., y M (n + 1)] T. According to the evolution of time in turn, a detection data matrix Y = [y (n), y (n + 1), y (n + 2), ...] is formed, and each row of the matrix is called a detection Data sequence, which corresponds to the long-term accumulation of mechanical vibration waves collected at a certain space detection point, such as fixedly accumulating the detection data of space detection point 1 (heart o) in the time dimension to form the detection of space detection point 1 The data sequence y 1 = [y 1 (n), y 1 (n + 1), y 1 (n + 2), ...].
上述时间上可控相关性是指检测数据序列根据不同的抽取频率，可以得到采集数据的时辰志、日志、周志和月志等，这样在时间轴上可以得到不同相关程度的数据。The above-mentioned time-controllable correlation means that the detection data sequence can obtain the time of day, log, week, and month of the collected data according to different extraction frequencies, so that data of different degrees of correlation can be obtained on the time axis.
上述振动参数的矢量时间序列的形成是指：假设对于空间检测点1(心脏o处)，某时刻n上采集的检测数据y 1(n)，假设其长度为L，前面L1长度的检测数据为被动检测数据，后面L-L1长度的检测数据为主动检测数据，根据被动检测数据、主动检测数据以及发射探测信号，反演出该检测点到振动源之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分等，组成参数矢量c 1(n)，M个空间检测点，则形成c(n)＝[c 1(n),c 2(n),c 3(n),...,c M(n)] T。同理，根据下一时刻n+1上采集的检测数据，M个空间检测点，形成c(n+1)＝[c 1(n+1),c 2(n+1),c 3(n+1),...,c M(n+1)] T。依次根据时间的演进，形成振动参数的矢量时间矩阵C＝[c(n),c(n+1),c(n+2),......]，该矩阵的每一行为某个空间检测点的振动参数的矢量时间序列。 The formation of the vector time series of the above-mentioned vibration parameters refers to: assuming that for the spatial detection point 1 (at the heart o), the detection data y 1 (n) collected at a certain time n, assuming its length is L, the detection data of the previous L1 length It is passive detection data, and the following L-L1 length detection data is active detection data. Based on the passive detection data, the active detection data, and the transmission of the detection signal, the muscle, fat, skin and other biological tissues between the detection point and the vibration source are inverted The sound attenuation parameters and multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, are composed of the parameter vector c 1 (n), and M space detection points form c (n ) = [C 1 (n), c 2 (n), c 3 (n), ..., c M (n)] T. Similarly, according to the detection data collected at the next moment n + 1, M spatial detection points form c (n + 1) = [c 1 (n + 1), c 2 (n + 1), c 3 ( n + 1), ..., c M (n + 1)] T. According to the evolution of time, a vector time matrix C = [c (n), c (n + 1), c (n + 2), ...] is formed according to the evolution of time. Vector time series of vibration parameters of three spatial detection points.
上述振动参数的矢量时间矩阵，是通过M个空间检测点不同时间上的累积采集数据计算和形成的，通过深度学习，在时空维度上建立振动参数的演进特性模型，获取人体振动的非线性时空特性。The vector time matrix of the above-mentioned vibration parameters is calculated and formed from the accumulated collection data of the M space detection points at different times. Through the deep learning, the evolution characteristic model of the vibration parameters is established in the space-time dimension to obtain the nonlinear space-time of the human vibration. characteristic.
当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动波参数不匹配原来学习的振动参数的演进特性模型则提示该两个空间检测点之间出现异常。如，A2空间检测点这次测的振动振动参数不匹配原来学习的振动参数的演进特性模型，则提示该空间检测点部位有可能异常。在OA通路上，上臂处采集的振动波数据和在肩颈处采集的振动波数据中间经过肌肉、皮肤、血管等一定的管道，通过长时采集的检测数据计算出的这两个空间检测点的振动波参数的矢量时间序列，通过人工智能深度学习训练这条管道的传导模型参数，当这次测的两个空间检测点的振动波参数不匹配原来学习的振动参数的演进特性模型则提示该管道及其部位有可能异常。When the measured vibration parameters of a certain space detection point do not match the previously learned vibration parameters, the evolutionary characteristic model indicates that the location of the space detection point may be abnormal. When the vibration wave parameters of some two space detection points do not match the originally learned The evolution characteristic model of vibration parameters indicates that an abnormality occurs between the two spatial detection points. For example, if the measured vibration parameters of the A2 space detection point do not match the evolution characteristic model of the original learned vibration parameters, it may indicate that the space detection point may be abnormal. On the OA path, the vibration wave data collected at the upper arm and the vibration wave data collected at the shoulder and neck pass through certain channels such as muscle, skin, blood vessels, etc., and these two spatial detection points calculated from the long-term collected detection data The vector time series of vibration wave parameters. The conduction model parameters of this pipeline are trained by artificial intelligence deep learning. When the measured vibration wave parameters of the two spatial detection points do not match the evolution characteristics of the original learned vibration parameters, the model prompts The pipe and its parts may be abnormal.
如图6所示，是本实施例人体多点机械振动参数随时间演进的混合测量方法的流程图。As shown in FIG. 6, it is a flowchart of a hybrid measurement method in which human body multi-point mechanical vibration parameters evolve with time in this embodiment.
在人体多点机械振动参数随时间演进的混合测量装置成为产品，服务于未知健康状态的人体之前，先在健康、亚健康、不同病症的的人体身上使用佩戴一段时间，采集各个部位的振动波数据，形成时辰志、日志、周志和月志等，用这些数据训练出一个标准通用的振动参数演进特性模型。当服务于未知健康状态的人体时，由于前期采集的数据有限而没有自己独 特的振动参数演进特性模型时，参考此标准通用的振动参数演进特性模型来评估用户的健康状态。Before the multi-point mechanical vibration parameters of the human body evolve with time, the hybrid measuring device becomes a product, and before serving human bodies with unknown health status, it is used for a period of time on healthy, sub-healthy, and different diseased human bodies to collect vibration waves of various parts. The data is used to form chronometer, diary, Zhou Zhi, Yue Zhi, etc., and use these data to train a standard universal vibration parameter evolution characteristic model. When serving a human body of unknown health status, due to the limited data collected in the previous period, there is no unique vibration parameter evolution characteristic model of its own, referring to this standard universal vibration parameter evolution characteristic model to evaluate the user's health status.
在后台处理器上已经存有标准通用的振动参数演进特性模型数据库时，人体多点机械振动参数随时间演进的混合测量，进行单次混合测量的使用方法包括如下步骤：When a standard universal vibration parameter evolution characteristic model database is already stored on the background processor, the hybrid measurement of human body's multi-point mechanical vibration parameters evolves with time, and the method of using a single hybrid measurement includes the following steps:
(1)多个振动波数据采集模块佩戴在身体的各个部位，做好空间检测点的标记，以及空间检测点的对应，即是和人机交互模块做好空间检测点的对应关系。(1) Multiple vibration wave data acquisition modules are worn on various parts of the body to mark the space detection points and correspond to the space detection points, that is, to correspond to the space detection points with the human-computer interaction module.
空间检测点标记的目的是便于在不同时间采集时找到同一空间检测点。空间检测点位置确定通过常见的振动传感器来采集某一被测点附近区域的机械振动波，然后比较，获得最大机械振动波处的位置则定位为这块区域的空间检测点，做好标记(可以用彩色笔或者喷墨笔做标记)。采集检测数据时，若标记还在，则下次时间点直接在此标记处采集，若标记不清晰，则用同样的方法重新确定此同一个空间检测点。The purpose of space detection point marking is to facilitate the finding of the same space detection point at different time acquisitions. Determine the position of the space detection point. Collect common mechanical vibration waves of the area near a measured point through common vibration sensors, and then compare. The position where the maximum mechanical vibration wave is obtained is located as the space detection point of this area and marked ( You can mark with a color pen or inkjet pen). When collecting detection data, if the mark is still present, the next time point is directly collected at this mark. If the mark is not clear, the same spatial detection point is re-determined by the same method.
多个振动波数据采集模块有不同的外形，植入于一个柔性带子上，柔性的带子为了适应不同的检测部位和区域有不同的外形，但是有的外形是一样的。为了区别是放置在哪个空间检测点的，即是采集的检测数据是对应哪个位置的，需要一个空间检测点的对应。如佩戴一个振动波数据采集模块于肚脐上，打开其开关，则其无线传输单元发一个通信握手信号给人机交互模块，人机交互模块收到后则提示用户输入检测部位，同时发“设置完毕”的反馈信息给对应的振动波数据采集模块。Multiple vibration wave data acquisition modules have different shapes and are implanted on a flexible tape. The flexible tapes have different shapes in order to adapt to different detection locations and areas, but some have the same shape. In order to distinguish which spatial detection point is placed, that is, which position the collected detection data corresponds to, a correspondence of a spatial detection point is required. If you wear a vibration wave data acquisition module on the navel and turn on the switch, the wireless transmission unit sends a communication handshake signal to the human-machine interaction module. After receiving the human-machine interaction module, it prompts the user to enter the detection location and sends “Settings” "Complete" feedback information to the corresponding vibration wave data acquisition module.
多个振动波数据采集模块佩戴的时候还要涂抹一定的耦合液，使超声波收发阵列和皮肤紧贴在一起。When multiple vibration wave data acquisition modules are worn, a certain coupling fluid must be applied to make the ultrasonic transceiver array and the skin close together.
(2)多个振动波数据采集模块同时开始采集工作，每个振动波数据采集模块采用被动及主动相结合的混合检测方式采集检测数据。(2) Multiple vibration wave data acquisition modules start acquisition at the same time. Each vibration wave data acquisition module uses a combination of passive and active detection methods to collect detection data.
人机交互模块发出控制指令给每个振动波数据采集模块，命令其开始采集工作。The human-machine interaction module sends a control instruction to each vibration wave data acquisition module, instructing it to start the acquisition.
温度传感器、压力传感器等采集所测部位的温度和压力；振动被动采集传感器被动采集所测部位的振动波；Temperature sensor, pressure sensor, etc. collect the temperature and pressure of the measured part; passive vibration acquisition sensor passively collects the vibration wave of the measured part;
超声波收发阵列每个收发组的接收探头先进行被动采集振动波，然后按照次序，每个收发组的发送探头发超声波，同组内的接收探头接收回波。振动波数据按照序号分类存储到缓存器。The receiving probes of each transmitting and receiving group of the ultrasonic transmitting and receiving array first passively collect vibration waves, and then in accordance with the order, the transmitting probes of each transmitting and receiving group send ultrasonic waves, and the receiving probes in the same group receive echoes. The vibration wave data is sorted and stored in the buffer according to the serial number.
每个收发组的发送探头发的超声波是经由人机交互模块通过无线传输单元发送过来的。The ultrasonic wave sent by the transmitting probe of each sending and receiving group is sent through the wireless transmission unit through the human-computer interaction module.
(3)人机交互模块汇总多个振动波数据采集模块采集的检测数据。(3) The human-machine interaction module summarizes the detection data collected by multiple vibration wave data acquisition modules.
每个振动波数据采集模块都通过自身的无线传输单元把缓存器里面存储的单次采集数据 加上自己的空间检测点标记传给人机交互模块。人机交互模块根据空间检测点标记来汇总每个振动波数据采集模块采集的数据。Each vibration wave data acquisition module transmits its single acquisition data stored in the buffer to its human-computer interaction module through its own wireless transmission unit. The human-machine interaction module summarizes the data collected by each vibration wave data acquisition module according to the space detection point marks.
(4)人机交互模块把汇总的采集检测数据传给后台服务器。(4) The human-computer interaction module transmits the collected collection and detection data to the background server.
人机交互模块采用无线或者有线的方式，甚至是用存储卡的方式，把汇总的采集数据加上个人标签信息传给后台服务器。The human-machine interaction module adopts wireless or wired mode, or even a memory card mode, and transmits the collected data plus personal tag information to the background server.
(5)后台服务器对振动波数据采集模块采集的所有数据提取振动参数，通过深度学习得到振动参数的矢量时间序列的演进模型，以及对当次的采集数据给出评价结果。(5) The background server extracts vibration parameters from all data collected by the vibration wave data acquisition module, obtains the evolution model of the vector time series of the vibration parameters through deep learning, and gives an evaluation result for the current collected data.
后台处理器对振动波数据采集模块采集的所有数据按照空间检测点来进行处理，对每个空间检测点获取的被动检测数据、主动检测数据以及发射探测信号，反演出该检测点到振动源之间的肌肉、脂肪、皮肤等生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分等，组成参数矢量。综合每个空间检测点的参数矢量，以及在原来获取的振动参数基础之上累积，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型。The background processor processes all the data collected by the vibration wave data acquisition module in accordance with the space detection points, and passively acquires the active detection data, actively detects the data, and transmits the detection signal for each space detection point, inverting the detection point to the vibration source. The acoustic attenuation parameters of biological tissues such as muscle, fat, skin, and the multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constitute the parameter vector. The parameter vector of each space detection point is integrated and accumulated on the basis of the previously acquired vibration parameters to form a vector time series of vibration parameters, and then the evolution characteristic model of the vector time series of vibration parameters is obtained through deep learning.
然后在有模型的基础上对单次振动参数进行智能评价。当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动波参数不匹配原来学习的振动参数的演进特性模型则提示该两点之间出现异常。Then based on the model, the single vibration parameters are intelligently evaluated. When the measured vibration parameters of a certain space detection point do not match the previously learned vibration parameters, the evolutionary characteristic model indicates that the location of the space detection point may be abnormal. When the vibration wave parameters of some two space detection points do not match the originally learned The evolution characteristic model of the vibration parameters indicates that an abnormality occurs between the two points.
(6)后台服务器把分析结果反馈给用户。(6) The background server feeds back the analysis results to the user.
后台服务器给出直观分析结果，同时传给人机交互界面进行显示。The background server gives an intuitive analysis result and transmits it to the human-computer interaction interface for display.
上述实施例为本发明较佳的实施方式，但本发明的实施方式并不受上述实施例的限制，其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化，均应为等效的置换方式，都包含在本发明的保护范围。The above embodiment is a preferred embodiment of the present invention, but the embodiment of the present invention is not limited by the above embodiment. Any other changes, modifications, substitutions, combinations, and modifications made without departing from the spirit and principle of the present invention, Simplified, all should be equivalent replacement methods, and all are included in the protection scope of the present invention.

Claims (9)

  1. 人体多点机械振动参数随时间演进的混合测量方法，其特征在于，在人体不同的空间检测点进行被动及主动相结合的混合检测方式，长时检测机械振动波，获取时间上可控相关性、空间上强相关性的检测数据序列，提取振动源的多维波参数与振动源到探测点之间的生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，对于单次的采集数据进行测试，得到评判结果，所述生物组织至少包括肌肉、脂肪、皮肤。A hybrid measurement method for the evolution of human body's multi-point mechanical vibration parameters over time, which is characterized by a combination of passive and active detection methods at different spatial detection points of the human body, long-term detection of mechanical vibration waves, and controllable temporal correlation And spatially strong correlation detection data sequence, extract the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the biological tissue between the vibration source and the detection point, form a vector time series of the vibration parameters, and then obtain the vibration parameter by deep learning. The evolution characteristic model of the vector time series is tested on a single collected data to obtain a judgment result. The biological tissue includes at least muscle, fat, and skin.
  2. 根据权利要求1所述的人体多点机械振动参数随时间演进的混合测量方法，其特征在于，所述被动及主动相结合的混合检测方式是指：在同一空间检测点既采取被动检测也采取主动检测；被动检测是采用振动传感器直接记录机械振动；主动检测是在此空间检测点向人体内部发射探测信号，然后接收反射信号，根据探测信号与反射信号的差异来判断振动的特性；The hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time according to claim 1, wherein the combination of passive and active detection means that both passive detection and active detection are adopted at the same space detection point. Active detection; passive detection uses a vibration sensor to directly record mechanical vibration; active detection sends a detection signal to the inside of the human body at this spatial detection point, and then receives a reflection signal to determine the characteristics of the vibration based on the difference between the detection signal and the reflection signal;
    检测数据序列是指：在某时刻n上采集的检测数据为y(n)＝[y 1(n),y 2(n),y 3(n),...,y M(n)] T，这里假设有M个空间检测点，y 1(n)是空间检测点1的检测数据，y 2(n)是空间检测点2的检测数据，.......，y M(n)是空间检测点M的检测数据；每个检测数据都包含两部分，一部分是被动检测获取的被动检测数据，一部分是主动检测获取的主动动检测数据； The detection data sequence means that the detection data collected at a certain time n is y (n) = [y 1 (n), y 2 (n), y 3 (n), ..., y M (n)] T , here it is assumed that there are M space detection points, y 1 (n) is the detection data of space detection point 1, y 2 (n) is the detection data of space detection point 2, ..., y M ( n) is the detection data of the spatial detection point M; each detection data contains two parts, one is the passive detection data obtained by passive detection, and the other is the active detection data obtained by active detection;
    在下一时刻n+1上采集的检测数据为y(n+1)＝[y 1(n+1),y 2(n+1),y 3(n+1),...,y M(n+1)] T；依次根据时间的演进，形成检测数据矩阵Y＝[y(n),y(n+1),y(n+2),......]，该矩阵的每一行则称为一个检测数据序列，它对应于某一空间检测点采集的机械振动波的长时累积，如对空间检测点1的检测数据进行时间维度上的积累，形成空间检测点1的检测数据序列y 1＝[y 1(n),y 1(n+1),y 1(n+2),......]； The detection data collected at the next moment n + 1 is y (n + 1) = [y 1 (n + 1), y 2 (n + 1), y 3 (n + 1), ..., y M (n + 1)] T ; according to the evolution of time, a detection data matrix Y = [y (n), y (n + 1), y (n + 2), ...] is formed, which matrix Each line of is called a detection data sequence, which corresponds to the long-term accumulation of mechanical vibration waves collected at a certain space detection point. For example, the detection data of space detection point 1 is accumulated in the time dimension to form space detection point 1. Detection data sequence y 1 = [y 1 (n), y 1 (n + 1), y 1 (n + 2), ...];
    时间上可控相关性是指检测数据序列根据不同的抽取频率，可以得到采集数据包括时辰志、日志、周志和月志，这样在时间轴上可以得到不同相关程度的数据；The temporally controllable correlation means that the detection data sequence can obtain the collected data including the time log, log, week log, and month log according to different extraction frequencies, so that data of different levels of correlation can be obtained on the time axis;
    人体不同的空间检测点采集的机械振动波，即检测数据序列是有空间关联的，如，在空间检测点1采集的检测数据y 1(n)，在空间检测点2采集的检测数据y 2(n)，则y 1(n)和y 2(n)的关联反映了空间检测点1到空间检测点2之间管道的状态；由于空间检测点1到空间检测点2经过多种介质，如经过肌肉、皮肤、血管等特殊的管道，故此管道的状态不局限于血管。 Mechanical vibration waves collected at different space detection points of the human body, that is, the detection data sequence is spatially related, for example, the detection data y 1 (n) collected at the space detection point 1 and the detection data y 2 collected at the space detection point 2 (n), then the association between y 1 (n) and y 2 (n) reflects the state of the pipeline between the space detection point 1 to the space detection point 2; since the space detection point 1 to the space detection point 2 pass through various media, Such as through muscle, skin, blood vessels and other special channels, so the state of the pipeline is not limited to blood vessels.
  3. 根据权利要求1所述的人体多点机械振动参数随时间演进的混合测量方法，其特征在于，同一空间检测点的位置确定是通过被动采集附近区域的机械振动波来比较，获得最大机械振动波处的位置则定位为这块区域的空间检测点，并且做好标记；采集检测数据时，若标记在，则直接在此标记处采集，若标记不清晰，则用同样的方法重新确定此同一个空间检测点。The hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time according to claim 1, characterized in that the determination of the position of the same space detection point is performed by passively collecting mechanical vibration waves in the nearby area for comparison to obtain the maximum mechanical vibration wave The location is located as the space detection point of this area and marked. If the detection data is collected, if the mark is in, it will be collected directly at this mark. If the mark is not clear, use the same method to re-determine the same A spatial detection point.
  4. 根据权利要求1所述的人体多点机械振动参数随时间演进的混合测量方法，其特征在于，振动参数的矢量时间序列的形成是指：对于某一个空间检测点，某时刻n上采集的检测数据y 1(n)，假设其长度为L，前面L1长度的检测数据为被动检测数据，后面L-L1长度的检测数据为主动检测数据，根据被动检测数据、主动检测数据以及发射探测信号，反演出该检测点到振动源之间的生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分，组成参数矢量c 1(n)，M个空间检测点，则形成c(n)＝[c 1(n),c 2(n),c 3(n),...,c M(n)] T，同理，根据下一时刻n+1上采集的检测数据，M个空间检测点，形成c(n+1)＝[c 1(n+1),c 2(n+1),c 3(n+1),...,c M(n+1)] T；依次根据时间的演进，形成振动参数的矢量时间矩阵C＝[c(n),c(n+1),c(n+2),......]，该矩阵的每一行为某个空间检测点的振动参数的矢量时间序列。 The hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time according to claim 1, characterized in that the formation of the vector time series of vibration parameters refers to: for a certain space detection point, the detection collected at a certain time n Data y 1 (n), assuming its length is L, the detection data of length L1 in the front is passive detection data, and the detection data of length L-L1 in the back is active detection data. Reverse the acoustic attenuation parameters of the biological tissue between the detection point and the vibration source, and calculate the multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constituting the parameter vector c 1 (n) , M space detection points, then form c (n) = [c 1 (n), c 2 (n), c 3 (n), ..., c M (n)] T , for the same reason, according to the following The detection data collected on n + 1 at a moment, M spatial detection points, forming c (n + 1) = [c 1 (n + 1), c 2 (n + 1), c 3 (n + 1), ..., c M (n + 1)] T ; according to the evolution of time, a vector time matrix C = [c (n), c (n + 1), c (n + 2), is formed according to the evolution of time. .....], each of the matrix Vector time series of vibration parameters at a certain spatial detection point.
  5. 根据权利要求1所述的人体多点机械振动参数随时间演进的混合测量方法，其特征在于，振动参数的矢量时间矩阵，是通过M个空间检测点不同时间上的累积采集数据计算和形成的，通过深度学习建立振动参数的矢量时间序列的演进特性模型，获取人体振动的非线性时空特性，The hybrid measurement method of human body multi-point mechanical vibration parameters evolved with time according to claim 1, characterized in that the vector time matrix of the vibration parameters is calculated and formed from the accumulated acquisition data at different times of the M space detection points , Establish the evolution characteristic model of the vector time series of vibration parameters through deep learning, and obtain the nonlinear spatiotemporal characteristics of human vibration,
    当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的矢量时间序列的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动波参数不匹配原来学习的振动参数的矢量时间序列的演进特性模型则提示该两个空间检测点之间出现异常，When the measured vibration parameters of a certain space detection point do not match the vector time series of the originally learned vibration parameters, the evolution characteristic model of the vector time series indicates that the location of the space detection point may be abnormal. When the vibration wave parameters of some two space detection points are not The evolution characteristic model of the vector time series that matches the previously learned vibration parameters indicates that an abnormality occurs between the two spatial detection points.
    上述采集的振动波数据在某个个体上还比较少时，可以用通用标准的振动波数据来训练，先在健康、亚健康、不同病症的的人体身上使用佩戴一段时间，采集各个部位的机械振动波，即检测数据，形成包括时辰志、日志、周志和月志，通过这些检测数据提取振动源的多维波参数与振动源到探测点之间的生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，通过标准的振动波数据训练出来的模型为标准通用的振动参数的演进特性模型，即是说，从当前采集的检测数据获取的振动参数 的矢量时间序列既可以基于自己前面的振动参数的演进特性模型做测试，也可以和标准通用的振动参数的演进特性模型做测试，后台处理器上存有标准通用的振动参数演进特性模型数据库。When the above-mentioned collected vibration wave data is relatively small on an individual, you can use the general standard vibration wave data for training. First, use it for a period of time on healthy, sub-healthy, and different diseased human bodies to collect the mechanical vibration of various parts. Waves, that is, detection data, form a time record, a log, a weekly log, and a month log. These detection data are used to extract the multidimensional wave parameters of the vibration source and the sound attenuation parameters of the biological tissue between the vibration source and the detection point to form a vector of vibration parameters. Time series, and then obtain the evolution characteristic model of vector time series of vibration parameters through deep learning. The model trained by standard vibration wave data is the standard universal vibration parameter evolution characteristic model, that is, the detection data collected from the current The obtained vector time series of the vibration parameters can be tested based on the evolution characteristic model of the vibration parameters in front of itself, or can be tested with the evolution characteristic model of the standard universal vibration parameters. The standard processor has the evolution of the standard universal vibration parameters. Characteristic model database.
  6. 人体多点机械振动参数随时间演进的混合测量装置，其特征在于，包括多个振动波数据采集模块、人机交互模块和后台处理器；振动波数据采集模块主要负责采集振动波数据，做成柔性的带状式绑在人体的各个部位；人机交互模块负责多个振动波数据采集模块的协同工作和开关控制，存储和传输多个振动波数据采集模块的采集数据，以及显示监测结果；后台处理器对多个振动波数据采集模块的检测数据提取振动源的多维波参数与振动源到探测点之间的生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，以及对每次的采集数据进行测评，给出评价结果，所述生物组织至少包括肌肉、脂肪和皮肤。The hybrid measuring device of human body multi-point mechanical vibration parameters evolves with time, which is characterized by including multiple vibration wave data acquisition modules, human-computer interaction modules and background processors; the vibration wave data acquisition module is mainly responsible for collecting vibration wave data and making Flexible bands are tied to various parts of the human body; the human-machine interaction module is responsible for the cooperative work and switch control of multiple vibration wave data acquisition modules, storing and transmitting the collected data of multiple vibration wave data acquisition modules, and displaying the monitoring results; The background processor extracts the multi-dimensional wave parameters of the vibration source and the sound attenuation parameters of the biological tissue between the vibration source and the detection point from the detection data of multiple vibration wave data acquisition modules to form a vector time series of the vibration parameters, and then obtains them through deep learning. The evolution characteristic model of the vector time series of the vibration parameters, and the evaluation is performed on each collected data to give an evaluation result. The biological tissue includes at least muscle, fat and skin.
  7. 根据权利要求6所述的人体多点机械振动参数随时间演进的混合测量装置，其特征在于，振动波数据采集模块包括多种采集传感器，多个超声波收发阵列，采集控制单元，缓存器，无线传输单元；它们植入于一个柔性带子上，工作时采集传感器和超声波收发阵列紧贴皮肤，超声波收发阵列还涂抹耦合剂以隔绝空气；采集传感器包括多种标量传感器和振动被动采集传感器；超声波收发阵列按照一定的方式排列成多个收发组，每个收发组采取中间探头发送超声波、周围的探头接收超声波的方式来排列，每个收发组轮流工作，可以只收不发，实现被动采集，也可以也发也收，实现主动采集；采集控制单元控制超声波收发阵列和振动被动采集传感器采集振动波数据，以及标量传感器采集标量数据，然后缓存到缓存器，随后通过无线传输单元传给人机交互模块；另外超声波收发阵列的发送数据也由人机交互模块通过无线传输单元发送过来通过无线传输单元接收，经由采集控制单元交由多个超声波收发阵列转换成超声发射出去。The hybrid measuring device of human body multi-point mechanical vibration parameters evolved with time according to claim 6, characterized in that the vibration wave data acquisition module comprises a plurality of acquisition sensors, a plurality of ultrasonic transceiver arrays, an acquisition control unit, a buffer, and a wireless Transmission units; they are implanted on a flexible strap, and the acquisition sensor and the ultrasonic transceiver array are close to the skin during operation. The ultrasonic transceiver array is also coated with a coupling agent to isolate the air; the acquisition sensors include a variety of scalar sensors and passive vibration acquisition sensors; The array is arranged into a plurality of sending and receiving groups in a certain way. Each sending and receiving group is arranged in such a way that the middle probe sends ultrasonic waves and the surrounding probes receive ultrasonic waves. Each sending and receiving group works in turn. It can also send and receive to realize active acquisition; the acquisition control unit controls the ultrasonic transceiver array and the passive vibration acquisition sensor to collect vibration wave data, and the scalar sensor collects scalar data, and then buffers it to the buffer, and then transmits it to the human-computer interaction through the wireless transmission unit Module In addition, the transmission data of the ultrasonic transceiver array is also sent by the human-computer interaction module through the wireless transmission unit and received by the wireless transmission unit, and is transferred to multiple ultrasonic transceiver arrays by the acquisition control unit to be converted into ultrasonic transmission.
  8. 根据权利要求6所述的人体多点机械振动参数随时间演进的混合测量装置，其特征在于，人机交互模块包括人机交互界面、控制单元、存储器和通信单元；人机交互界面负责接收用户的输入指令和参数，控制单元根据指令和参数控制遥控多个振动波数据采集模块的协同工作，负责按照用户的要求产生特定的发送数据和按照特定的时序供超声波收发阵列的发送探头发送，以及对每个振动波数据采集模块采集的数据进行对应分类后给存储器进行缓存，控制单元控制超声波收发阵列的发送探头按照特定的时序发送超声波，发送时序严格可控，可以同时发送，也可以异步发送；通信单元主要实现人机交互模块和多个振动波数据采 集模块的无线通信，传输控制指令和数据的双向传输；以及实现人机交互模块和后台处理器的通信，传输多个振动波数据采集模块的单次数据采集，The hybrid measuring device of human body multi-point mechanical vibration parameters evolved with time according to claim 6, characterized in that the human-machine interaction module comprises a human-machine interaction interface, a control unit, a memory and a communication unit; the human-machine interaction interface is responsible for receiving users According to the instructions and parameters, the control unit controls the coordinated work of the remote control of multiple vibration wave data acquisition modules, and is responsible for generating specific transmission data according to the user's requirements and for the transmission probes of the ultrasonic transceiver array according to a specific timing, and The data collected by each vibration wave data acquisition module is correspondingly classified and buffered in the memory. The control unit controls the transmitting probe of the ultrasonic transceiver array to send ultrasonic waves according to a specific timing. The transmission timing is strictly controllable, and can be sent simultaneously or asynchronously. ; The communication unit mainly realizes the wireless communication between the human-machine interaction module and multiple vibration wave data acquisition modules, and the two-way transmission of transmission control instructions and data; and the communication between the human-machine interaction module and the background processor, which transmits multiple vibration wave data acquisition Module Single data acquisition,
    后台处理器对振动波数据采集模块采集的所有数据先进行简单的滤波去噪、去相关处理后，然后按照不同的空间检测点分类，然后对每个点的振动波数据提取振动源的多维波参数与振动源到探测点之间的生物组织的声衰减参数，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型，在有模型的基础上对单次数据采集进行智能分析，给出直观分析结果，同时传给人机交互界面进行显示。The background processor first performs simple filtering and denoising on all the data collected by the vibration wave data acquisition module, and then classifies them according to different spatial detection points, and then extracts the multi-dimensional waves of the vibration source from the vibration wave data of each point. The parameters and the sound attenuation parameters of the biological tissue between the vibration source and the detection point form a vector time series of the vibration parameters, and then the depth evolution model of the vector time series of the vibration parameters is obtained. Based on the model, the single time The data is collected for intelligent analysis, and the intuitive analysis results are given, and transmitted to the human-computer interaction interface for display.
  9. 根据权利要求6所述的人体多点机械振动参数随时间演进的混合测量装置，其特征在于，在后台处理器上已经存有标准通用的振动参数演进特性模型数据库的情形下，人体多点机械振动参数随时间演进的混合测量，进行单次混合测量的使用方法包括如下步骤：The hybrid measuring device of human body multi-point mechanical vibration parameters evolved with time according to claim 6, wherein the human body multi-point mechanical body has a standard universal vibration parameter evolution characteristic model database stored on a background processor. Hybrid measurement of vibration parameters evolved over time. The method of using a single hybrid measurement includes the following steps:
    (1)多个振动波数据采集模块佩戴在身体的各个部位，做好空间检测点的标记，以及空间检测点的对应，即是和人机交互模块做好空间检测点的对应关系；(1) Multiple vibration wave data acquisition modules are worn on various parts of the body to mark the space detection points and correspond to the space detection points, that is, to correspond to the space detection points with the human-computer interaction module;
    空间检测点标记的目的是便于在不同时间采集时找到同一空间检测点，空间检测点位置确定通过常见的振动传感器来采集某一被测点附近区域的机械振动波，然后比较，获得最大机械振动波处的位置则定位为这块区域的空间检测点，做好标记；采集检测数据时，若标记还在，则下次时间点直接在此标记处采集，若标记不清晰，则用同样的方法重新确定此同一个空间检测点；The purpose of the space detection point marking is to facilitate the finding of the same space detection point at different time acquisitions. The position of the space detection point is determined by using common vibration sensors to collect mechanical vibration waves in the vicinity of a certain measured point, and then compared to obtain the maximum mechanical vibration. The position of the wave is positioned as the spatial detection point of this area, and a mark is made; when collecting the detection data, if the mark is still present, the next time point is directly collected at this mark. If the mark is not clear, use the same Method to re-determine this same space detection point;
    多个振动波数据采集模块有不同的外形，植入于一个柔性带子上，柔性的带子为了适应不同的检测部位和区域有不同的外形，但是有的外形是一样的，为了区别是放置在哪个空间检测点的，即是采集的检测数据是对应哪个位置的，需要一个空间检测点的对应，每个振动波数据采集模块在人体上放置好之后，打开其开关，则其无线传输单元发一个通信握手信号给人机交互模块，人机交互模块收到后则提示用户输入检测部位，同时发“设置完毕”的反馈信息给对应的振动波数据采集模块；Multiple vibration wave data acquisition modules have different shapes and are implanted on a flexible tape. The flexible tapes have different shapes in order to adapt to different detection locations and areas, but some have the same shape. In order to distinguish which one is placed, For the space detection point, which is the location of the collected detection data, a space detection point is required. After each vibration wave data acquisition module is placed on the human body, the switch is turned on, and its wireless transmission unit sends a The communication handshake signal is sent to the human-machine interaction module. After receiving, the human-machine interaction module prompts the user to input the detection part, and at the same time sends a "set-up" feedback message to the corresponding vibration wave data acquisition module;
    多个振动波数据采集模块佩戴的时候还要涂抹一定的耦合液，使超声波收发阵列和皮肤紧贴在一起；When multiple vibration wave data acquisition modules are worn, a certain coupling fluid must be applied to make the ultrasonic transceiver array and the skin close together;
    (2)多个振动波数据采集模块同时开始采集工作，每个振动波数据采集模块采用被动及主动相结合的混合检测方式采集检测数据；(2) Multiple vibration wave data acquisition modules start collecting data at the same time, and each vibration wave data acquisition module uses a combination of passive and active detection methods to collect detection data;
    人机交互模块发出控制指令给每个振动波数据采集模块，命令其开始采集工作；The human-computer interaction module sends a control instruction to each vibration wave data acquisition module, instructing it to start the acquisition work;
    温度传感器、压力传感器等采集所测部位的温度和压力；振动被动采集传感器被动采集所测部位的振动波；Temperature sensor, pressure sensor, etc. collect the temperature and pressure of the measured part; passive vibration acquisition sensor passively collects the vibration wave of the measured part;
    超声波收发阵列每个收发组的接收探头先进行被动采集振动波，然后按照次序，每个收发组的发送探头发超声波，同组内的接收探头接收回波，振动波数据按照序号分类存储到缓存器；The receiving probes of each sending and receiving group of the ultrasonic transmitting and receiving array first passively collect vibration waves, and then in accordance with the order, the sending probes of each sending and receiving group send ultrasonic waves, and the receiving probes in the same group receive echoes. The vibration wave data is stored in the buffer according to the serial number. Device
    每个收发组的发送探头发的超声波是经由人机交互模块通过无线传输单元发送过来的；The ultrasonic wave sent by the transmitting probe of each sending and receiving group is sent through the wireless transmission unit through the human-computer interaction module;
    (3)人机交互模块汇总多个振动波数据采集模块采集的检测数据；(3) The human-machine interaction module summarizes the detection data collected by multiple vibration wave data acquisition modules;
    每个振动波数据采集模块都通过自身的无线传输单元把缓存器里面存储的单次采集数据加上自己的空间检测点标记传给人机交互模块，人机交互模块根据空间检测点标记来汇总每个振动波数据采集模块采集的数据；Each vibration wave data acquisition module transmits the single acquisition data stored in the buffer plus its own space detection point mark to the human-machine interaction module through its own wireless transmission unit, and the human-machine interaction module summarizes the space detection point mark Data collected by each vibration wave data acquisition module;
    (4)人机交互模块把汇总的采集检测数据传给后台服务器；(4) The human-computer interaction module transmits the collected collection and detection data to the background server;
    人机交互模块采用无线或者有线的方式，或者用存储卡的方式，把汇总的采集数据加上个人标签信息传给后台服务器；The human-computer interaction module adopts wireless or wired mode or memory card to transmit the collected data plus personal tag information to the background server;
    (5)后台服务器对振动波数据采集模块采集的所有数据提取振动参数，通过深度学习得到振动参数的矢量时间序列的演进模型，以及对当次的采集数据给出评价结果；(5) The background server extracts vibration parameters from all the data collected by the vibration wave data acquisition module, obtains the evolution model of the vector time series of the vibration parameters through deep learning, and gives an evaluation result of the current collected data;
    后台处理器对振动波数据采集模块采集的所有数据按照空间检测点来进行处理，对每个空间检测点获取的被动检测数据、主动检测数据以及发射探测信号，反演出该检测点到振动源之间的生物组织的声衰减参数，以及计算出振动源的多维波参数，包括振动源的频率、相位、幅度、谐波成分，组成参数矢量，综合每个空间检测点的参数矢量，以及在原来获取的振动参数基础之上累积，形成振动参数的矢量时间序列，再通过深度学习得到振动参数的矢量时间序列的演进特性模型；The background processor processes all the data collected by the vibration wave data acquisition module in accordance with the space detection points, and passively acquires the active detection data, actively detects the data, and transmits the detection signal for each space detection point, inverting the detection point to the vibration source. Acoustic attenuation parameters of biological tissues and multi-dimensional wave parameters of the vibration source, including the frequency, phase, amplitude, and harmonic components of the vibration source, constituting the parameter vector, integrating the parameter vector of each space detection point, and the original The acquired vibration parameters are accumulated on the basis of which the vector time series of the vibration parameters is formed, and the evolution characteristic model of the vector time series of the vibration parameters is obtained through deep learning;
    然后在有模型的基础上对单次振动参数进行智能评价；当某个空间检测点某次测的振动参数不匹配原来学习的振动参数的演进特性模型则提示该空间检测点部位有可能异常，当某两个空间检测点的振动波参数不匹配原来学习的振动参数的演进特性模型则提示该两点之间出现异常；Then, based on the model, the single vibration parameters are intelligently evaluated; when the measured vibration parameters of a certain space detection point do not match the previously learned vibration parameters, the evolution characteristic model indicates that the space detection point may be abnormal. When the vibration wave parameters of a certain two space detection points do not match the evolution characteristic model of the originally learned vibration parameters, an abnormality occurs between the two points;
    (6)后台服务器把分析结果反馈给用户；(6) The background server feeds back the analysis results to the user;
    后台服务器给出直观分析结果，同时传给人机交互界面进行显示。The background server gives an intuitive analysis result and transmits it to the human-computer interaction interface for display.
PCT/CN2018/099985 2018-08-10 2018-08-10 Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time WO2020029255A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/099985 WO2020029255A1 (en) 2018-08-10 2018-08-10 Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/099985 WO2020029255A1 (en) 2018-08-10 2018-08-10 Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time

Publications (1)

Publication Number Publication Date
WO2020029255A1 true WO2020029255A1 (en) 2020-02-13

Family

ID=69415378

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/099985 WO2020029255A1 (en) 2018-08-10 2018-08-10 Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time

Country Status (1)

Country Link
WO (1) WO2020029255A1 (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1421905A1 (en) * 2001-08-20 2004-05-26 Japan Science and Technology Agency Tissue identifying method in ultrasonography and ultrasonograph
CN101773387A (en) * 2009-01-08 2010-07-14 香港中文大学 Body feeling network-based sleeveless driven pulse pressure measurement and automatic calibration device
CN104873186A (en) * 2015-04-17 2015-09-02 中国科学院苏州生物医学工程技术研究所 Wearable artery detection device and data processing method thereof
CN105286919A (en) * 2015-10-13 2016-02-03 广州丰谱信息技术有限公司 Vascular condition detection method and apparatus based on heart point fluctuation conduction property
CN105997019A (en) * 2016-05-09 2016-10-12 鲍崇智 A body sensor network-based multi-dimensional heartbeat information synchronous collection method and system
CN107348949A (en) * 2017-08-10 2017-11-17 上海理工大学 More time serieses couple non-contact cardiovascular physiology parametric analysis system and method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1421905A1 (en) * 2001-08-20 2004-05-26 Japan Science and Technology Agency Tissue identifying method in ultrasonography and ultrasonograph
CN101773387A (en) * 2009-01-08 2010-07-14 香港中文大学 Body feeling network-based sleeveless driven pulse pressure measurement and automatic calibration device
CN104873186A (en) * 2015-04-17 2015-09-02 中国科学院苏州生物医学工程技术研究所 Wearable artery detection device and data processing method thereof
CN105286919A (en) * 2015-10-13 2016-02-03 广州丰谱信息技术有限公司 Vascular condition detection method and apparatus based on heart point fluctuation conduction property
CN105997019A (en) * 2016-05-09 2016-10-12 鲍崇智 A body sensor network-based multi-dimensional heartbeat information synchronous collection method and system
CN107348949A (en) * 2017-08-10 2017-11-17 上海理工大学 More time serieses couple non-contact cardiovascular physiology parametric analysis system and method

Similar Documents

Publication Publication Date Title
CN104873186B (en) A kind of wearable artery detection device and its data processing method
EP3363366B1 (en) A device that detects vascular condition based on point fluctuation conduction characteristics of the heart
JP6443517B2 (en) Biological information measuring device and method of operating biological information measuring device or biological information measuring system
WO2011135507A1 (en) Visualization of myocardial infarct size in diagnostic ecg
CN107296628A (en) The detection method of the real-time detecting system of internal fistula thrombus, real-time detection apparatus and its VPV
CA3076914A1 (en) System and method for fusing ultrasound with additional signals
US11980448B2 (en) Apparatus and methods of monitoring maternal and fetal heart rate
CN113160921A (en) Construction method and application of digital human cardiovascular system based on hemodynamics
CN107106118A (en) The method for detecting dicrotic notch
JP2023534001A (en) Electrocardiogram generation device and method based on hostile generation network algorithm
US20230355204A1 (en) Wearable ultrasound patch for monitoring subjects in motion using machine learning and wireless electronics
CN102512147A (en) Non-destructive detector for human arteriosclerosis
CN109222924B (en) Cardiovascular characteristic noninvasive quantitative evaluation method and device based on acousto-optic combined spectrum
CN109223042B (en) Hybrid measuring device for human body multipoint mechanical vibration parameter evolution along with time
KR101879634B1 (en) Monitoring system for cardiopulmonary vessel
CN110916715A (en) Heart sound collection system based on little acceleration sensor
WO2020029255A1 (en) Method and device for combined measurement of changes in mechanical vibration parameters at multiple points on human body over time
JP2024516573A (en) Physiological parameter sensing system and method - Patents.com
CN201267466Y (en) Arteriosclerosis detector
CN210408410U (en) Wearable pulse information acquisition device
CN106137245A (en) A kind of auscultation method with reference to multiple cardiographic detector signal analysis
JP2022547784A (en) Non-invasive, real-time, beat-to-beat ambulatory blood pressure monitoring
CN205054184U (en) Probe wireless data transmission's arteriosclerosis apparatus
US20230210393A1 (en) Method and device for multidimensional analysis of the dynamics of cardiac activity
CN109009015A (en) System based on multi-parameters sampling

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18929730

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 24/06/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18929730

Country of ref document: EP

Kind code of ref document: A1