CN112790748A - Central arterial pressure waveform reconstruction system and method - Google Patents
Central arterial pressure waveform reconstruction system and method Download PDFInfo
- Publication number
- CN112790748A CN112790748A CN202011613837.5A CN202011613837A CN112790748A CN 112790748 A CN112790748 A CN 112790748A CN 202011613837 A CN202011613837 A CN 202011613837A CN 112790748 A CN112790748 A CN 112790748A
- Authority
- CN
- China
- Prior art keywords
- output
- layer
- long
- pressure
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000004872 arterial blood pressure Effects 0.000 title claims abstract description 97
- 238000000034 method Methods 0.000 title claims abstract description 83
- 230000015654 memory Effects 0.000 claims abstract description 84
- 230000002093 peripheral effect Effects 0.000 claims abstract description 69
- 238000004364 calculation method Methods 0.000 claims abstract description 48
- 238000012937 correction Methods 0.000 claims abstract description 33
- 210000004027 cell Anatomy 0.000 claims description 42
- 210000001367 artery Anatomy 0.000 claims description 41
- 238000012545 processing Methods 0.000 claims description 37
- 238000013528 artificial neural network Methods 0.000 claims description 31
- 210000002302 brachial artery Anatomy 0.000 claims description 24
- 230000007787 long-term memory Effects 0.000 claims description 22
- 238000005259 measurement Methods 0.000 claims description 22
- 230000006403 short-term memory Effects 0.000 claims description 22
- 230000035487 diastolic blood pressure Effects 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 20
- 210000002321 radial artery Anatomy 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000012544 monitoring process Methods 0.000 claims description 7
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000003321 amplification Effects 0.000 claims description 4
- 210000001715 carotid artery Anatomy 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 4
- 230000010355 oscillation Effects 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 230000017531 blood circulation Effects 0.000 claims description 3
- 238000012886 linear function Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 239000010410 layer Substances 0.000 description 89
- 238000012546 transfer Methods 0.000 description 17
- 230000036772 blood pressure Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 230000035488 systolic blood pressure Effects 0.000 description 8
- 230000000747 cardiac effect Effects 0.000 description 7
- 230000002146 bilateral effect Effects 0.000 description 6
- 230000035485 pulse pressure Effects 0.000 description 5
- 210000000709 aorta Anatomy 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 2
- 210000001105 femoral artery Anatomy 0.000 description 2
- 230000000004 hemodynamic effect Effects 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001575 pathological effect Effects 0.000 description 2
- 239000002356 single layer Substances 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 208000024172 Cardiovascular disease Diseases 0.000 description 1
- 208000001953 Hypotension Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 230000007211 cardiovascular event Effects 0.000 description 1
- 208000015606 cardiovascular system disease Diseases 0.000 description 1
- 230000004087 circulation Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 206010012601 diabetes mellitus Diseases 0.000 description 1
- 230000002526 effect on cardiovascular system Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 230000036543 hypotension Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000011514 reflex Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—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
- A61B5/021—Measuring pressure in heart or blood vessels
- A61B5/022—Measuring pressure in heart or blood vessels by applying pressure to close blood vessels, e.g. against the skin; Ophthalmodynamometers
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
Abstract
The invention discloses a central arterial pressure waveform reconstruction system and a method, wherein a peripheral arterial pulse wave measuring module of the reconstruction system sends measured peripheral arterial pulse waves to a pulse wave correcting module for correction, and outputs corrected pulse wave data to a central arterial pressure reconstruction calculating module; the central arterial pressure reconstruction computing module consists of an input layer, a computing layer and an output layer; the input layer receives the corrected peripheral arterial pulse wave and inputs the wave to the calculation layer, the calculation layer is composed of a plurality of long and short time memory units, and each long and short time memory unit receives the signal of the input layer at the current moment and the output of the long and short time memory unit at the last moment; each long and short time memory unit is output to the output layer and is simultaneously used as the input of the long and short time memory unit at the next time; the output layer outputs the finally reconstructed central artery pressure pulse wave to the data display module. The invention can automatically realize the end-to-end central arterial pressure waveform reconstruction, and the reconstruction accuracy rate is obviously superior to that of the traditional method.
Description
Technical Field
The invention relates to a central artery pressure waveform reconstruction system and a central artery pressure waveform reconstruction method, in particular to a system and a method for realizing non-invasive reconstruction of a peripheral artery to a central aorta blood pressure waveform by using a bilateral long-time memory artificial neural network, and belongs to the crossing field of artificial intelligence and medical instrument technology.
Background
The central artery pressure waveform refers to the curve of the ascending aorta root blood pressure along with the time change, and the curve contains rich human body physiological and pathological information. By extracting the parameter indexes of the central arterial pressure waveform, such as systolic pressure, diastolic pressure, pressure equalizing, pulse pressure, reflex intensity, enhancement index, ejection time, cardiac output and the like, the physiological and pathological conditions of the testee can be evaluated and predicted. Moreover, some clinical studies have shown that: compared with peripheral arterial blood pressure, the central arterial blood pressure can reflect cardiac load more and is more closely related to occurrence of cardiovascular events. Therefore, the accurate reconstruction of the central arterial pressure waveform is an important premise for extracting the parameter indexes, is the key for accurately monitoring the cardiovascular system diseases, and has very important academic value and clinical application significance.
Currently, the "gold standard" of the central arterial pressure waveform measurement method is to use a catheter to intervene in peripheral arteries, such as the radial artery and the femoral artery, and to extend a sensor inside the blood vessel to the vicinity of the root of the ascending aorta to perform continuous blood pressure waveform measurement. The catheter intervention method has certain invasiveness, is mainly used in specific occasions such as emergency treatment, cardiovascular intervention operation, intensive care and the like, and has certain limitation in application scenes. For this reason, some noninvasive methods are proposed one after another and applied to some medical device systems. The method mainly comprises the following steps: transfer function method, N-point moving average method, wave division method, blind source system identification method, etc. The system devices related to the methods mainly acquire blood pressure signals of finger artery, radial artery, carotid artery, brachial artery or femoral artery through a pressure sensor or a sensor in an inflatable cuff, then establish a physical model or a mathematical model to analyze and diagnose the data, and measure and obtain the blood pressure waveform of the central artery.
The transfer function method is the earliest non-invasive central artery pressure waveform reconstruction method. This method was first proposed by Karamanoglu et al in European Heart Journal in 1993 by acquiring radial and central arterial pressure waveforms, then taking a fourier transform and dividing the two to obtain a transfer function between radial and central arterial pressure, and averaging all cases to obtain an average or universal transfer function. The transfer function was applied to the fourier transform result of the radial artery of any one case and inverse fourier transform was performed, thereby calculating the central arterial pressure waveform. In 1997, Chen et al proposed a new transfer function method based on the autoregressive model ARX in the journal of Circulation, with better results. This method has been used in the development of the SphygmoCor arterial pulse wave analysis system by atcorpedical, australia. Since these scholars only propose a universal transfer function and do not show individual differences, the accuracy rate of the transfer function is not ideal in clinical experiments of some populations (such as sports population, diabetes, hypotension and the like). In order to improve accuracy, some scholars propose an individualized transfer function method. Some of the methods realize individuation of a transfer function by using specific parameters obtained by a radial artery descending branch, and some methods realize parameter estimation of a model by establishing mathematical models such as a lumped model, an elastic cavity model, a transmission line model, an electric network model and the like of a human artery system, and then simulate or calculate a central arterial pressure waveform by using the model. Such methods require certain hemodynamic knowledge to construct and estimation of model parameters introduces a potential risk of error. Based on this, commercialization of individualized transfer functions has not been fully realized.
The N-point moving average method is a convenient central arterial pressure method proposed in the Journal of the American College by Williams et al 2011. The method adopts rectangular waves with certain window width to carry out moving average on radial artery so as to obtain central arterial pressure, and the precision reaches the level of a universal transfer function method. In the ieee transactions on biological Engineering journal, we published in 2018 proof and experiments that the N-point moving average method is a special general transfer function method, indicating that window adjustment substantially modifies transfer time and reflection coefficient of the transfer function. Although the N-point moving average method is simple and efficient to implement and has good noise resistance, the method can only obtain the systolic pressure of the central arterial pressure, and cannot obtain waveforms and other parameters.
The wave division method is a new method proposed by Stergiopulos and Westerhof et al in 1998. According to the method, a radial artery pressure waveform is decomposed into incident waves and reflected waves, then the pulse wave propagation time is calculated, the reflected waves are translated by utilizing the time and are superposed with the incident waves, and therefore a central artery pressure waveform is reconstructed. The method also requires the establishment of an on-line model of the artery and parameter estimation.
The blind system identification method proposed by Swamy et al in 2007, which reconstructs the central arterial pressure waveform through two or more peripheral pulse waves. The method needs to measure two or more paths of signals, is inconvenient for practical clinical application and limits the application and popularization of the method.
In conclusion, compared with the wave division method, the transfer function method has better application prospect in the N-point moving average method and the blind system identification method. However, most of the two methods propose a simulation model with good physiological significance on the basis of hemodynamics, then perform parameter estimation of the model, and finally perform central arterial pressure waveform calculation and reconstruction, so that the methods have the problems of difficult simulation model establishment, inaccurate parameter estimation, large reconstruction error and the like.
Disclosure of Invention
The invention provides a central artery pressure waveform reconstruction system and a method, aiming at the problems of difficult simulation model establishment, inaccurate parameter estimation, larger reconstruction error and the like faced by the existing noninvasive central artery pressure waveform reconstruction method and device.
The technical scheme of the invention is realized as follows:
a central arterial pressure waveform reconstruction system comprises a data acquisition control module, a peripheral arterial pulse wave measuring module, a pulse wave correcting module, a central arterial pressure reconstruction calculating module and a data display module;
the data acquisition control module is respectively connected with the peripheral artery pulse wave measurement module and the pulse wave correction module and is used for controlling the measurement and processing sequence of the two modules, including the acquisition of peripheral artery pulse wave signals and the start and stop control of the pulse wave correction module, controlling the data output of the pulse wave correction module to the central artery pressure reconstruction calculation module and outputting the reconstruction result to the display module for final display;
the peripheral artery pulse wave measuring module is used for measuring peripheral artery pulse waves and comprises a pulse signal sensor, a lead and a signal processing circuit, wherein the pulse signal sensor obtains pulse signals of blood flow in an artery, the pulse signal sensor transmits the signals to the signal processing circuit through the lead, the signal processing circuit carries out preprocessing on the pulse signals, the preprocessing comprises signal processing such as amplification and filtering, and the processed results are sent to the pulse wave correcting module;
the pulse wave correction module consists of a cuff, an air channel conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, wherein the cuff is connected with the air channel conduit, the air channel conduit is connected with the inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air channel conduit, and the air pressure sensor is arranged in the cuff or the air channel conduit and is used for monitoring the air pressure change in the cuff or the air channel conduit; the method comprises the following steps that an oscillometric method is adopted in a brachial artery signal processing and calculating module to measure the average pressure and diastolic pressure of brachial arteries, the average pressure and diastolic pressure are used for correcting waveforms input by a peripheral artery pulse wave measuring module, namely, stretching processing of equal average pressure and diastolic pressure is carried out on peripheral artery pulse waves, and the pulse wave correcting module outputs corrected pulse wave data to a central artery pressure reconstruction calculating module;
the central arterial pressure reconstruction computing module consists of an input layer, a computing layer and an output layer; the input layer receives the corrected peripheral arterial pulse waves and serially inputs sampling points of the pulse waves into the calculation layer one by one, the calculation layer is composed of a plurality of long and short time memory units, each long and short time memory unit establishes a forward self-circulation connection or a reverse self-circulation connection, and the number of the long and short time memory units in the forward self-circulation connection and the number of the reverse self-circulation connectionThe number of the long and short time memory units is the same and the memory units are paired one by one; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; the output layer is composed of fully-connected shallow neural networks, and the shallow neural networks output finally reconstructed central artery pressure pulse waves to the data display module one by one according to the sequence of sampling points;
and the data display module receives the output of the central arterial pressure reconstruction calculation module and displays the relevant waveforms and parameters.
The pulse signal sensor is a piezoelectric sensor or a photoelectric sensor.
The computing layer is formed by sequentially connecting one or more levels of sub computing layers, each level of sub computing layer is the same in structure and is formed by a forward layer and a reverse layer, the forward layer comprises one or more forward self-circulation connections, the reverse layer comprises one or more reverse self-circulation connections, the number of the forward self-circulation connections in the same level of sub computing layer is consistent with the number of the reverse self-circulation connections in the same level of sub computing layer and corresponds to the number of the reverse self-circulation connections one by one, the last level of sub computing layer outputs to the output layer, and the output of each level of sub computing layer comprises the output of all long and short time memory units forming the forward self-circulation connections in the level of sub computing layer and the output of all long and short time memory units forming the reverse self-circulation connections.
The invention also provides a central artery pressure waveform reconstruction method which comprises the following steps,
1) measuring peripheral arterial pulse waves and correcting;
2) sending the corrected peripheral arterial pulse waves into a central arterial pressure reconstruction calculation module for reconstruction processing, wherein the central arterial pressure reconstruction calculation module consists of an input layer, a calculation layer and an output layer; the input layer receives the corrected peripheral arterial pulse wave and serially inputs the sampling points of the pulse wave to the calculating layer one by one, and the calculating layer records a plurality of time periodsThe memory units are formed, each long-time memory unit establishes a forward self-circulation connection or a reverse self-circulation connection, and the number of the long-time memory units in the forward self-circulation connection is the same as that of the long-time memory units in the reverse self-circulation connection and the long-time memory units in the reverse self-circulation connection are matched one by one; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; the output layer is composed of fully-connected shallow neural networks, and the shallow neural networks output finally reconstructed central artery pressure pulse waves one by one according to the sequence of sampling points;
the specific reconstruction process is
2.1) forgetting to leave the ratio f of useful information in the memory of this unittAs calculated in the formula (1),
ft=σ(Wf[ht-1,PAPWt]+bf) (1)
wherein, WfWeight matrix for forgetting gate, ht-1For memorizing the output of the cell for the duration of the last moment in the self-circulating connection, PAPWtPeripheral arterial pulse wave input at time t, bfA deviation item for a forgotten door; sigma refers to a Sigmoid activation function, and takes a value from 0 to 1;
2.2) screening and retaining newly input peripheral arterial pulse wave information itIndicating, acquiring the state of an input cellGet the cell state C at time ttThe calculation formula is shown in formulas (2) - (4);
it=σ(Wi[ht-1,PAPWt]+bi) (2)
wherein, WiAs a weight matrix of the input gates, biIs the bias item of the input gate; ct-1Memorizing the cell state of the cell for the last moment in the self-circulation connection; wc and bc are respectively a weight matrix and an offset term of a signal entering an input gate;
2.3) finally, based on the cell state C at that momenttThe output proportion of the output gate is determined, the output of the time-long memory unit at the time t is finally determined, and the calculation formula is as shown in the formula (5);
ht=σ(Wo[ht-1,PAPWt]+bo)*tanh(Ct) (5)
wherein, WOAs a weight matrix of output gates, boFor the biased entry of the output gate, htThe output of the long-time memory unit at the final time t;
2.4) the input of the output layer is the output of all the long-time memory units forming the forward self-circulation connection and the output of all the long-time memory units forming the reverse self-circulation connection, the input information is transmitted to each neuron of the hidden layer in the fully-connected shallow neural network and then transmitted to a single neuron of the output layer in the fully-connected shallow neural network, and the calculation process is shown as formula (6);
CAPWt=linear(WO′*tanh(WH*ht+bH)+bO′) (6)
wherein CAPWt is the final output of the central arterial pressure reconstruction calculation module, and the final output is the central arterial pressure wave obtained by reconstruction; wHAnd bHWeights and biases, W, from input layer to hidden layer in fully-connected shallow neural networks, respectivelyo′And bo′Weights and offsets from a hidden layer to an output layer in the fully-connected shallow neural network are respectively set; linear represents a linear function; h istFor forming the output and reverse of long-and-short-term memory cells connected in a forward self-circulation mannerMerging the outputs of the long and short term memory cells connected by self-circulation.
The correction in the step 1) is completed by a pulse wave correction module, the pulse wave correction module consists of a cuff, an air passage conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, the cuff is connected with the air passage conduit, the air passage conduit is connected with an inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air passage conduit, and the air pressure sensor is installed in the cuff or the air passage conduit and is used for monitoring the air pressure change in the cuff or the air passage conduit; before correction, the cuff is bound at the left brachial artery of a tester, during correction, the inflation and deflation speed of an inflation and deflation motor and the air pressure in the cuff are controlled, then an air pressure sensor is used for measuring the oscillation change curve of the air pressure inside the cuff or in an air passage conduit, then the average pressure and the diastolic pressure of the brachial artery are measured by adopting an oscillometric method in a brachial artery signal processing and calculating module, and the input peripheral arterial pulse wave is corrected by utilizing the average pressure and the diastolic pressure, namely the peripheral arterial pulse wave is subjected to stretching processing of the average pressure and the diastolic pressure, so that the peripheral arterial pulse wave is corrected.
The peripheral arteries include, but are not limited to, the radial artery, the finger artery, and the carotid artery.
The reconstructed central arterial pressure waveform is displayed through a data display module, and the data display module simultaneously displays relevant parameters of the central arterial pressure and peripheral arterial pulse signals and generates a measurement report.
Compared with the prior art, the invention has the following beneficial effects:
the invention can automatically realize the central artery pressure waveform reconstruction from end to end (from the input end to the output end) without manually extracting characteristics, establishing an intermediate simulation model and estimating parameters thereof. In addition, the reconstruction accuracy rate of the invention is obviously superior to that of the traditional method, because the invention adopts the long-time memory unit of the recurrent neural network, the invention can effectively memorize the long-period rule in the peripheral arterial pressure waveform, and adopts the bidirectional structure (namely, the forward and reverse self-circulation connection structure), and the invention can effectively utilize the subsequent associated information in the acquired waveform sequence to realize the reconstruction of the high-precision central arterial pressure waveform. Therefore, compared with the prior art, the method has remarkable advantages in both modeling automation and high-precision reconstruction.
Drawings
Fig. 1 is a block diagram of a system for reconstructing a cardiac pulse pressure waveform according to the present invention.
FIG. 2 is a schematic diagram of a single-layer bilateral long-and-short-term memory network model for cardiac pulse pressure reconstruction according to the present invention.
FIG. 3 is a schematic diagram of a long-short term memory cell in the dual-edge long-short term memory network model according to the present invention.
FIG. 4 is a schematic diagram of a fully-connected artificial neural network in the bilateral long-time memory network model of the present invention.
Fig. 5-schematic diagram of the detailed operation steps of the cardiac pulse pressure waveform reconstruction in the present invention.
FIG. 6 is a graph comparing the reconstructed results of the method with the golden standard for a portion of patients' central artery pressures.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
As shown in fig. 1, the system for reconstructing a waveform of cardiac arterial pressure in the present invention comprises a data acquisition control module, a peripheral arterial pulse wave measurement module, a pulse wave correction module, a central arterial pressure reconstruction calculation module and a data display module;
the data acquisition control module is respectively connected with the peripheral artery pulse wave measurement module and the pulse wave correction module and is used for controlling the measurement and processing sequence of the two modules, including the acquisition of peripheral artery pulse wave signals and the start and stop control of the pulse wave correction module, controlling the data output of the pulse wave correction module to the central artery pressure reconstruction calculation module and outputting the reconstruction result to the display module for final display;
the peripheral artery pulse wave measuring module is used for measuring peripheral artery pulse waves and comprises a pulse signal sensor, a lead and a signal processing circuit, wherein the pulse signal sensor obtains pulse signals of blood flow in an artery, the pulse signal sensor transmits the signals to the signal processing circuit through the lead, the signal processing circuit carries out preprocessing on the pulse signals, the preprocessing comprises signal processing such as amplification and filtering, and the processed results are sent to the pulse wave correcting module; the pulse signal sensor is a piezoelectric sensor or a photoelectric sensor. Peripheral arteries include, but are not limited to, radial arteries, finger arteries, carotid arteries, and the like.
The pulse wave correction module consists of a cuff, an air channel conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, wherein the cuff is connected with the air channel conduit, the air channel conduit is connected with the inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air channel conduit, and the air pressure sensor is arranged in the cuff or the air channel conduit and is used for monitoring the air pressure change in the cuff or the air channel conduit; the method comprises the following steps that an oscillometric method is adopted in a brachial artery signal processing and calculating module to measure the average pressure and diastolic pressure of brachial arteries, the average pressure and diastolic pressure are used for correcting waveforms input by a peripheral artery pulse wave measuring module, namely, stretching processing of equal average pressure and diastolic pressure is carried out on peripheral artery pulse waves, and the pulse wave correcting module outputs corrected pulse wave data to a central artery pressure reconstruction calculating module;
the central arterial pressure reconstruction computing module consists of an input layer, a computing layer and an output layer; the input layer receives the corrected peripheral arterial pulse waves and serially inputs sampling points of the pulse waves into the calculation layer one by one, the calculation layer is composed of a plurality of long-time memory units, each long-time memory unit establishes forward self-circulation connection or reverse self-circulation connection, and the number of the long-time memory units in the forward self-circulation connection is the same as that of the long-time memory units in the reverse self-circulation connection and the long-time memory units in the reverse self-circulation connection are in one-to-one pairing; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; the output layer is composed of fully-connected shallow neural networks, and the shallow neural networks output finally reconstructed central artery pressure pulse waves to the data display module one by one according to the sequence of sampling points;
and the data display module receives the output of the central arterial pressure reconstruction calculation module, and displays the relevant waveform and parameters for a user to check the measurement result.
In order to make reconstruction more accurate, the computing layer is formed by sequentially connecting one or more levels of sub-computing layers, each level of sub-computing layer has the same structure and is formed by a forward layer and a reverse layer, the forward layer comprises one or more forward self-circulation connections, the reverse layer comprises one or more reverse self-circulation connections, and the number of the forward self-circulation connections and the number of the reverse self-circulation connections in the sub-computing layer of the same level are consistent and in one-to-one correspondence; and the output of each level of sub-computation layer comprises the output of the long and short time memory unit in the forward self-circulation connection and the output of the long and short time memory unit in the reverse self-circulation connection of the sub-computation layer.
The method for reconstructing the cardiac pulse pressure waveform comprises the following steps,
1) measuring peripheral arterial pulse waves and correcting;
2) sending the corrected peripheral arterial pulse waves into a central arterial pressure reconstruction calculation module for reconstruction processing, wherein the central arterial pressure reconstruction calculation module consists of an input layer, a calculation layer and an output layer; the input layer receives the corrected peripheral arterial pulse waves and serially inputs sampling points of the pulse waves into the calculation layer one by one, the calculation layer is composed of a plurality of long-time memory units, each long-time memory unit establishes forward self-circulation connection or reverse self-circulation connection, and the number of the long-time memory units in the forward self-circulation connection is the same as that of the long-time memory units in the reverse self-circulation connection and the long-time memory units in the reverse self-circulation connection are in one-to-one pairing; each long-time and short-time memory unit consists of an input gate, a forgetting gate and an output gate; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; output layerThe system is composed of a fully-connected shallow neural network, and the shallow neural network outputs finally reconstructed central arterial pressure pulse waves one by one according to the sequence of sampling points;
the specific reconstruction process is
2.1) forgetting to leave the ratio f of useful information in the memory of this unittAs calculated in the formula (1),
ft=σ(Wf[ht-1,PAPWt]+bf) (1)
wherein, WfWeight matrix for forgetting gate, ht-1For memorizing the output of the cell for the duration of the last moment in the self-circulating connection, PAPWtPeripheral arterial pulse wave input at time t, bfA deviation item for a forgotten door; sigma refers to a Sigmoid activation function, and the function takes values from 0 to 1;
2.2) screening and retaining newly input peripheral arterial pulse wave information itIndicating, acquiring the state of an input cellGet the cell state C at time tt(Is an intermediate quantity, in order to calculate CtAmount of (d), the calculation formula is as in formulas (2) - (4);
it=σ(Wi[ht-1,PAPWt]+bi) (2)
wherein, WiAs a weight matrix of the input gates, biIs the bias item of the input gate; ct-1Memorizing the cell state of the cell for the last moment in the self-circulation connection; wc and bc are signal entries, respectivelyThe weight and offset of the input gate;
2.3) finally, based on the cell state C at that momenttThe output proportion of the output gate is determined, the output of the time-long memory unit at the time t is finally determined, and the calculation formula is as shown in the formula (5);
ht=σ(Wo[ht-1,PAPWt]+bo)*tanh(Ct) (5)
wherein, WOAs a weight matrix of output gates, boFor the biased entry of the output gate, htThe output of the long-time memory unit at the final time t;
2.4) the input of the output layer is the output of the long and short time memory unit in the forward self-circulation connection and the output of the long and short time memory unit in the reverse self-circulation connection, the input information is transmitted to each neuron of the hidden layer in the fully-connected shallow neural network and then transmitted to a single neuron of the output layer in the fully-connected shallow neural network, and the calculation process is shown as a formula (6);
CAPWt=linear(WO′*tanh(WH*ht+bH)+bO′) (6)
wherein, CAPWtThe central arterial pressure is finally output by a central arterial pressure reconstruction calculation module, namely the central arterial pressure wave obtained by reconstruction; wHAnd bHWeights and biases, W, from input layer to hidden layer in fully-connected shallow neural networks, respectivelyo′And bo′Weights and offsets from a hidden layer to an output layer in the fully-connected shallow neural network are respectively set; linear represents a linear function; h istIs the combination of the output of the long and short duration memory cells in the forward self-loop connection and the output of the long and short duration memory cells in the reverse self-loop connection.
In the invention, the core of the reconstruction of the central arterial pressure waveform is completed by adopting a single-layer or multi-layer bilateral long-time and short-time memory artificial neural network. Fig. 2 is a schematic diagram of a multilayer bilateral long-term and short-term memory artificial neural network model, which is structurally composed of an input layer, a computing layer (multilayer) and an output layer, wherein the input layer receives peripheral arterial pulse waves and respectively sends the peripheral arterial pulse waves to the computing layer. The computation layer is composed of several long-short-time memory units (LSTM) and eachThe long and short time memory unit establishes a forward self-circulation connection or a reverse self-circulation connection, and the calculation layer outputs the forward self-circulation connection or the reverse self-circulation connection to the output layer. The output layer is formed by a fully-connected artificial neural network (FNN), and comprises an input layer, a hidden layer and an output layer. FIG. 3 is a schematic diagram of a single LSTM cell having three input signals (peripheral arterial signal at the current time, output of the LSTM cell at a time immediately preceding the self-circulation connection, cell state of the LSTM cell at a time immediately preceding the self-circulation connection) and two outputs (cell state C of the LSTM cell itself)tAnd htOutput); self cell state CtAs input to the LSTM cell at the next instant in the self-circulating connection, where htThe output is divided into two paths, one path is used as the input of the LSTM unit at the next moment in the self-circulation connection (the horizontal path in the figure), and the other path (the vertical upward path) is used as the input of the output layer or the input of the LSTM unit at each moment in the next layer of calculation layer. The previous time is the same as the previous time in the normal sense for the forward self-loop connection (in the chronological order of the events, the previous time is the previous time relative to the later time), and the previous time is the opposite of the previous time in the normal sense for the reverse self-loop connection, which is the characteristic of the forward self-loop connection and the reverse self-loop connection.
FIG. 4 is a schematic diagram of a fully-connected artificial neural network in the bilateral long-time and short-time memory network model according to the present invention.
The correction in the step 1) is completed by a pulse wave correction module, the pulse wave correction module consists of a cuff, an air channel conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, the cuff is connected with the air channel conduit, the air channel conduit is connected with an inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air channel conduit, and the air pressure sensor is installed in the cuff or the air channel conduit and used for monitoring the air pressure change in the cuff or the air channel conduit; before correction, the cuff is bound at the left brachial artery of a tester, during correction, the inflation and deflation speed of an inflation and deflation motor and the air pressure in the cuff are controlled, then an air pressure sensor is used for measuring the oscillation change curve of the air pressure inside the cuff or in an air passage conduit, then the average pressure and the diastolic pressure of the brachial artery are measured by adopting an oscillometric method in a brachial artery signal processing and calculating module, and the input peripheral arterial pulse wave is corrected by utilizing the average pressure and the diastolic pressure, namely the peripheral arterial pulse wave is subjected to stretching processing of the average pressure and the diastolic pressure, so that the peripheral arterial pulse wave is corrected.
In order to visually display, the reconstructed central arterial pressure waveform is displayed by the data display module, and the data display module simultaneously displays the relevant parameters of the central arterial pressure and the peripheral arterial pulse signals and generates a measurement report.
In actual measurement, the following sequence of operations is followed, see fig. 5:
the first step is as follows: after a tester lies and has a rest for 5-10 minutes, the cuff is bound at the brachial artery of the left arm, then the piezoelectric sensor or the photoelectric sensor is placed at the radial artery, the finger artery or other peripheral arteries, and the position of the sensor is adjusted to find the optimal position and fix the optimal position;
the second step is that: the data acquisition control module starts and controls the peripheral artery pulse wave measurement module to carry out measurement, so that the peripheral artery pulse wave signals are acquired, and the peripheral artery pulse wave measurement module carries out processing such as amplification and filtering of the signals;
the third step: the pulse wave correction module is started to work through the data acquisition control module, the inflation and deflation speed of the inflation and deflation motor and the air pressure in the cuff are controlled, then the pulse wave correction module measures the oscillation change curve of the air pressure in the cuff by using the air pressure sensor, and then the measurement of the systolic pressure, the average pressure and the diastolic pressure of the brachial artery blood pressure is completed by adopting an oscillometric method; and finally, correcting the peripheral pulse wave by adopting the correction calculation of the pulse wave correcting module, namely stretching operation of equal average pressure and diastolic pressure.
The fourth step: starting a central arterial pressure reconstruction calculation module, inputting the corrected peripheral pulse waves into an input layer of the central arterial pressure reconstruction calculation module through control, realizing reconstruction calculation of blood pressure waveforms from peripheral arteries to central aorta by utilizing a bidirectional long-time and short-time memory network, and sending the reconstructed central arterial pressure waveforms to a display module;
the fifth step: the data display module reads the sent waveform data from the cache through the module, displays each blood pressure waveform and related parameters on a display screen, and simultaneously generates a measurement report.
Example (b):
the method of the invention is used for testing the clinical experiments of 62 patients, and the effect is obviously improved compared with the traditional method. In clinical experiments, 62 patients respectively measure the blood pressure waveforms of a radial artery and a central artery through invasive intervention of a catheter, the invasive radial artery pulse wave is used as the input of the invention, the central artery pressure waveform is reconstructed through the invention, then the central artery pressure measured with the wound is compared, and the reconstruction effect of the central artery pressure of part of the patients is shown in fig. 6. In FIG. 6, RAPWinvRadial pulse wave, CAPW, for invasive measurementinvCentral arterial pressure waveform for invasive measurement, CAPWBiLSTMFor the central arterial pressure waveform reconstructed by the method, the overall effect of the invention is good as can be seen from the comparison of fig. 6.
In order to compare the difference between the present method and the conventional method, two general transfer function methods (GTF) are used simultaneouslyFFTAnd GTFARX) The results of the central arterial pressure reconstruction waveform and the related indexes of the one-way long-short time memory network (LSTM) and the method (BilSTM) are compared with the standard measurement results of the invasive gold, and the results are shown in Table 1. The linear regression and the Pearson correlation coefficient are the comparison of the systolic pressure, the waveform root mean square error is the error after the whole wave reconstruction, and the mean deviation of the systolic pressure represents the mean value and the variance of the systolic pressure value output by the model and the actually measured systolic pressure value. As can be seen from Table 1, BilSTM and GTFARXThe method is best in error of central artery pressure waveform reconstruction and Pearson correlation coefficient, and the mean-difference of the BilSTM in the systolic pressure is the smallest.
TABLE 1 statistical comparison of Central arterial pressure reconstructed waveforms and indices for the general method and the method
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all the modifications or equivalent substitutions should be covered by the claims of the present invention.
Claims (7)
1. A central arterial pressure waveform reconstruction system characterized by: the device comprises a data acquisition control module, a peripheral artery pulse wave measurement module, a pulse wave correction module, a central artery pressure reconstruction calculation module and a data display module;
the data acquisition control module is respectively connected with the peripheral artery pulse wave measurement module and the pulse wave correction module and is used for controlling the measurement and processing sequence of the two modules, including the acquisition of peripheral artery pulse wave signals and the start and stop control of the pulse wave correction module, controlling the data output of the pulse wave correction module to the central artery pressure reconstruction calculation module and outputting the reconstruction result to the display module for final display;
the peripheral artery pulse wave measuring module is used for measuring peripheral artery pulse waves and comprises a pulse signal sensor, a lead and a signal processing circuit, wherein the pulse signal sensor obtains pulse signals of blood flow in an artery, the pulse signal sensor transmits the signals to the signal processing circuit through the lead, the signal processing circuit carries out preprocessing on the pulse signals, the preprocessing comprises signal processing such as amplification and filtering, and the processed results are sent to the pulse wave correcting module;
the pulse wave correction module consists of a cuff, an air channel conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, wherein the cuff is connected with the air channel conduit, the air channel conduit is connected with the inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air channel conduit, and the air pressure sensor is arranged in the cuff or the air channel conduit and is used for monitoring the air pressure change in the cuff or the air channel conduit; the method comprises the following steps that an oscillometric method is adopted in a brachial artery signal processing and calculating module to measure the average pressure and diastolic pressure of brachial arteries, the average pressure and diastolic pressure are used for correcting waveforms input by a peripheral artery pulse wave measuring module, namely, stretching processing of equal average pressure and diastolic pressure is carried out on peripheral artery pulse waves, and the pulse wave correcting module outputs corrected pulse wave data to a central artery pressure reconstruction calculating module;
the central arterial pressure reconstruction computing module consists of an input layer, a computing layer and an output layer; the input layer receives the corrected peripheral arterial pulse waves and serially inputs sampling points of the pulse waves into the calculation layer one by one, the calculation layer is composed of a plurality of long-time memory units, each long-time memory unit establishes forward self-circulation connection or reverse self-circulation connection, and the number of the long-time memory units in the forward self-circulation connection is the same as that of the long-time memory units in the reverse self-circulation connection and the long-time memory units in the reverse self-circulation connection are in one-to-one pairing; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; the output layer is composed of fully-connected shallow neural networks, and the shallow neural networks output finally reconstructed central artery pressure pulse waves to the data display module one by one according to the sequence of sampling points;
and the data display module receives the output of the central arterial pressure reconstruction calculation module and displays the relevant waveforms and parameters.
2. The central arterial pressure waveform reconstruction system of claim 1, wherein: the pulse signal sensor is a piezoelectric sensor or a photoelectric sensor.
3. The central arterial pressure waveform reconstruction system of claim 1, wherein: the computing layer is formed by sequentially connecting one or more levels of sub computing layers, each level of sub computing layer is the same in structure and is formed by a forward layer and a reverse layer, the forward layer comprises one or more forward self-circulation connections, the reverse layer comprises one or more reverse self-circulation connections, the number of the forward self-circulation connections in the same level of sub computing layer is consistent with the number of the reverse self-circulation connections in the same level of sub computing layer and corresponds to the number of the reverse self-circulation connections one by one, the last level of sub computing layer outputs to the output layer, and the output of each level of sub computing layer comprises the output of all long and short time memory units forming the forward self-circulation connections in the level of sub computing layer and the output of all long and short time memory units forming the reverse self-circulation connections.
4. A central artery pressure waveform reconstruction method is characterized by comprising the following steps: the method comprises the following steps of,
1) measuring peripheral arterial pulse waves and correcting;
2) sending the corrected peripheral arterial pulse waves into a central arterial pressure reconstruction calculation module for reconstruction processing, wherein the central arterial pressure reconstruction calculation module consists of an input layer, a calculation layer and an output layer; the input layer receives the corrected peripheral arterial pulse waves and serially inputs sampling points of the pulse waves into the calculation layer one by one, the calculation layer is composed of a plurality of long-time memory units, each long-time memory unit establishes forward self-circulation connection or reverse self-circulation connection, and the number of the long-time memory units in the forward self-circulation connection is the same as that of the long-time memory units in the reverse self-circulation connection and the long-time memory units in the reverse self-circulation connection are in one-to-one pairing; each long and short term memory unit receives the signal input of the current time input layer and the output of the long and short term memory unit at the previous time; the output of the last moment of this long and short term memory cell includes cell state CtAnd output h at the last momentt(ii) a Each long-short time memory unit is output to the output layer and is simultaneously used as the input of the long-short time memory unit at the next time in the self-circulation connection; the output layer is composed of fully-connected shallow neural networks, and the shallow neural networks output finally reconstructed central artery pressure pulse waves one by one according to the sequence of sampling points;
the specific reconstruction process is
2.1) forgetting to leave the ratio f of useful information in the memory of this unittAs calculated in the formula (1),
ft=σ(Wf[ht-1,PAPWt]+bf) (1)
wherein, WfWeight matrix for forgetting gate, ht-1For memorizing the output of the cell for the duration of the last moment in the self-circulating connection, PAPWtPeripheral arterial pulse wave input at time t, bfA deviation item for a forgotten door; sigma refers to a Sigmoid activation function, and takes a value from 0 to 1;
2.2) screening and retaining newly input peripheral arterial pulse wave information itIndicating, acquiring the state of an input cellGet the cell state C at time ttThe calculation formula is shown in formulas (2) - (4);
it=σ(Wi[ht-1,PAPWt]+bi) (2)
wherein, WiAs a weight matrix of the input gates, biIs the bias item of the input gate; ct-1Memorizing the cell state of the cell for the last moment in the self-circulation connection; wc and bc are respectively a weight matrix and an offset term of a signal entering an input gate;
2.3) finally, based on the cell state C at that momenttThe output proportion of the output gate is determined, the output of the time-long memory unit at the time t is finally determined, and the calculation formula is as shown in the formula (5);
ht=σ(Wo[ht-1,PAPWt]+bo)*tanh(Ct) (5)
wherein, W0As a weight matrix of output gates, boFor the biased entry of the output gate, htFor final t-time long-and-short memory listThe output of the element;
2.4) the input of the output layer is the output of all the long-time memory units forming the forward self-circulation connection and the output of all the long-time memory units forming the reverse self-circulation connection, the input information is transmitted to each neuron of the hidden layer in the fully-connected shallow neural network and then transmitted to a single neuron of the output layer in the fully-connected shallow neural network, and the calculation process is shown as formula (6);
CAPWt=linear(WO′*tanh(WH*ht+bH)+bO′) (6)
wherein, CAPWtThe central arterial pressure is the final output of the central arterial pressure reconstruction calculation module, and the final output is the central arterial pressure wave obtained by reconstruction; wHAnd bHWeights and biases, W, from input layer to hidden layer in fully-connected shallow neural networks, respectivelyO′And bO′Weights and offsets from a hidden layer to an output layer in the fully-connected shallow neural network are respectively set; linear represents a linear function; h istIs a combination of the outputs of the long and short term memory cells constituting the forward self-looping connection and the outputs of the long and short term memory cells constituting the reverse self-looping connection.
5. The method of claim 4, wherein the central artery pressure waveform reconstruction method comprises: the correction in the step 1) is completed by a pulse wave correction module, the pulse wave correction module consists of a cuff, an air passage conduit, an air pressure sensor, an inflation and deflation motor and a brachial artery signal processing and calculating module, the cuff is connected with the air passage conduit, the air passage conduit is connected with an inflation and deflation motor, the inflation and deflation motor performs inflation and deflation on the cuff through the air passage conduit, and the air pressure sensor is arranged in the cuff or the air passage conduit and is used for monitoring the air pressure change in the cuff or the air passage conduit; before correction, the cuff is bound at the left brachial artery of a tester, during correction, the inflation and deflation speed of an inflation and deflation motor and the air pressure in the cuff are controlled, then an air pressure sensor is used for measuring the oscillation change curve of the air pressure inside the cuff or in an air passage conduit, then the average pressure and the diastolic pressure of the brachial artery are measured by adopting an oscillometric method in a brachial artery signal processing and calculating module, and the input peripheral arterial pulse wave is corrected by utilizing the average pressure and the diastolic pressure, namely the peripheral arterial pulse wave is subjected to stretching processing of the average pressure and the diastolic pressure, so that the peripheral arterial pulse wave is corrected.
6. The method of claim 4, wherein the central artery pressure waveform reconstruction method comprises: the peripheral arteries include, but are not limited to, the radial artery, the finger artery, and the carotid artery.
7. The method of claim 4, wherein the central artery pressure waveform reconstruction method comprises: the reconstructed central arterial pressure waveform is displayed through a data display module, and the data display module simultaneously displays relevant parameters of the central arterial pressure and peripheral arterial pulse signals and generates a measurement report.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011613837.5A CN112790748A (en) | 2020-12-30 | 2020-12-30 | Central arterial pressure waveform reconstruction system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011613837.5A CN112790748A (en) | 2020-12-30 | 2020-12-30 | Central arterial pressure waveform reconstruction system and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112790748A true CN112790748A (en) | 2021-05-14 |
Family
ID=75804662
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011613837.5A Pending CN112790748A (en) | 2020-12-30 | 2020-12-30 | Central arterial pressure waveform reconstruction system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112790748A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113413145A (en) * | 2021-07-30 | 2021-09-21 | 北京阅影科技有限公司 | Method, device and computer readable storage medium for acquiring central arterial pressure |
CN113499048A (en) * | 2021-07-22 | 2021-10-15 | 重庆理工大学 | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103505191A (en) * | 2012-06-20 | 2014-01-15 | 法玛科技顾问股份有限公司 | Method and device for estimating pulse pressure of central aorta by using pressure pulse wave vibration signals of tourniquet |
CN103876723A (en) * | 2014-04-01 | 2014-06-25 | 江苏理工学院 | Method of obtaining blood pressure value by noninvasive radial artery wave calculating pulse wave transmission time |
CN104000573A (en) * | 2014-05-29 | 2014-08-27 | 东北大学 | Body surface two point pulse wave based central arterial pulse monitoring system and method |
CN104188642A (en) * | 2014-09-24 | 2014-12-10 | 中国科学院合肥物质科学研究院 | Device for noninvasive acquiring centre artery pressure waveform and relevant parameters and acquiring method thereof |
CN109512412A (en) * | 2018-11-21 | 2019-03-26 | 重庆理工大学 | A kind of central aortic blood pressure measuring device |
CN111493855A (en) * | 2020-04-21 | 2020-08-07 | 重庆理工大学 | System and method for non-invasive measurement of individualized cardiac output |
-
2020
- 2020-12-30 CN CN202011613837.5A patent/CN112790748A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103505191A (en) * | 2012-06-20 | 2014-01-15 | 法玛科技顾问股份有限公司 | Method and device for estimating pulse pressure of central aorta by using pressure pulse wave vibration signals of tourniquet |
CN103876723A (en) * | 2014-04-01 | 2014-06-25 | 江苏理工学院 | Method of obtaining blood pressure value by noninvasive radial artery wave calculating pulse wave transmission time |
CN104000573A (en) * | 2014-05-29 | 2014-08-27 | 东北大学 | Body surface two point pulse wave based central arterial pulse monitoring system and method |
CN104188642A (en) * | 2014-09-24 | 2014-12-10 | 中国科学院合肥物质科学研究院 | Device for noninvasive acquiring centre artery pressure waveform and relevant parameters and acquiring method thereof |
CN109512412A (en) * | 2018-11-21 | 2019-03-26 | 重庆理工大学 | A kind of central aortic blood pressure measuring device |
CN111493855A (en) * | 2020-04-21 | 2020-08-07 | 重庆理工大学 | System and method for non-invasive measurement of individualized cardiac output |
Non-Patent Citations (2)
Title |
---|
CHAKSHU, N.K., SAZONOV, I. & NITHIARASU, P.: "Towards enabling a cardiovascular digital twin for human systemic circulation using inverse analysis", 《BIOMECH MODEL MECHANOBIOL》 * |
高扬: "《人工智能与机器人先进技术丛书 智能摘要与深度学习》", 31 July 2019 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113499048A (en) * | 2021-07-22 | 2021-10-15 | 重庆理工大学 | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM |
CN113499048B (en) * | 2021-07-22 | 2022-07-08 | 重庆理工大学 | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM |
CN113413145A (en) * | 2021-07-30 | 2021-09-21 | 北京阅影科技有限公司 | Method, device and computer readable storage medium for acquiring central arterial pressure |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108498089B (en) | Noninvasive continuous blood pressure measuring method based on deep neural network | |
US11844593B2 (en) | System and method for non-invasive determination of blood pressure dip based on trained prediction models | |
US11298029B2 (en) | Blood pressure measuring apparatus, blood pressure measuring method, electronic device, and computer readable storage medium | |
CN106889979B (en) | Continuous blood pressure measuring system based on electrocardiosignals and blood oxygen volume waves | |
CN106821356A (en) | High in the clouds continuous BP measurement method and system based on Elman neutral nets | |
CN112790748A (en) | Central arterial pressure waveform reconstruction system and method | |
CN113143230B (en) | Peripheral arterial blood pressure waveform reconstruction system | |
CN103876723A (en) | Method of obtaining blood pressure value by noninvasive radial artery wave calculating pulse wave transmission time | |
CN104000573A (en) | Body surface two point pulse wave based central arterial pulse monitoring system and method | |
CN107854123A (en) | One kind is without oversleeve continuous blood pressure monitoring method and apparatus | |
Paviglianiti et al. | Noninvasive arterial blood pressure estimation using ABPNet and VITAL-ECG | |
CN114145724A (en) | Method for dynamically monitoring blood pressure based on ECG (electrocardiogram) and PPG (photoplethysmography) multiple physiological characteristic parameters | |
CN101224108A (en) | Noninvasive intracranial pressure monitoring equipment | |
CN109512410A (en) | A kind of more physiological signal Fusion Features without cuff continuous BP measurement method | |
Alastruey et al. | Arterial pulse wave modeling and analysis for vascular-age studies: a review from VascAgeNet | |
CN113662520B (en) | Wearable continuous blood pressure measuring system based on uncertainty quantification strategy | |
CN113040738B (en) | Blood pressure detecting device | |
CN201019769Y (en) | Non-invasive intracranial pressure monitoring equipment | |
US20150157217A1 (en) | Analysis System for Cardiac Information and Analyzing Method Thereof | |
CN116392091A (en) | Continuous arterial blood pressure waveform detection system and device based on single-channel finger tip PPG | |
CN113499048B (en) | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM | |
CN115836847A (en) | Blood pressure prediction device and equipment | |
US20220133159A1 (en) | Non-Invasive, Continuous, Accurate and Cuff-Less Measurement of Blood Pressure and Other Cardiovascular Variables by Pulse Wave Acquisition and Analysis Using Non-Invasive Sensors | |
CN204813858U (en) | Blood pressure appearance based on pulse back wave transmission time | |
CN114587310A (en) | Method for realizing invasive blood pressure waveform estimation based on photoplethysmography |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210514 |