CN109157202B - Cardiovascular disease early warning system based on multi-physiological signal deep fusion - Google Patents
Cardiovascular disease early warning system based on multi-physiological signal deep fusion Download PDFInfo
- Publication number
- CN109157202B CN109157202B CN201811085130.4A CN201811085130A CN109157202B CN 109157202 B CN109157202 B CN 109157202B CN 201811085130 A CN201811085130 A CN 201811085130A CN 109157202 B CN109157202 B CN 109157202B
- Authority
- CN
- China
- Prior art keywords
- physiological
- physiological signal
- signals
- signal
- frequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/0205—Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/024—Detecting, measuring or recording pulse rate or heart rate
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
- A61B5/14542—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue for measuring blood gases
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
- A61B5/681—Wristwatch-type devices
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6813—Specially adapted to be attached to a specific body part
- A61B5/6824—Arm or wrist
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
Abstract
The invention provides a cardiovascular disease early warning system based on multi-physiological signal deep fusion, and relates to the technical field of wearable health medical monitoring. The system comprises a wearing device, a multi-physiological signal acquisition device, a transmission device, intelligent terminal equipment and a cloud server; the multi-physiological signal acquisition device and the transmission device are both arranged on the wearing device; the multi-physiological signal acquisition device acquires physiological signals of a detected person and transmits acquired physiological signal data to the intelligent terminal equipment through the transmission device; the intelligent terminal device is internally provided with a program for judging whether the detected person has cardiovascular disease risk or not and transmitting the processed physiological signal to the cloud server; the cloud server is internally provided with a program for classifying and diagnosing diseases of multiple physiological signals of a wearer and feeding back the signals to the intelligent terminal; the cardiovascular disease early warning system based on the multi-physiological signal deep fusion provides more accurate reference for the assessment of cardiac function and the diagnosis of cardiac diseases.
Description
Technical Field
The invention relates to the technical field of wearable health medical monitoring, in particular to a cardiovascular disease early warning system based on multi-physiological signal deep fusion.
Background
Cardiovascular disease has become a very common disease and seriously threatens human health. However, under the big background that medical resources in China are relatively deficient and aging is aggravated, intelligent and personalized medical diagnosis modes have huge development prospects. In the future, wearable health monitoring equipment capable of fusing multiple physiological parameters will become a prime force of digital mobile medical treatment.
Among the numerous physiological signals, cardiac electrical signals are an important means of detecting heart diseases, especially cardiovascular diseases with paroxysmal and random nature. The information of the pulse wave in the aspects of shape, intensity, speed and the like reflects the important physiological and pathological information of the cardiovascular system of the human body. The electrocardio and pulse signals belong to weak signals, and have low amplitude and low frequency, so that the electrocardio and pulse signals are easily interfered by various kinds in the process of extracting the heart pulse signals.
The invention patent with patent number 201710169969.5 provides an electrocardio monitoring device based on electronic epidermis, which detects heart rate and electrocardio, processes and transmits signals. The above patents provide reference schemes for body building and medical treatment starting from measuring electrocardio and heart rate parameters. However, on the premise of obtaining accurate electrocardiosignals, the combined monitoring and reference of multiple physiological signals are not realized, and the deep fusion is carried out according to the detection of multiple physiological parameters and the analysis of electrocardiosignals of different patients, so that the screening diagnosis and the personalized cardiovascular disease risk early warning are carried out on the diseases; the invention patent with patent number 201510873447.4 provides a physiological signal detection system based on multi-channel flexible fusion, which solves the problem of conflicting information fusion existing in multi-channel binary detection, and the invention patent with patent number 201680078461.8 provides a physiological parameter signal fusion processing method, device and system, which solves the joint judgment mechanism among multiple physiological parameters.
Disclosure of Invention
The invention aims to solve the technical problem of providing a cardiovascular disease early warning system based on multi-physiological signal deep fusion to realize early warning on cardiovascular disease risks aiming at the defects of the prior art.
A cardiovascular disease early warning system based on multi-physiological signal deep fusion comprises a wearing device, a multi-physiological signal acquisition device, a transmission device, intelligent terminal equipment and a cloud server; the wearing device is worn on the wrist of the detected person, and the multi-physiological-signal acquisition device and the transmission device are both arranged on the wearing device; the multi-physiological signal acquisition device is used for acquiring physiological signals of a detected person; the transmission device is used for transmitting the physiological signal data acquired by the multiple physiological signal acquisition devices to the intelligent terminal equipment; the intelligent terminal device is internally provided with a program used for processing physiological signals of a detected person, judging whether the detected person has cardiovascular disease risk or not, carrying out disease risk early warning, assisting a doctor to diagnose and transmitting the processed physiological signals to the cloud server; the cloud server is internally provided with a program, and based on a deep learning disease intelligent classification method, the disease classification and diagnosis are carried out on multiple physiological signals of a wearer according to the trained disease types, and the results are fed back to the intelligent terminal;
the built-in programs of the intelligent terminal comprise a signal intelligent preprocessing unit, a feature extraction unit, a parameter calculation and analysis unit and a risk early warning unit; the signal intelligent preprocessing unit is used for preprocessing multiple physiological signals acquired by the multiple physiological signal acquisition devices and transmitting the preprocessed multiple physiological signals to the cloud server; the feature extraction unit is used for extracting feature points of various preprocessed human physiological signal waveforms; the parameter calculating and analyzing unit comprises a time domain parameter calculating and analyzing module and a frequency domain parameter calculating and analyzing module; the time domain parameter calculation and analysis module calculates and analyzes the time domain parameters of the physiological signals according to the multi-physiological signal feature extraction result; the frequency domain parameter calculation and analysis module calculates and analyzes frequency domain parameters according to time-frequency analysis results of the multiple physiological signals and calculates amplitude-frequency characteristics of the multiple physiological signals after Fourier transform; the risk early warning unit carries out cardiovascular disease risk early warning according to real-time-frequency domain analysis results of various physiological signal parameters and feedback results of the cloud server, and assists doctors in diagnosing;
the program built in the cloud server comprises a multi-channel time-frequency analysis unit and a depth analysis calculation unit for multi-physiological signal fusion; the multi-channel time-frequency analysis unit comprises a time domain analysis module and a frequency domain analysis module; the time domain analysis module respectively performs identity mapping and downsampling transformation on the acquired multi-physiological signal data to obtain the identity mapped multi-physiological signal data, and downsampling transformation is used for generating time sequence sketches with different time scales, so that a plurality of input time sequences with different downsampling rates are obtained; the frequency domain analysis module removes high-frequency interference and random noise from the collected multi-physiological signal data by adopting a low-frequency filter with a plurality of smoothness values, and obtains a multi-frequency input time sequence by utilizing moving averages of different windows according to different smoothness values; the multi-physiological signal fusion depth analysis and calculation unit inputs the result obtained by the multi-channel time-frequency analysis unit into a convolutional neural network for convolution operation; information is gathered in a full connection layer or other classifiers to obtain a classification result; and finally, comparing the multi-physiological-signal measurement data of the wearer with the trained features of different cardiovascular diseases to obtain a cardiovascular disease risk classification result, and feeding the result back to the intelligent terminal.
Preferably, the wearing device comprises an elastic wrist band, a housing and a dial; the elastic wrist strap is made of soft materials, and two buttons are arranged in the middle of the outer side of the elastic wrist strap; a dial plate is arranged on the outer side of the elastic wrist strap, and magic tapes for fixing the elastic wrist strap are arranged at two ends of the elastic wrist strap; the casing is arranged outside the dial plate to form a dial plate structure and is used for packaging a hardware circuit; the back of the dial plate is provided with a metal buckle, and the metal buckle is fixedly connected with a button on the elastic wrist strap to connect the elastic wrist strap and the dial plate; the multiple physiological signal acquisition devices are fixed on the elastic wristbands.
Preferably, the multiple physiological signal acquisition devices comprise an electrocardio acquisition device and a pulse acquisition device; the electrocardiosignal acquisition device comprises a metal electrode and two conductive fabric electrodes, wherein the two conductive fabric electrodes and the metal electrode form three electrodes for acquiring electrocardiosignals; the two conductive fabric electrodes are respectively embedded into the inner side of the elastic wrist band, and the metal electrodes are arranged on the back of the dial plate; the pulse acquisition device is embedded on the surface of the shell, a photoelectric pulse blood oxygen sensor is adopted to respectively receive the reflected light intensity of two beams of red light and infrared light passing through the fingers of a testee, and a photoelectric volume method is adopted to acquire pulse waves to obtain two paths of different pulse wave waveforms; the output ends of the electrocardio acquisition device and the pulse acquisition device are both connected with a transmission device.
Preferably, the signal intelligent preprocessing unit comprises a first processing module and a second processing module; the first processing module removes low-quality untrusted signals in the multi-physiological signals; the second processing module is used for carrying out signal processing on a high-quality credible signal in the multiple physiological signals; the second processing module comprises a baseline drift removing module, a power frequency interference removing module and a high-frequency interference removing module; the baseline wander removing module is used for removing baseline wander from high-quality credible signals in the multiple physiological signals to obtain physiological signals with the baseline wander removed; the power frequency interference removing module is used for removing power frequency interference from the multiple physiological signals with the baseline drift removed to obtain signals with the power frequency interference removed; and the high-frequency interference removing module is used for removing high-frequency noise interference on the multi-physiological signal without power frequency interference to obtain a signal without the high-frequency noise interference, and the signal is used as an input signal for feature extraction.
Preferably, the transmission device is connected with the intelligent terminal device by adopting a wireless transceiving module, so that real-time communication between the multi-physiological-signal acquisition device and the intelligent terminal device is realized.
Preferably, the cardiovascular disease early warning system based on multi-physiological signal deep fusion further comprises a filtering device and an AD conversion device which are both arranged in the dial plate; the filtering device and the AD conversion device are connected with the output end of the multi-physiological signal acquisition device, filtering and AD conversion are carried out on the output signals of the multi-physiological signal acquisition device, and the multi-physiological signals after AD conversion are transmitted to the intelligent terminal equipment through the wireless transceiving module.
Preferably, the cardiovascular disease early warning system based on multi-physiological signal deep fusion further comprises a power management module for providing reference voltage for the filtering device and the AD conversion device; the power management module comprises a power management circuit and a battery; the power management circuit adopts the power management chip to realize the safe charge and discharge of lithium cell to provide +3V regulated voltage, the steady voltage chip converts +3V voltage into +2.5V voltage, for filter equipment and AD conversion device power supply and provide stable reference voltage.
Preferably, the cardiovascular disease early warning system based on multi-physiological signal deep fusion further comprises a storage module; the storage module selects a large-capacity storage medium, ensures that data is continuously stored for more than 24 hours, and is used for storing the AD-converted multi-physiological-signal original signals, the calculated multi-physiological-signal parameters and the time-frequency-domain analysis results of the multi-physiological signals.
Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: the cardiovascular disease early warning system based on the multi-physiological-signal deep fusion has the advantages that the tightness of a wristband strap can be adjusted, the equipment volume is small, the used conductive fabric electrode is soft in material and good in fit with the skin of a human body, discomfort of the traditional equipment to the skin is greatly reduced, great convenience and comfort are provided for long-term wearing and monitoring of the equipment, the anti-interference capability is high, and the measurement can be stably carried out for a long time; can synchronously collect multiple channels of important physiological signals such as electrocardio, pulse, blood pressure and the like. The precision of multiple physiological signals such as electrocardio, pulse, heart rate and the like is compared with the synchronous acquisition of professional equipment, more accurate and rich disease judgment basis can be provided, and the wearable multiple physiological signal monitoring device is a clinical wearable multiple physiological signal monitoring device. Compared with the existing wearable equipment, the wearable equipment has more physiological signals acquired synchronously and higher accuracy; the cardiovascular disease risk is early-warned by adopting two methods, and real-time disease pre-diagnosis is carried out in the intelligent terminal equipment according to the time-frequency domain parameter calculation results of various physiological signals; and carrying out disease classification and diagnosis on multiple physiological signals of the wearer in a cloud server based on a deep learning disease intelligent classification method, and feeding back results to an intelligent terminal to carry out cardiovascular disease risk early warning. Provides more accurate reference for the assessment of the heart function and the diagnosis of heart diseases.
Drawings
Fig. 1 is a block diagram of a cardiovascular disease early warning system based on multi-physiological signal deep fusion according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a wearing device according to an embodiment of the present invention, wherein (a) is a schematic perspective structural diagram, and (b) is a schematic plan structural diagram;
fig. 3 is a schematic perspective view of a dial according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a dial provided in an embodiment of the present invention, where (a) is a schematic structural diagram of a front face of the dial, and (b) is a schematic structural diagram of a back face of the dial;
fig. 5 is a flowchart of a cardiovascular disease risk early warning method using the cardiovascular disease early warning system and apparatus based on multi-physiological signal deep fusion according to an embodiment of the present invention;
fig. 6 is a flowchart illustrating that the cloud server performs intelligent disease classification on the preprocessed multiple physiological signals according to the embodiment of the present invention;
fig. 7 is a block diagram of a depth analysis computing unit for multi-physiological signal fusion according to an embodiment of the present invention.
In the figure, 701, conductive fabric electrodes; 702. magic tape; 703. a button; 801. a pulse acquisition device; 802. a metal electrode; and a metal button 803.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
A cardiovascular disease early warning system based on multi-physiological signal deep fusion is shown in figure 1 and comprises a wearing device, a multi-physiological signal acquisition device, a transmission device, intelligent terminal equipment and a cloud server; the wearing device is worn on the wrist of the detected person, and the multiple physiological signal acquisition devices and the transmission device are arranged on the wearing device; the multi-physiological signal acquisition device is used for acquiring physiological signals of a detected person; the transmission device is used for transmitting the physiological signal data acquired by the multiple physiological signal acquisition devices to the intelligent terminal equipment; the intelligent terminal device is internally provided with a program used for processing physiological signals of a detected person, judging whether the detected person has a disease risk or not, carrying out cardiovascular disease risk early warning, assisting a doctor in diagnosing, and transmitting the processed physiological signals to the cloud server; the cloud server is internally provided with a program, and based on a deep learning disease intelligent classification method, the disease classification and diagnosis are carried out on multiple physiological signals of a wearer according to the trained disease types, and the results are fed back to the intelligent terminal;
the built-in program of the intelligent terminal comprises a signal intelligent preprocessing unit, a feature extraction unit, a parameter calculation and analysis unit and a risk early warning unit; the signal intelligent preprocessing unit is used for preprocessing the multiple physiological signals acquired by the multiple physiological signal acquisition devices and transmitting the preprocessed multiple physiological signals to the cloud server; the feature extraction unit is used for extracting feature points of various preprocessed physiological signal waveforms of the human body; the parameter calculating and analyzing unit comprises a time domain parameter calculating and analyzing module and a frequency domain parameter calculating and analyzing module; the time domain parameter calculation and analysis module calculates and analyzes the time domain parameters of the physiological signals according to the multi-physiological signal feature extraction result; the frequency domain parameter calculation and analysis module calculates and analyzes frequency domain parameters according to time-frequency analysis results of the multiple physiological signals and calculates amplitude-frequency characteristics of the multiple physiological signals after Fourier transform; the risk early warning unit carries out disease risk early warning according to real-time-frequency domain analysis results of various physiological signal parameters and feedback results of the cloud server, and assists a doctor in diagnosis;
the program built in the cloud server comprises a multi-channel time-frequency analysis unit and a depth analysis calculation unit for multi-physiological signal fusion; the multi-channel time-frequency analysis unit comprises a time domain analysis module and a frequency domain analysis module; the time domain analysis module respectively performs identity mapping and downsampling transformation on the acquired multi-physiological signal data to obtain the identity mapped multi-physiological signal data, and downsampling transformation is used for generating time sequence sketches with different time scales, so that a plurality of input time sequences with different downsampling rates are obtained; the frequency domain analysis module removes high-frequency interference and random noise from the collected multi-physiological signal data by adopting a low-frequency filter with a plurality of smoothness values, and obtains a multi-frequency input time sequence by utilizing moving averages of different windows according to different smoothness values; the multi-physiological signal fusion depth analysis and calculation unit inputs the result obtained by the multi-channel time-frequency analysis unit into a convolutional neural network for convolution operation; information is gathered in a full connection layer or other classifiers to obtain a classification result; and finally, comparing the multi-physiological-signal measurement data of the wearer with the trained features of different cardiovascular diseases to obtain a cardiovascular disease risk classification result, and feeding the result back to the intelligent terminal.
The cardiovascular disease early warning system based on the multi-physiological signal deep fusion further comprises a filtering device and an AD conversion device which are both arranged in the dial plate; the filtering device and the AD conversion device are connected with the output ends of the multiple physiological signal acquisition devices, filtering and AD conversion are carried out on output signals of the multiple physiological signal acquisition devices, and the multiple physiological signals after AD conversion are transmitted to the intelligent terminal equipment through the Bluetooth wireless receiving and transmitting module.
The cardiovascular disease early warning system based on the multi-physiological signal deep fusion further comprises a power management module and a storage module, wherein the power management module and the storage module are used for providing reference voltage for the filtering device and the AD conversion device; the power management module comprises a power management circuit and a battery; the power management circuit adopts a power management chip to realize safe charging and discharging of the lithium battery and provide +3V stable voltage, and the voltage stabilizing chip converts the +3V voltage into +2.5V voltage to supply power for the filtering device and the AD conversion device and provide stable reference voltage; the storage module selects a large-capacity storage medium to ensure that data is continuously stored for more than 24 hours, and is used for storing the original signals of the multiple physiological signals after AD conversion, the parameters of the multiple physiological signals obtained by calculation and the time-frequency domain analysis results of the multiple physiological signals.
The wearing device is shown in fig. 2 and comprises an elastic wrist band, a shell and a dial plate; the elastic wrist strap is made of soft materials, has comfort and scalability, is suitable for different wrist sizes, is bound on the left wrist, and is provided with two buttons 703 at the middle position outside the elastic wrist strap; a dial plate is arranged on the outer side of the elastic wrist strap, magic tapes 702 for fixing the elastic wrist strap are arranged at two ends of the elastic wrist strap, and the shell is arranged outside the dial plate to form a dial plate structure shown in figure 3 and used for packaging hardware circuits; the metal button 803 is arranged on the back of the dial plate, the metal button 803 is fixedly connected with the button 703 on the elastic wrist strap, and the elastic wrist strap and the dial plate are connected, so that the comfortable wearing and the long-term stable signal acquisition of a wearer are guaranteed; many physiological signal collection system fix on the elasticity wristband, can individualized adjustment bandage elasticity, and the user also can not produce uncomfortable sense for long-term the dress, and the dress travelling comfort is high to have stronger interference killing feature, can stable measurement for a long time.
The multi-physiological signal acquisition device comprises an electrocardio acquisition device and a pulse acquisition device; the electrocardio-acquisition device comprises a metal electrode 802 and two conductive fabric electrodes 701, wherein the two conductive fabric electrodes 701 and the metal electrode 802 form three electrodes for acquiring electrocardio signals; two conductive fabric electrodes 701 are respectively embedded into the inner side of the elastic wrist strap; as shown in fig. 4, a metal electrode 802 is mounted on the back of the dial; the pulse acquisition device is embedded on the surface of the shell, a photoelectric pulse blood oxygen sensor is adopted to respectively receive the reflected light intensity of two beams of red light and infrared light passing through the finger of a testee, and a photoelectric volume method is adopted to acquire pulse waves to obtain two paths of different pulse wave waveforms; the output ends of the electrocardio acquisition device and the pulse acquisition device are both connected with the transmission device. When the cardiovascular disease early warning system based on multi-physiological signal deep fusion carries out disease risk early warning, the right hand of the testee is arranged on the front surface of the dial plate, one finger tip of the right hand is arranged on the pulse acquisition device 801, and other fingers of the right hand are arranged on the metal electrode 802 of the electrocardio acquisition device, so that multi-channel physiological signals can be synchronously acquired. This equipment is small, wears the convenience, compares with current wearable equipment, and the physiological parameter who obtains is more, can provide more accurate, abundant disease judgement basis, really accomplishes accurate cardiovascular disease risk early warning, supplementary doctor's treatment. In the aspect of obtaining electrocardiosignals, the signal acquisition sensor adopting the dry electrode (conductive fabric and metal film) is different from a commonly used disposable electrode patch, and the dry electrode can be repeatedly used, so that the resource waste is avoided. Meanwhile, the conductive fabric is softer and has good fit with human skin, so that discomfort of the traditional equipment to the skin is greatly reduced, and great convenience and comfort are brought to a user needing to monitor for a long time.
The signal intelligent preprocessing unit comprises a first processing module and a second processing module, wherein the first processing module removes low-quality unreliable signals in the multi-physiological signals; the second processing module is used for carrying out signal processing on the high-quality credible signals in the multiple physiological signals; the second processing module comprises a baseline drift removing module, a power frequency interference removing module and a high-frequency interference removing module; the baseline wander removing module is used for removing baseline wander from high-quality credible signals in the multiple physiological signals to obtain physiological signals with the baseline wander removed; the power frequency interference removing module is used for removing power frequency interference from the multiple physiological signals with the baseline drift removed to obtain signals with the power frequency interference removed; and the high-frequency interference removing module is used for removing high-frequency noise interference on the multi-physiological signal without power frequency interference to obtain a signal without the high-frequency noise interference, and the signal is used as an input signal for feature extraction.
The transmission device is connected with the intelligent terminal equipment by adopting a wireless transceiving module, so that the real-time communication between the multi-physiological-signal acquisition device and the intelligent terminal equipment is realized.
In this embodiment, taking an electrocardiographic signal and a pulse signal as an example, the cardiovascular disease early warning system based on multi-physiological signal deep fusion of the present invention is used to perform cardiovascular disease risk early warning, as shown in fig. 5, and includes the following steps:
step 1, a multi-physiological signal acquisition device acquires various physiological signals of a human body, such as electrocardiosignals, pulse wave signals and the like, filtering and AD conversion are carried out through a filtering device and an AD conversion device, and the multi-physiological signals after AD conversion are wirelessly transmitted to intelligent terminal equipment;
step 2, the signal intelligent preprocessing unit carries out signal preprocessing on the collected multiple physiological signals (such as electrocardio, pulse wave and the like), and the specific method comprises the following steps:
the first processing module removes low-quality unreliable signals in multiple physiological signals (such as electrocardio, pulse and the like); specifically, taking the electrocardiographic signal as an example, if the amplitude change of the electrocardiographic signal exceeds a set threshold, the electrocardiographic signal is a low-quality unreliable signal, otherwise, the baseline drift degree is determined: if the baseline drift degree does not exceed the set threshold, the electrocardiosignal is a high-quality credible signal, otherwise, the electrocardiosignal is a low-quality credible signal;
the second processing module is used for processing high-quality credible signals in multiple physiological signals (such as electrocardio, pulse and the like); the baseline wander removing module processes multiple physiological signals (such as electrocardio, pulse and the like) with baseline wander degrees larger than a set value by adopting a mean filtering method to obtain baseline wander signals, and processes the physiological signals with baseline wander degrees smaller than the set value by adopting a curve fitting method to obtain baseline wander signals; the baseline wander signal subtracted from the original signal is used to obtain a corrected signal with baseline wander removed. The power frequency interference removing module is used for selecting a wavelet basis function to decompose the multiple physiological signals (such as electrocardio, pulse and the like) with the baseline drift removed; carrying out Fourier decomposition on the wavelet coefficients of each layer; finding wavelet coefficients of frequencies corresponding to power frequency interference of 50Hz to 60Hz, and setting the wavelet coefficients to be zero, wherein the wavelet coefficients of other frequencies are kept unchanged; and performing inverse wavelet transform according to the current wavelet coefficients, and reconstructing signals to obtain the physiological signals without power frequency interference. The high-frequency interference removing module is used for selecting a wavelet basis function to carry out multi-layer wavelet decomposition transformation on the physiological signal without power frequency interference according to the noise characteristics of various physiological signals; respectively carrying out threshold soft threshold processing on wavelet coefficients of all layers, reducing the wavelet coefficients of a high-frequency part, and removing high-frequency noise interference; and performing wavelet inverse transformation on the wavelet coefficient subjected to threshold value processing, and reconstructing a physiological signal to obtain the physiological signal without high-frequency noise interference.
Step 3, the intelligent terminal and the cloud server respectively process the preprocessed multiple physiological signals, and the specific method comprises the following steps:
the intelligent terminal processes the preprocessed multiple physiological signals and comprises the following steps:
the characteristic extraction unit extracts characteristic points of the acquired multi-physiological signal waveform;
taking electrocardiosignals and pulse wave signals as examples, extracting QRS feature points from the electrocardiosignals, determining the number n of wavelet decomposition layers according to sampling frequency, performing n-layer wavelet decomposition by using wavelet basis, and extracting and reconstructing nth-layer high-frequency signals;
detecting an electrocardiosignal R wave: setting a threshold value, ensuring that the threshold value is below the peak of the R wave and other parts of the electrocardiosignals are below the threshold value, finding the R wave larger than the threshold value at the moment, and marking the R wave on the electrocardiosignals in real time; detecting an electrocardiosignal Q wave: moving the abscissa of the R wave crest backward by a unit length, circularly detecting minimum value points in the interval, namely Q waves, and marking; detecting electrocardiosignal S wave: the abscissa of the R wave crest is moved forward by a unit length, and minimum value points are circularly detected in the interval, namely S waves, and are marked. Similar to the electrocardio signals, performing multi-layer wavelet decomposition on the pulse wave signals by using wavelet bases, and extracting and reconstructing high-frequency signals; and defining a threshold, wherein the wave crest exceeding the threshold in the high-frequency signal is a main wave P, and the wave crest lower than the threshold in the high-frequency signal is a heavy wave crest.
The parameter calculation and analysis unit calculates and analyzes the real-time-frequency domain parameters of the multi-physiological signals after the characteristics are extracted;
the time domain parameter calculation and analysis module calculates and analyzes the time domain parameters of all physiological signals according to the feature extraction results of multiple physiological signals (such as electrocardio, pulse waves and the like); the present embodiment takes the electrocardiographic signal and the pulse wave signal as an example. Calculating electrocardiosignal parameters: RR intervals, QRS complex wave width, instantaneous heart rate and most adjacent m RR interval mean values, and performing HRV time domain analysis on the electrocardiosignals; calculating pulse wave signal parameters: PP (main wave) interval, real-time blood pressure value and blood oxygen saturation, and performing PRV time domain analysis on pulse waves; finally, calculating the cardiac output per minute according to the heart rate parameter calculated by the electrocardiosignal and the stroke output parameter calculated by the pulse wave signal;
the frequency domain parameter calculating and analyzing module calculates and analyzes frequency domain parameters according to time frequency analysis results (such as time frequency analysis of HRV, PRV and the like) of the multiple physiological signals, and calculates amplitude-frequency characteristics of the multiple physiological signals after Fourier transformation. In the embodiment, by taking electrocardiosignals and pulse wave signals as examples, an HRV spectrum analysis chart is obtained by calculation according to RR interval parameters of the electrocardiosignals; and calculating a PRV spectrum analysis chart according to the pulse wave signal PP interval parameters.
In this embodiment, the time domain analysis of the HRV is performed according to the RR interval parameter of the electrocardiographic signal, specifically, the RR interval calculated by using the continuously acquired waveform of the electrocardiographic signal is used as the cardiac cycle of the electrocardiographic signal. Depicting the relation of the change of the cardiac cycle along with the time, solving a fitting function of the heart cycle by an interpolation method, reflecting the tiny difference change between the cardiac cycles, namely time domain analysis of the HRV; the time domain analysis of the PRV is carried out according to the PP interval parameters of the pulse wave signals, which is similar to the HRV analysis, and particularly, the time domain analysis of the PRV can be obtained by using the PP interval calculated by continuously obtained pulse wave signal waveforms as a cardiac cycle.
Calculating a real-time blood pressure value and blood oxygen saturation according to the pulse wave signals; the specific method for calculating the real-time blood pressure value according to the pulse wave signals comprises the steps of extracting characteristic points of a main wave peak and a heavy wave peak of the pulse wave signals, calculating characteristic parameters of the main wave rising slope, the systolic period, the diastolic period and the like of the pulse waves, estimating systolic pressure and diastolic pressure values, and displaying the values in real time.
The specific method for calculating the blood oxygen saturation according to the pulse wave signal comprises the following steps: according to the reflected light intensity of two beams of red light and infrared light passing through the finger of a subject and the different absorption degrees of oxygenated hemoglobin and hemoglobin to different light, the ratio of the absorption amount of the infrared light to the absorption amount of the red light is measured by a spectrophotometry method, and the blood oxygen saturation is calculated.
The cloud server carries out disease intelligent classification on the preprocessed multiple physiological signals, as shown in fig. 6, including:
the time domain analysis module respectively performs identity mapping and downsampling transformation on the acquired multi-physiological signal data (such as electrocardio, pulse and the like) to obtain the identity mapped multi-physiological signal data, and uses the downsampling transformation to generate time sequence sketches with different time scales, so as to obtain a plurality of input time sequences with different downsampling rates.
The frequency domain analysis module removes high-frequency interference and random noise from the collected multi-physiological signal data (such as electrocardio and pulse wave) by adopting a low-frequency filter with a plurality of smoothness values, and obtains a multi-frequency input time sequence by utilizing the moving average of different windows according to different smoothness values.
The depth analysis computing unit for multi-physiological signal fusion, as shown in fig. 7, obtains 3 kinds of transformation data: the multi-physiological signals after the identity mapping, the multi-scale time sequence after the downsampling transformation and the multi-frequency time sequence after the frequency spectrum transformation are used as the input of 3 parallel convolution neural networks, transformation data is input into one convolution neural network, each neural network is used as a feature learning tool, and features obtained through deep analysis are extracted by utilizing downsampling. And integrating the data and then sending the data to a full connection layer or other classifiers for classification, and obtaining the classification accuracy. In the embodiment, the arrhythmia is classified according to the electrocardiosignals, the convolutional neural network with proper parameters and depth is designed, the electrocardiosignal training data containing normal heartbeat and arrhythmia (bundle branch block, atrial premature beat, ventricular premature beat and the like) is divided into a training set and a testing set of a convolutional neural network model, part of samples are randomly selected as the training set, the rest of samples are used as the testing set to be trained in a supervision mode, feature learning and classification are carried out by utilizing deep learning, the training is stopped when the accuracy rate reaches the expectation, and the classification result is obtained. And carrying out disease classification on arteriosclerosis according to the pulse wave signals, designing a convolutional neural network model to carry out classification training on arteriosclerosis diseases on normal pulse wave signals and arteriosclerosis pulse wave signals, and obtaining a classification result. And finally, feeding back the disease classification and diagnosis results to the intelligent terminal.
4, a risk early warning unit in the intelligent terminal device carries out disease risk early warning according to real-time-frequency domain analysis results of various physiological signal parameters and classification results of the cloud server;
in this embodiment, according to the RR interval, the QRS complex wave width, the instantaneous heart rate, and the average value of the nearest m RR intervals of the electrocardiographic signal, the following rules are used to perform cardiovascular disease risk early warning:
(a) if at least one of the following conditions is satisfied: 1) the RR interval is less than or equal to p times of the mean value of the nearest m RR intervals (according to the practical situation, p can be about 0.8); 2) if the QRS complex wave width is more than or equal to t seconds (for example, t can be set to be about 0.12 seconds), ventricular premature beat risk early warning is carried out;
(b) if at least one of the following conditions is satisfied: 1) the RR intervals are less than or equal to p times of the mean value of the nearest m (such as 5, 7 and 9) RR intervals; 2) if the QRS complex wave width is less than t seconds, performing atrial premature beat risk early warning;
(c) if at least one of the following conditions is satisfied: 1) p times of the mean value of the nearest m RR intervals is less than the RR intervals; 2) the RR interval is less than or equal to p times (q can be about 1.2 according to actual conditions) of the mean value of m nearest RR intervals; 3) if the QRS complex wave width is more than or equal to t seconds, performing restraint branch block heart beat risk early warning;
(d) except the above conditions, sinus heartbeat risk early warning is carried out;
according to the heart beat classification, risk early warning can be further carried out on heart rate rhythms;
sinus rhythm: all are sinus beats;
nodal tachycardia: more than three sinus heartbeats, wherein the heart rate of each heart beat is more than 120;
sinus bradycardia: more than three sinus heartbeats, and the heart rate of each heart beat is less than 50;
sinus arrest: the RR intervals of two sinus beats exceed a certain time (generally, the RR interval is more than 1.6 seconds);
in the morning of the house: one or more atrial premature beats;
in the morning of the room: one or more ventricular premature beats;
paired chamber early stage: two consecutive ventricular premature beats;
bundle branch block: three or more continuous branch block heartbeats appear;
ventricular premature beat bigeminy: each sinus heartbeat is followed by a ventricular premature beat and is repeated at least three times;
triple ventricular premature beat law: every two sinus beats are followed by one ventricular premature beat or every sinus beat is followed by two ventricular premature beats, and the steps are repeated at least twice;
atrial tachycardia: three or more atrial premature beat heartbeats continuously appear, and the heart rate of each beat is more than 120;
ventricular tachycardia: three or more ventricular premature beats occur in succession, with a heart rate of each beat greater than 120.
According to the real-time blood pressure value and the blood oxygen saturation of the pulse wave signal, carrying out disease risk early warning: and when the real-time blood pressure value or the blood oxygen saturation exceeds the set normal blood pressure range or the set normal blood oxygen saturation at n continuous sampling moments, carrying out blood pressure abnormity early warning or blood oxygen saturation abnormity early warning.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (7)
1. A cardiovascular disease early warning system based on multi-physiological signal deep fusion is characterized in that: the system comprises a wearing device, a multi-physiological signal acquisition device, a transmission device, intelligent terminal equipment and a cloud server; the wearing device is worn on the wrist of the detected person, and the multi-physiological-signal acquisition device and the transmission device are both arranged on the wearing device; the multi-physiological signal acquisition device is used for acquiring physiological signals of a detected person; the transmission device is used for transmitting the physiological signal data acquired by the multiple physiological signal acquisition devices to the intelligent terminal equipment; the intelligent terminal device is internally provided with a program used for processing physiological signals of a detected person, judging whether the detected person has cardiovascular disease risk or not, carrying out disease risk early warning, assisting a doctor to diagnose and transmitting the processed physiological signals to the cloud server; the cloud server is internally provided with a program, and based on a deep learning disease intelligent classification method, the disease classification and diagnosis are carried out on multiple physiological signals of a wearer according to the trained disease types, and the results are fed back to the intelligent terminal;
the multi-physiological signal acquisition device comprises an electrocardio acquisition device and a pulse acquisition device; the electrocardiosignal acquisition device comprises a metal electrode and two conductive fabric electrodes, wherein the two conductive fabric electrodes and the metal electrode form three electrodes for acquiring electrocardiosignals; the pulse acquisition device adopts a photoelectric pulse blood oxygen sensor to respectively receive the reflected light intensity of two beams of red light and infrared light passing through the finger of a subject, and adopts a photoelectric volume method to acquire pulse waves to obtain two paths of different pulse wave waveforms; the output ends of the electrocardio acquisition device and the pulse acquisition device are both connected with a transmission device;
the built-in programs of the intelligent terminal comprise a signal intelligent preprocessing unit, a feature extraction unit, a parameter calculation and analysis unit and a risk early warning unit; the signal intelligent preprocessing unit is used for preprocessing multiple physiological signals acquired by the multiple physiological signal acquisition devices and transmitting the preprocessed multiple physiological signals to the cloud server; the feature extraction unit is used for extracting feature points of various preprocessed human physiological signal waveforms; extracting QRS characteristic points from the electrocardiosignal, determining the number n of wavelet decomposition layers according to the sampling frequency, performing n-layer wavelet decomposition by using a wavelet base, and extracting and reconstructing an nth-layer high-frequency signal; detecting an electrocardiosignal R wave: setting a threshold value, ensuring that the threshold value is below the peak of the R wave, finding the R wave larger than the threshold value, and marking the R wave on the electrocardiosignal in real time; detecting an electrocardiosignal Q wave: moving the abscissa of the R wave crest backward by a unit length, circularly detecting minimum value points in the interval, namely Q waves, and marking; detecting electrocardiosignal S wave: moving the horizontal coordinate of the R wave crest forward by a unit length, circularly detecting minimum value points in the interval, namely S waves, and marking; similar to the electrocardio signals, performing multi-layer wavelet decomposition on the pulse wave signals by using wavelet bases, and extracting and reconstructing high-frequency signals; defining a threshold, wherein a wave crest exceeding the threshold in the high-frequency signal is a main wave P, and a wave crest lower than the threshold in the high-frequency signal is a heavy wave crest; the parameter calculating and analyzing unit comprises a time domain parameter calculating and analyzing module and a frequency domain parameter calculating and analyzing module; the time domain parameter calculation and analysis module calculates and analyzes the time domain parameters of the physiological signals according to the multi-physiological signal feature extraction result; the frequency domain parameter calculation and analysis module calculates and analyzes frequency domain parameters according to time-frequency analysis results of the multiple physiological signals and calculates amplitude-frequency characteristics of the multiple physiological signals after Fourier transform; the risk early warning unit carries out disease risk early warning according to real-time-frequency domain analysis results of various physiological signal parameters and feedback results of the cloud server, and assists a doctor in diagnosis;
the program built in the cloud server comprises a multi-channel time-frequency analysis unit and a depth analysis calculation unit for multi-physiological signal fusion; the multi-channel time-frequency analysis unit comprises a time domain analysis module and a frequency domain analysis module; the time domain analysis module respectively performs identity mapping and downsampling transformation on the acquired multi-physiological signal data to obtain the identity mapped multi-physiological signal data, and downsampling transformation is used for generating time sequence sketches with different time scales, so that a plurality of input time sequences with different downsampling rates are obtained; the frequency domain analysis module removes high-frequency interference and random noise from the collected multi-physiological signal data by adopting a low-frequency filter with a plurality of smoothness values, and obtains a multi-frequency input time sequence by utilizing moving averages of different windows according to different smoothness values; the multi-physiological signal fusion depth analysis and calculation unit inputs the result obtained by the multi-channel time-frequency analysis unit into a convolutional neural network for convolution operation; information is gathered in the full connection layer to obtain a classification result; and finally, comparing the multi-physiological-signal measurement data of the wearer with the trained features of different cardiovascular diseases to obtain a cardiovascular disease risk classification result, and feeding the result back to the intelligent terminal.
2. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the wearing device comprises an elastic wrist strap, a shell and a dial plate; the elastic wrist strap is made of soft materials, and two buttons are arranged in the middle of the outer side of the elastic wrist strap; a dial plate is arranged on the outer side of the elastic wrist strap, and magic tapes for fixing the elastic wrist strap are arranged at two ends of the elastic wrist strap; the casing is arranged outside the dial plate to form a dial plate structure and is used for packaging a hardware circuit; the back of the dial plate is provided with a metal buckle, and the metal buckle is fixedly connected with a button on the elastic wrist strap to connect the elastic wrist strap and the dial plate; the multiple physiological signal acquisition devices are fixed on the elastic wristbands.
3. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the signal intelligent preprocessing unit comprises a first processing module and a second processing module; the first processing module removes low-quality untrusted signals in the multi-physiological signals; the second processing module is used for carrying out signal processing on a high-quality credible signal in the multiple physiological signals; the second processing module comprises a baseline drift removing module, a power frequency interference removing module and a high-frequency interference removing module; the baseline wander removing module is used for removing baseline wander from high-quality credible signals in the multiple physiological signals to obtain physiological signals with the baseline wander removed; the power frequency interference removing module is used for removing power frequency interference from the multiple physiological signals with the baseline drift removed to obtain signals with the power frequency interference removed; and the high-frequency interference removing module is used for removing high-frequency noise interference on the multi-physiological signal without power frequency interference to obtain a signal without the high-frequency noise interference, and the signal is used as an input signal for feature extraction.
4. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the transmission device is connected with the intelligent terminal equipment by adopting a wireless transceiving module, so that real-time communication between the multi-physiological-signal acquisition device and the intelligent terminal equipment is realized.
5. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the cardiovascular disease early warning system based on the multi-physiological signal deep fusion further comprises a filtering device and an AD conversion device which are both arranged in the dial plate; the filtering device and the AD conversion device are connected with the output end of the multi-physiological signal acquisition device, filtering and AD conversion are carried out on the output signals of the multi-physiological signal acquisition device, and the multi-physiological signals after AD conversion are transmitted to the intelligent terminal equipment through the wireless transceiving module.
6. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the cardiovascular disease early warning system based on the multi-physiological signal deep fusion also comprises a power supply management module for providing reference voltage for the filtering device and the AD conversion device; the power management module comprises a power management circuit and a battery; the power management circuit adopts the power management chip to realize the safe charge and discharge of lithium cell to provide +3V regulated voltage, the steady voltage chip converts +3V voltage into +2.5V voltage, for filter equipment and AD conversion device power supply and provide stable reference voltage.
7. The cardiovascular disease early warning system based on multi-physiological signal deep fusion of claim 1, characterized in that: the cardiovascular disease early warning system based on the multi-physiological signal deep fusion also comprises a storage module; the storage module selects a large-capacity storage medium, ensures that data is continuously stored for more than 24 hours, and is used for storing the AD-converted multi-physiological-signal original signals, the calculated multi-physiological-signal parameters and the time-frequency-domain analysis results of the multi-physiological signals.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811085130.4A CN109157202B (en) | 2018-09-18 | 2018-09-18 | Cardiovascular disease early warning system based on multi-physiological signal deep fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811085130.4A CN109157202B (en) | 2018-09-18 | 2018-09-18 | Cardiovascular disease early warning system based on multi-physiological signal deep fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109157202A CN109157202A (en) | 2019-01-08 |
| CN109157202B true CN109157202B (en) | 2021-06-01 |
Family
ID=64879608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811085130.4A Active CN109157202B (en) | 2018-09-18 | 2018-09-18 | Cardiovascular disease early warning system based on multi-physiological signal deep fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109157202B (en) |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9833607B2 (en) | 2014-10-30 | 2017-12-05 | West Affum Holdings Corp. | Wearable cardiac defibrillation system with flexible electrodes |
| US11540762B2 (en) | 2014-10-30 | 2023-01-03 | West Affum Holdings Dac | Wearable cardioverter defibrtillator with improved ECG electrodes |
| CN109645963A (en) * | 2019-02-25 | 2019-04-19 | 深圳奥思顿健康科技有限公司 | It is a kind of for detecting the intelligent analysis system of vital sign and diagnosis by feeling the pulse |
| CN110037668B (en) * | 2019-04-10 | 2022-02-01 | 常熟理工学院 | System for judging age, health state and malignant arrhythmia identification by combining pulse signal time-space domain with model |
| CN110236906A (en) * | 2019-05-20 | 2019-09-17 | 安徽通灵仿生科技有限公司 | A kind of reaction type multi-modal data fusion synchronization accessory heart control system |
| CN110192842A (en) * | 2019-05-30 | 2019-09-03 | 华南理工大学 | System and method is monitored based on the integrated data of physiological index of cloud side end |
| CN110200621A (en) * | 2019-06-06 | 2019-09-06 | 中山大学孙逸仙纪念医院 | A kind of cardiovascular event early warning analysis method |
| CN112263249A (en) * | 2019-07-08 | 2021-01-26 | 复旦大学附属中山医院 | ECG-based method and device for enhancing blood oxygen saturation monitoring |
| CN110584599A (en) * | 2019-08-07 | 2019-12-20 | 王满 | Wavelet transformation data processing system and method based on cardiac function dynamic monitoring |
| CN111419214A (en) * | 2020-03-31 | 2020-07-17 | 联想(北京)有限公司 | Electrocardio abnormality detection method, terminal and server |
| CN111248877A (en) * | 2020-04-18 | 2020-06-09 | 赵宏杰 | Myocardial infarction early warning system |
| CN111839480B (en) * | 2020-07-14 | 2024-03-29 | 广州智康云科技有限公司 | Detection processing system and method of robot and robot |
| CN111956215A (en) * | 2020-07-29 | 2020-11-20 | 鲁东大学 | QRS point detection method of low-quality electrocardiogram |
| WO2022041225A1 (en) * | 2020-08-31 | 2022-03-03 | 中山大学 | Server for use in assessing cardiovascular state, wearable device, and method for cardiovascular state assessment |
| CN112505798B (en) * | 2020-11-27 | 2022-07-01 | 河北雄安太芯电子科技有限公司 | Object detection method based on terahertz |
| CN112657050A (en) * | 2020-12-01 | 2021-04-16 | 北京工业大学 | Artificial heart control system based on Internet of things edge calculation and operation method |
| WO2023019467A1 (en) * | 2021-08-18 | 2023-02-23 | 中国人民解放军总医院第二医学中心 | Glove for measuring multiple physiological parameters, and system for detecting risk of suffering from hypertensive disease |
| CN113892926A (en) * | 2021-09-08 | 2022-01-07 | 杭州回车电子科技有限公司 | Heart rate signal processing method and device based on head pulse wave and electronic device |
| CN113876308A (en) * | 2021-11-02 | 2022-01-04 | 南华大学 | Intelligent internet-of-things cardiovascular disease medical health early warning system |
| CN114224348A (en) * | 2021-11-29 | 2022-03-25 | 江南大学 | Combined neural network electrocardiogram classification method and system |
| CN117224092B (en) * | 2023-11-16 | 2024-02-09 | 常熟理工学院 | Photoelectric volume pulse wave interference band real-time detection method and system based on decision tree |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002165768A (en) * | 2000-12-04 | 2002-06-11 | Casio Comput Co Ltd | Measurement device to be set on the arm |
| CN101264011A (en) * | 2007-03-12 | 2008-09-17 | 三星电子株式会社 | Method and apparatus for cufflessly and non-invasively measuring wrist blood pressure |
| EP2542147A2 (en) * | 2010-03-04 | 2013-01-09 | Neumitra LLC | Devices and methods for treating psychological disorders |
| CN204009405U (en) * | 2014-09-04 | 2014-12-10 | 康泰医学***（秦皇岛）股份有限公司 | The intelligent watch of collection passometer, electrocardio, blood-oxygen functional |
| CN105708441A (en) * | 2014-12-03 | 2016-06-29 | 中国科学院苏州纳米技术与纳米仿生研究所 | Wearable finger stall and electrocardiogram and pulse condition acquisition terminal |
| CN106066933A (en) * | 2016-05-27 | 2016-11-02 | 成都信汇聚源科技有限公司 | The method obtaining atrial fibrillation identification artificial neural network weighted value matrix |
| CN107334466A (en) * | 2017-08-08 | 2017-11-10 | 西安交通大学 | A kind of apparatus and method of wearable chronic disease intelligent monitoring and early warning |
| CN107510452A (en) * | 2017-09-30 | 2017-12-26 | 扬美慧普（北京）科技有限公司 | A kind of ECG detecting method based on multiple dimensioned deep learning neutral net |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103284702A (en) * | 2013-06-17 | 2013-09-11 | 中国科学院苏州纳米技术与纳米仿生研究所 | Electrocardiogram and pulse wave relation analysis method and method and device of fusion analysis |
| CN104545844B (en) * | 2014-12-25 | 2017-02-22 | 中国科学院苏州生物医学工程技术研究所 | Multi-parameter sleep monitoring and intelligent diagnosis system based on 4G mobile communication technology and application method of multi-parameter sleep monitoring and intelligent diagnosis system |
| CN104706349B (en) * | 2015-04-13 | 2017-03-22 | 大连理工大学 | Electrocardiosignal construction method based on pulse wave signals |
-
2018
- 2018-09-18 CN CN201811085130.4A patent/CN109157202B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002165768A (en) * | 2000-12-04 | 2002-06-11 | Casio Comput Co Ltd | Measurement device to be set on the arm |
| CN101264011A (en) * | 2007-03-12 | 2008-09-17 | 三星电子株式会社 | Method and apparatus for cufflessly and non-invasively measuring wrist blood pressure |
| EP2542147A2 (en) * | 2010-03-04 | 2013-01-09 | Neumitra LLC | Devices and methods for treating psychological disorders |
| CN204009405U (en) * | 2014-09-04 | 2014-12-10 | 康泰医学***（秦皇岛）股份有限公司 | The intelligent watch of collection passometer, electrocardio, blood-oxygen functional |
| CN105708441A (en) * | 2014-12-03 | 2016-06-29 | 中国科学院苏州纳米技术与纳米仿生研究所 | Wearable finger stall and electrocardiogram and pulse condition acquisition terminal |
| CN106066933A (en) * | 2016-05-27 | 2016-11-02 | 成都信汇聚源科技有限公司 | The method obtaining atrial fibrillation identification artificial neural network weighted value matrix |
| CN107334466A (en) * | 2017-08-08 | 2017-11-10 | 西安交通大学 | A kind of apparatus and method of wearable chronic disease intelligent monitoring and early warning |
| CN107510452A (en) * | 2017-09-30 | 2017-12-26 | 扬美慧普（北京）科技有限公司 | A kind of ECG detecting method based on multiple dimensioned deep learning neutral net |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109157202A (en) | 2019-01-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109157202B (en) | Cardiovascular disease early warning system based on multi-physiological signal deep fusion | |
| Bolanos et al. | Comparison of heart rate variability signal features derived from electrocardiography and photoplethysmography in healthy individuals | |
| US7025729B2 (en) | Apparatus for detecting sleep apnea using electrocardiogram signals | |
| CN106725382A (en) | Sleep state judgement system and method based on action and HRV measurements | |
| GB2454960A (en) | Data recording for patient status analysis | |
| KR20120108575A (en) | Apparatus and method improving accuracy of wrist blood pressure by using multiple bio-signal | |
| CN103445767B (en) | The full-automatic autonomic nervous function detector of sensor monitoring interactive controlling | |
| CN110236508A (en) | A kind of non-invasive blood pressure continuous monitoring method | |
| CN107530015B (en) | Vital sign analysis method and system | |
| CN109288515A (en) | Periodical monitoring method and device based on premature beat signal in wearable ECG signal | |
| CN110881967A (en) | Non-invasive multi-segment peripheral arterial vessel elastic function detection method and instrument thereof | |
| CN107582040B (en) | Method and device for monitoring heart rhythm | |
| CN115299899A (en) | Activity recognition and beat-to-beat blood pressure monitoring, analyzing and early warning system based on multiple sensors | |
| CN210408412U (en) | Portable dynamic cardiovascular parameter acquisition equipment | |
| Qiu et al. | A wireless wearable sensor patch for the real-time estimation of continuous beat-to-beat blood pressure | |
| CN114903445A (en) | Intelligent monitoring and early warning system for cardiovascular and cerebrovascular diseases | |
| von Rosenberg et al. | Smart helmet: monitoring brain, cardiac and respiratory activity | |
| CN114343596A (en) | Continuous blood pressure measuring system and method based on reverse blood flow propagation time | |
| CN111887829A (en) | Human health information detection and evaluation system and operation method thereof | |
| Such et al. | On-body sensors for personal healthcare | |
| Paket et al. | ECG Monitoring System Using Textile Electrodes | |
| CN210095711U (en) | Noninvasive continuous blood pressure measuring equipment | |
| KH et al. | Wearable ECG electrodes for detection of heart rate and arrhythmia classification | |
| CN112535483B (en) | Heart sympathetic activity detection device based on high-frequency electrocardio | |
| Mao et al. | Motion Artifact Reduction In Photoplethysmography For Reliable Signal Selection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |