CN116761544A - Method and device for intelligent respiration monitoring by electrocardiogram, respiration acoustics and chest acceleration - Google Patents
Method and device for intelligent respiration monitoring by electrocardiogram, respiration acoustics and chest acceleration Download PDFInfo
- Publication number
- CN116761544A CN116761544A CN202180089672.2A CN202180089672A CN116761544A CN 116761544 A CN116761544 A CN 116761544A CN 202180089672 A CN202180089672 A CN 202180089672A CN 116761544 A CN116761544 A CN 116761544A
- Authority
- CN
- China
- Prior art keywords
- respiratory
- electrocardiogram
- acoustics
- acceleration
- chest
- 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
- 238000000034 method Methods 0.000 title claims abstract description 35
- 230000001133 acceleration Effects 0.000 title claims abstract description 30
- 230000029058 respiratory gaseous exchange Effects 0.000 title claims abstract description 25
- 238000012544 monitoring process Methods 0.000 title abstract description 15
- 230000000241 respiratory effect Effects 0.000 claims abstract description 49
- 230000036387 respiratory rate Effects 0.000 claims abstract description 27
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 230000035565 breathing frequency Effects 0.000 claims description 6
- 230000036391 respiratory frequency Effects 0.000 claims description 4
- 238000012886 linear function Methods 0.000 claims description 3
- 238000012887 quadratic function Methods 0.000 claims description 3
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 claims 2
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 claims 2
- 230000009466 transformation Effects 0.000 claims 2
- 208000037656 Respiratory Sounds Diseases 0.000 abstract description 8
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 abstract description 6
- 230000006866 deterioration Effects 0.000 abstract description 6
- 230000004064 dysfunction Effects 0.000 abstract description 5
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 abstract description 4
- 238000005259 measurement Methods 0.000 abstract description 4
- 229910052760 oxygen Inorganic materials 0.000 abstract description 4
- 239000001301 oxygen Substances 0.000 abstract description 4
- 229910002092 carbon dioxide Inorganic materials 0.000 abstract description 3
- 239000001569 carbon dioxide Substances 0.000 abstract description 3
- 238000009423 ventilation Methods 0.000 abstract description 3
- 208000004756 Respiratory Insufficiency Diseases 0.000 abstract description 2
- 206010038687 Respiratory distress Diseases 0.000 abstract description 2
- 238000006213 oxygenation reaction Methods 0.000 abstract description 2
- 201000004193 respiratory failure Diseases 0.000 abstract description 2
- 238000012549 training Methods 0.000 description 8
- 230000004202 respiratory function Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 239000000203 mixture Substances 0.000 description 3
- 230000002441 reversible effect Effects 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000013186 photoplethysmography Methods 0.000 description 2
- 238000009531 respiratory rate measurement Methods 0.000 description 2
- 208000007848 Alcoholism Diseases 0.000 description 1
- 241000004176 Alphacoronavirus Species 0.000 description 1
- 206010003658 Atrial Fibrillation Diseases 0.000 description 1
- 208000025721 COVID-19 Diseases 0.000 description 1
- 208000006545 Chronic Obstructive Pulmonary Disease Diseases 0.000 description 1
- 206010013654 Drug abuse Diseases 0.000 description 1
- 241000124008 Mammalia Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 206010001584 alcohol abuse Diseases 0.000 description 1
- 208000025746 alcohol use disease Diseases 0.000 description 1
- 230000036626 alertness Effects 0.000 description 1
- 208000008784 apnea Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 208000006673 asthma Diseases 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000002876 beta blocker Substances 0.000 description 1
- 229940097320 beta blocking agent Drugs 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000747 cardiac effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000029142 excretion Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000007620 mathematical function Methods 0.000 description 1
- 230000000474 nursing effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 231100000915 pathological change Toxicity 0.000 description 1
- 230000036285 pathological change Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000002106 pulse oximetry Methods 0.000 description 1
- 230000036412 respiratory physiology Effects 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 208000011117 substance-related disease Diseases 0.000 description 1
- 230000002889 sympathetic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
-
- 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
- A61B5/02405—Determining heart rate variability
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/113—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb occurring during breathing
- A61B5/1135—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb occurring during breathing by monitoring thoracic expansion
-
- 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
- A61B5/72—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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B7/00—Instruments for auscultation
- A61B7/003—Detecting lung or respiration noise
Abstract
Respiration is the basis of life. Continuous monitoring of respiratory rate and its pattern is critical to detecting the onset of respiratory failure; however, most hospital wards are unable to monitor this vital sign. Respiratory distress is caused by either an oxygenation dysfunction (insufficient oxygen inhalation) or a ventilation dysfunction (insufficient carbon dioxide removal). Methods and apparatus for monitoring respiratory rate by extracting information from chest acceleration, electrocardiogram and respiratory sounds are disclosed. The application also discloses a method of quantifying the degree of respiratory deterioration by defining a smart respiratory index defined by combining at least three parameters extracted from physiological measurements such as electrocardiogram, respiratory sounds and chest acceleration. These methods are implemented by a small wireless patch attached to the upper part of the patient's chest and communicate with external proprietary software and monitors via bluetooth.
Description
Technical Field
The present application relates generally to apparatus and methods for determining the respiratory quality and effort of a subject and respiratory frequency (RR). More particularly, the present application relates to a wireless device and method for obtaining an index representing the probability of respiratory deterioration.
Background
Respiration is the basis of life. The lungs are responsible for breathing, the process of supplying oxygen to the body and removing carbon dioxide from the body.
Respiratory distress is caused by either an oxygenation dysfunction (insufficient oxygen inhalation) or a ventilation dysfunction (insufficient carbon dioxide excretion).
These forms of malfunction may manifest as irregular breathing frequencies and/or abnormal breathing patterns. Continuous monitoring of RR and patterns is critical to detecting the onset of respiratory failure; however, most hospital wards fail to properly monitor such vital signs. When measuring RR, this measurement is performed very slowly. Patients in a general ward are typically checked every six to eight hours and often have significant human error because nurses typically observe the patient's breath for 15 seconds and then multiply the result by 4 to obtain an RR per minute. The world health organization advocates counting the number of breaths in one minute. If RR is above or below the acceptable range of 12-16 breaths/min, then a respiratory problem is indicated.
Current devices for measuring RR are very expensive and are connected to the patient by cables, so they are only used in special units, such as Intensive Care Units (ICU).
RR dies in hospitals 21% of patients between 25 and 29 per minute (Respiratory rate: the neglected vital sign. Michelle A Critikos, rinaldo Belllomo, ken Hillman, jack Chen, simon Finfer and Arthas Flabous. MJA 2008; 188:657-659). The best way to maintain alertness and reduce complications to a patient's clinical state is to continuously monitor the patient's RR and Tidal Volume (TV). Coronavirus epidemic (Covid-19) has clearly shown that respiration monitoring is necessary in a nursing home outside the hospital and even in a home care setting.
U.S. patent application "Acoustic sensor and ventilation monitoring system" to Joseph et al (US 2020/0054277A 1) discloses a method of monitoring respiration using an acoustic measurement device. After a broad description of respiratory physiology, authors propose a device integrated by two elements. One of the elements is attached to the patient and the second element is attached to the first element. The first element is composed of a sound sensor, an accelerometer and a transmitter. The second element is comprised of a rechargeable battery. In the device, a plurality of sensors may be integrated to measure temperature, heart rate and oxygen saturation. The device may also be connected to a smart watch. With information from different sensors, they propose a risk factor index to determine changes in respiratory function of mammals due to physiological pathological changes and drug or alcohol abuse. They describe a quadratic equation to calculate their respiratory risk factor index. The main difference from our current patent application is that Joseph et al uses only accelerometers to estimate body movement on the one hand, and accelerometers to estimate respiratory rate on the other hand, the formulas of respiratory risk index and smart respiratory index differ significantly. Furthermore, in Joseph et al, respiratory rate is estimated by respiratory acoustics, while in this patent application respiratory rate is estimated by the chest acceleration measured by an accelerometer and the first derivative of the respiratory acoustic envelope (breath sound envelope).
Both Joseph and this patent application use sensors as microphones and accelerometers to evaluate the condition of breathing sounds and movements. However, the parameters analyzed by Joseph et al in both of these works and the derivation formulas for calculating risk factor indices are quite different from the SRI of the present disclosure. Joseph et al use an accelerometer to evaluate the patient's body movement, and in this disclosure, the accelerometer is used to calculate the respiratory rate.
Joseph et al use breath sounds to calculate the respiratory rate and TV, while in this disclosure use the envelope to evaluate the variability of TV.
Finally, for Joseph et al, body movement and speech add up to score highest, while slow breathing and lack of movement score lowest. U.S. application "Portable device with multiple integrated sensors for vital signs scanning" (US 2015/0313484 A1) discloses a portable device with multiple integrated sensors. The patient application of the application is different, it uses neither a thermometer nor photoplethysmography (PPG), and the application defines an index of respiratory quality.
The US application "Mobile frontend system for comprehensive cardiac diagnosis" (US 2015/0065814 A1) is significantly different from the present patent application, and the purpose of US2015/0065814 is to comprehensively diagnose heart problems.
U.S. patent application "Mesh network personal emergency appliance" (US 2008/0001735 A1) discloses a system comprising one or more wireless nodes forming a wireless mesh network, the difference of the present application being that it does not form a wireless mesh network.
U.S. patent application "Monitoring, predicting and treating clinical episodes" (US 2008/0275349 A1) discloses a device for sensing a physiological parameter of a subject and being able to sense a wide range of body movements, in contrast to the present patent application which does not disclose sensing a wide range of body movements.
US application "Physiological acoustic monitoring system" (US 8821415B 2) discloses a method of using acoustic signals to assess respiratory frequency, however this application uses only one or two acoustic sensors (microphones) and thus is quite different from the present application which also uses an Electrocardiogram (ECG) and an accelerometer.
U.S. patent "METHODS AND SYSTEMS FOR MONITORING RESPIRATION" (U.S. Pat. No. 6,918,878 B2) discloses a method for determining the respiratory rate of a patient, the method comprising a plurality of parts. The respiratory rate can be determined by measuring the S2 division of the heart. S2 split can be identified by observing the time of heart sounds. Other respiration-related information, such as respiratory phase and the occurrence of apneas, may also be identified. This type of respiratory monitor may be useful for monitoring subacute patients and for monitoring outpatients. The sensor for the respiration monitor and the electrode for the ECG monitor may be combined into a single probe.
The present patent application does not include S2 splitting and is therefore different from patent application US 6,918,878.
U.S. patent "System and method for monitoring respiratory rate measurements" (US 20180214090 A1) discloses a system and method for determining multi-parameter confidence in respiratory rate measurements using multiple physiological parameter inputs. This disclosure differs from the present patent application in that the present application combines breath acoustics with ECG R-R interval variation (Reaks).
Disclosure of Invention
In a preferred embodiment, a patch (1, 13, 27, 42) is attached to the chest of a patient, said patch (1, 13, 27, 42) containing at least a sensor for an Electrocardiogram (ECG), a sensor for measuring the Respiratory Rate (RR) by an accelerometer, a sensor for measuring RR by a microphone. The ECG (2, 14, 29) is further processed using a Heart Rate (HR) extraction algorithm, such as, but not limited to, HR using a Fast Fourier Transform (FFT) (6, 19, 34) to extract the ECG, and Rpeaks (7, 20, 35) is also determined by the FFT (spectral analysis).
Respiratory sounds (respiratory acoustics) from microphones (3, 15, 30) are processed through envelope waveform respiration extraction formulas (8, 21, 36) to obtain a respiratory frequency, referred to as RespR. Equation (9, 22, 37) is used to calculate the tidal volume variability (TVv). The acceleration signal (4, 16, 31) of the chest is analyzed by evaluating the Hilbert transform model of respiration, called RespRacc (10, 23, 38). The relationship between the three parameters Rpeaks, respR and RespRacc may change as respiration worsens, so cross-mutual information (cross mutual information) (5, 18, 32) is calculated and the variable CMIbreath is generated. The parameters extracted from the measurements are fed into a classifier, which may be but is not limited to an adaptive fuzzy neural network system (ANFIS) (11, 25, 40). The output of the classifier is called the intelligent respiratory index (SRI) (12, 26, 41).
Estimation of RespR by envelope formula
The acoustic signal recorded by the microphone is input to a spline function or other curve function to evaluate the amplitude envelope, as shown in fig. 7. By calculating the peaks in the envelope curve, respR can be calculated, see also the example in fig. 6.
Estimation of tidal volume variability (TVv)
Previously, the relationship between the respiratory airflow F and the energy E of the respiratory (tracheal) sound may be represented by the form a=kf α Where k and α are constants, different study teams make different suggestions for the value of the index. The amplitude-airflow relationship of such sounds has been used for respiratory monitoring, in particular for qualitative and quantitative assessment of respiratory airflow and for continuous respiratory rate estimation.
However, we have found that the estimation can be improved by adding the derivative of the breathing envelope to the equation, thus defining the following flow equation:
F＝k 1 max (dA envelope/dt) +k 2 A β
The volume can thus be estimated as the integral of the flow F over the inspiration time,
tidal volume variability is defined herein as varying over time, for example, if tidal volume increases from 10 to 12, then tidal volume variability becomes 20%.
Determination of RespRacc by applying Hilbert transform to acceleration signals
A new method of extracting respiration from acceleration signals using Hilbert Vibration Decomposition (HVD) is presented. The maximum energy component of the acquired acceleration signal is proportional to the respiration signal.
Determination of Smart breath index (SRI)
In one embodiment, SRI is a linear or quadratic function of RespR, HR, RRv and TVv. The formula may be:
SRI＝K1*RespR+k2*HR+k3*RRV+k4*TVv+K5*RespR*HR*RRV*TVv，
more secondary factors may be included in the quadratic equation.
The constants k1 to k4 should be within the following range:
0.30<k1<0.7，
0.2<k2<0.4，
0.05<k3<0.2，
0.01<k4<0.1。
combining parameters by classifier to define SRI
In a second embodiment, as shown in fig. 1, the apparatus combines parameters using a classifier, such as, but not limited to, using an ANFIS model, to define an SRI definition. Parameters extracted from at least 3 sensors (ECG, breath sounds, chest acceleration) are used as input to an adaptive fuzzy neural network system (ANFIS). In a third embodiment, as shown in FIG. 2, the apparatus combines parameters using an ANFIS model to define SRI. Parameters extracted from at least 4 sensors (ECG, respiratory sounds, chest acceleration, pulse oximeter) are used as inputs to an adaptive fuzzy neural network system (ANFIS).
In a fourth embodiment, as shown in FIG. 3, the apparatus combines parameters using an ANFIS model to define SRI. Parameters extracted from at least 4 sensors (ECG, respiratory sounds, chest acceleration, pulse oximeter) and demographic data (age, sex, height, weight) and clinical data (chronic obstructive pulmonary disease, asthma, sympathetic dysfunction, atrial fibrillation, beta-blockers, pacemakers) of the patient are used as inputs to the adaptive fuzzy neural network system (ANFIS).
ANFIS overview
ANFIS is a hybrid of a fuzzy logic system and a neural network that does not assume any mathematical function that controls the relationship between input and output. ANFIS employs a data-driven approach in which training data determines the behavior of the system.
Five layers of ANFIS have the following functions:
in layer 1, three parameters are stored per cell to define a bell-shaped membership function. Each unit is connected to exactly one input unit and the membership degree of the obtained input value is calculated.
In layer 2, each rule is represented by one element in layer 2. Each cell is connected to a cell in the previous layer that is associated with the rule premise. The inputs to the cell are membership degrees, which are multiplied to determine the degree of implementation of the rule represented.
In layer 3, there is one unit to calculate its relative implementation by normalizing the equation for each rule. Each cell is connected to all regular cells in layer 2.
In layer 4, the cells of layer 4 are connected to all input cells, and one cell is connected to exactly one cell of layer 3. Each unit calculates the output of the rule.
In layer 5, the output unit of layer 5 calculates the final output by summing all the outputs of layer 4.
ANFIS applies a standard learning process of neural network theory. Reverse transfer is used to learn the precondition parameters, i.e., membership functions, while least squares estimation is used to determine coefficients for linear combinations in rule conclusions. The steps in the learning process are passed on twice. In the first pass, forward pass, the input pattern is passed and the optimal conclusion parameters are estimated by an iterative least mean square process, while the premise parameters remain unchanged in the current loop of the training set. In the second pass, the reverse pass, the pattern is passed again and in this pass the reverse pass is used to modify the precondition parameters while the conclusion parameters remain unchanged. The process then iterates according to the number of training cycles required. If the preconditions are initially properly selected based on expert knowledge, a training period is usually sufficient because the LMS algorithm can determine the optimal conclusion parameters in one pass, and if the preconditions parameters do not change significantly under the gradient descent method, the LMS calculation of the conclusion parameters will not produce other results. For example, in a 2-input, 2-rule system, rule 1 is defined as:
if x is A and y is B, f 1 ＝p 1 x+q 1 y+r 1
Where p, q and r are all linear, referred to as conclusion parameters or unique conclusions. The first order f is most common, as the higher order Sugeno fuzzy model introduces great complexity but no significant advantage.
Number of categories
The inputs to the ANFIS system are obfuscated into a plurality of predetermined categories. The number of categories should be greater than or equal to two. The number of categories may be determined by different methods. In conventional fuzzy logic, categories are defined by experts. This method can only be applied if the expert has a clear view of where landmarks (landarrays) between the two categories can be placed. ANFIS optimizes the location of landmarks, however, if the initial value of the parameters defining the class is close to the optimal value, the gradient descent method will reach its minimum faster. By default, the initial landmark of an ANFIS is selected by dividing the interval of all data from minimum to maximum into n equidistant intervals, where n is the number of categories. The number of categories may also be selected by means of various clustering methods or Markov models, by plotting the data as a histogram and intuitively deciding on a sufficient number of categories, or by sorting by Fuzzy Inductive Reasoning (FIR). The present application selects the default value of ANFIS and finds more than 3 categories during the verification phase that would result in instability, thus using 2 or 3 categories.
Number of inputs
Both the number of categories and the number of inputs increase the complexity of the model, i.e. the number of parameters. For example, a system with 4 inputs, each blurred into 3 categories, consisting of 36 precondition (non-linear) parameters and 405 conclusion (linear) parameters, is calculated by the following two formulas:
premise = number of categories x number of inputs x 3
Conclusion = number of categories Number of inputs X (number of inputs +1)
To obtain a meaningful solution for the parameters, the number of input-output pairs should typically be much larger (at least 10 times) than the number of parameters.
Stability criterion
Unfortunately, there is currently no definition of stability criteria for a neuro-fuzzy system. The most useful tool for ensuring stability is the experience obtained by testing in the context of a specific data set, using certain neuro-fuzzy systems such as ANFIS and using extreme data obtained, for example, by simulation.
Number of training periods
The ANFIS uses Root Mean Square Error (RMSE) to verify the training results and may calculate RMSE verification errors from a set of verification data after each training period. A period is defined as one update of the precondition parameters and conclusion parameters. The increased number of cycles generally reduces training errors.
Drawings
Fig. 1: the extracted parameters are fed to an ANFIS, which is a mixture of neural networks and fuzzy logic systems. The inputs include at least the following 3 parameters: HR (6), RRv (7), respR (8), TVv (9), respRacc (10), cross-information (CMIbreath) between HR, RRv and RespRacc (5). The output of the ANFIS model (11) is an intelligent respiratory index (SRI) (12), which is a number of units free from 0 to 100, wherein the decreasing value corresponds to the degree of deterioration of the respiratory function of the patient.
Fig. 2: the extracted parameters are fed to an ANFIS, which is a mixture of neural networks and fuzzy logic systems. The inputs include at least the following 3 parameters: HR (19), RRv (20), respR (21), TVv (22), respRacc (23), pulse oximeter (17), cross-information (CMIbreath) between HR, RRv and RespRacc (18). The output of the ANFIS (25) model is the intelligent respiratory index (SRI) (26), which is a number of units free from 0 to 100, with a reduced value corresponding to the degree of deterioration of the respiratory function of the patient.
Fig. 3: the extracted parameters are fed to an ANFIS, which is a mixture of neural networks and fuzzy logic systems. The inputs include at least the following 3 parameters: HR (34), RRv (35), respR (36), TVv (37), respRacc (38), pulse oximetry (39), cross-interaction information (CMIbreath) (32) between HR, RRv and RespRacc, demographic profile data such as gender, age and Body Mass Index (BMI). The output of the ANFIS model (40) is an intelligent respiratory index (SRI) (41), which is a number of units free from 0 to 100, wherein the reduced value corresponds to a degree of deterioration of the respiratory function of the patient.
Fig. 4: this figure shows how the breathing patch (42) is attached to a person's chest, and the location of the patch.
Fig. 5: the patch consists of an amplifier (43) for the ECG (51), an accelerometer (52), a microphone (53), a radio transmitter module (e.g. a bluetooth low energy module (46)), a battery (45) and four electrodes (47-50) which are simultaneously used for attaching the patch to the patient.
Fig. 6: this figure shows the digital processing of acquired signals such as ECG (54), respiratory sounds (55) and chest movement (56). The figure also shows the parameters obtained from each signal, HR, RRv from ECG (54), respR, TVv from microphone (55), respRacc from accelerometer (56).
Fig. 7: this figure shows a schematic diagram of the cyclical respiratory sounds of inspiration and expiration, wherein the amplitude envelope is also plotted.
Fig. 8: this table represents the relationship between clinical status and intelligent respiratory index (SRI). SRI is a progressive scale where 100 corresponds to normal respiratory function and a decreasing value reflects a deterioration of respiratory function, denoted by 0 when breathing ceases.
Fig. 9: this figure shows one of the Graphical User Interfaces (GUIs) of the display, where SRI is the most important parameter and thus the largest parameter. The GUI also displays RR and HR.
Claims (20)
1. A method of determining the respiratory quality of a patient by combining parameters extracted from an electrocardiogram (2, 14, 29), chest movement and acceleration (4, 16, 31) and respiratory acoustics (3, 15, 30), characterized in that the method comprises the steps of:
a) Measuring an electrocardiogram (2, 14, 29) using the sensor;
b) Measuring chest movement and acceleration (4, 16, 31) using sensors;
c) Measuring respiratory acoustics (3, 15, 30) using a sensor;
d) Calculating heart rate variability (7, 20, 35) from an electrocardiogram (2, 14, 29) using a fast fourier transform or a Choi-Williams distribution;
e) Calculating a tidal volume variation (9, 22, 37) based on the amplitude of the respiratory acoustics (8, 21, 36) and the first derivative of the amplitude envelope;
f) Calculating a respiratory rate from the chest movement and acceleration (10, 23, 38);
g) Calculating cross-mutual information (5, 18, 32) between heart rate variability (7, 20, 35), respiratory acoustics (3, 15, 30), chest movement (4, 16, 31) and tidal volume variability (9, 22, 37) of an electrocardiogram as input to an intelligent respiratory index (12, 26, 41);
h) At least 3 parameters extracted from the electrocardiogram (2, 14, 29), the respiratory acoustics (3, 15, 30), the chest movement (4, 16, 31) and the cross mutual information (5, 18, 32) of the three are combined into an index (12, 26, 41) of the respiratory quality level using an adaptive fuzzy neural network system (11, 25, 40) or any other classifier.
2. The method according to claim 1, wherein step a is characterized by measuring through a patch consisting of 2 or more electrodes to determine an electrocardiogram (2, 14, 29) located on the upper or lower chest of the subject.
3. Method according to claim 1, wherein step b is characterized in that chest acceleration is recorded by means of an accelerometer (4, 16, 31) integrated in the patch, whereby the breathing frequency is calculated based on the acceleration (10, 23, 38).
4. Method according to claim 1, wherein step c is characterized by recording breath acoustics (3, 15, 30), inspiration and expiration by means of a microphone integrated in the patch (53), thereby calculating the breathing frequency (8, 21, 36).
5. The method of claim 1, wherein step e is characterized by estimating the flow by the following formula,
F＝k 1 max (dA bag)Collaterals/dt) +k 2 A β
Where F is the flow rate, A is the amplitude of the respiratory acoustics (3, 15, 30), so the tidal volume variation (9, 22, 37) is the integral of the flow rate over time,
6. method according to claim 1, wherein step f is characterized by extracting breath from the acceleration (10, 23, 38) by applying a hilbert transformation to the acceleration signal to calculate the breathing frequency, wherein the maximum energy component of the acquired acceleration signal is proportional to the breathing frequency.
7. The method according to claim 1, wherein step g is characterized by calculating cross-information (5, 18, 32) between heart rate variability (7, 20, 35) extracted from an electrocardiogram (2, 14, 29), features extracted from respiratory acoustics (3, 15, 30) and features extracted from accelerations (4, 16, 31).
8. Method according to claim 1, wherein step h is characterized in that heart rate variability (7, 20, 35) extracted from the electrocardiogram (2, 14, 29) and tidal volume variability (9, 22, 37) calculated from the respiratory acoustics (3, 15, 30), and cross-correlation information (5, 18, 32) are used as inputs to an adaptive fuzzy neural network system (11, 25, 40) or other classifier, and the output thereof is then an indicator (12, 26, 41) of respiratory quality.
9. The method according to claim 1, wherein the formula of the index (12, 26, 41) of respiratory quality, step h, is a linear or quadratic function of respiratory frequency (8, 21, 36), heart rate (6, 19, 34), heart rate variability (7, 20, 35) and tidal volume variability (9, 22, 37), the formula being as follows:
SRI＝k1*RespR+k2*HR+K3*RRV+k4*TVv+k5*RespR*HR*RRV*TVv，
wherein the constants k1 to k4 should be within the following range:
0.30<k1<0.7，
0.2<k2<0.4，
0.05<k3<0.2，
0.01<k4<0.1。
10. the method according to claims 1 to 10, wherein the method is integrated in a wireless patch (1, 13, 27, 42), the patch (1, 13, 27, 42) comprising interconnected electrocardiogram sensor (51) and amplifier (43), microphone (53), accelerometer (52), battery (55) and radio transmitter, e.g. bluetooth low energy module (46), transmitting data to an external viewer device (57).
11. A device for determining the respiratory quality of a patient, wherein the device is configured to combine parameters extracted from an electrocardiogram (2, 14, 29), chest movement and acceleration (4, 16, 31) and respiratory acoustics (3, 15, 30), characterized in that the device performs the steps of:
a) Measuring an electrocardiogram (2, 14, 29) using the sensor;
b) Measuring chest movement and acceleration (4, 16, 31) using sensors;
c) Measuring respiratory acoustics (3, 15, 30) using a sensor;
d) Calculating heart rate variability (7, 20, 35) from an electrocardiogram (2, 14, 29) using a fast fourier transform or a Choi-Williams distribution;
e) Calculating a tidal volume variation (9, 22, 37) based on the amplitude of the respiratory acoustics (8, 21, 36) and the first derivative of the amplitude envelope;
f) Calculating a respiratory rate from the chest movement and acceleration (10, 23, 38);
g) Calculating cross-mutual information (5, 18, 32) between heart rate variability (7, 20, 35), respiratory acoustics (8, 21, 36), chest movement (4, 16, 31) and tidal volume variability (9, 22, 37) of an electrocardiogram as input to an intelligent respiratory index (12, 26, 41);
h) At least 3 parameters extracted from the electrocardiogram (2, 14, 29), the respiratory acoustics (3, 15, 30), the chest movement (4, 16, 31) and the cross mutual information (5, 18, 32) of the three are combined into an index (12, 26, 41) of the respiratory quality level using an adaptive fuzzy neural network system (11, 25, 40) or any other classifier.
12. The device according to claim 11, wherein step a is characterized by a patch (1, 13, 27, 42) consisting of 2 or more electrodes (47, 48, 49, 50) to determine an electrocardiogram (2, 14, 29) located on the upper or lower chest of the subject.
13. The device according to claim 11, wherein step b is characterized in that the device is configured to record chest acceleration by integrating an accelerometer (4, 16, 31) in the patch (42) to calculate the respiratory rate based on the acceleration (10, 23, 38).
14. The device according to claim 11, wherein step c is characterized in that the device is configured to calculate the breathing frequency (8, 21, 36) by integrating a microphone (53) in the patch (1, 13, 27, 42) to record breath acoustics (3, 15, 30), inspiration and expiration.
15. The apparatus of claim 11, wherein step e is characterized in that the apparatus is configured to estimate the flow by the following formula:
F＝k 1 max (dA envelope/dt) +k 2 A β ，
Where F is flow, A is the amplitude of the respiratory acoustics, so the tidal volume variation is the integral of flow over time,
16. the device according to claim 11, wherein step f is characterized in that the device is configured to extract respiration from the acceleration (10, 23, 38) by applying a hilbert transformation to the acceleration signal (4, 16, 31) to calculate the respiration rate, wherein the maximum energy component of the acquired acceleration signal is proportional to the respiration rate.
17. The device according to claim 11, wherein step g is characterized in that the device is configured for calculating cross-information (5, 18, 32) between heart rate variability (7, 20, 35) extracted from an electrocardiogram (2, 14, 29), features extracted from respiratory acoustics (3, 15, 30) and features extracted from accelerations (4, 16, 31).
18. The device according to claim 11, wherein step h is characterized in that the device is configured to use the heart rate variability extracted from the electrocardiogram (7, 20, 35) and the tidal volume variability calculated from the respiratory acoustics (9, 22, 37), and the cross-over information (5, 18, 32) as inputs to an adaptive fuzzy neural network system (11, 25, 40) or other classifier, and its output is an indicator (12, 26, 41) of the respiratory quality.
19. The device according to claim 11, wherein step h is characterized in that the formula of the index of respiratory quality level (12, 26, 41) is a linear or quadratic function of respiratory rate (8, 21, 36), heart rate (6, 19, 34), heart rate variability (7, 20, 35) and tidal volume variability (9, 22, 37), the formula being as follows:
SRI＝k1*RespR+k2*HR+K3*RRV+k4*TVv+k5*RespR*HR*RRV*TVv，
wherein the constants k1 to k4 should be within the following range:
0.30<k1<0.7，
0.2<k2<0.4，
0.05<k3<0.2，
0.01<k4<0.1。
20. the device according to claims 11 to 19, wherein the device is integrated in a wireless patch (1, 13, 27, 42), the patch (1, 13, 27, 42) comprising interconnected electrocardiogram sensor (51) and amplifier (43), microphone (53), accelerometer (52), battery (45) and radio transmitter, e.g. bluetooth low energy module (46), for transmitting data to an external viewer device (57).
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/IB2021/059526 WO2023062420A1 (en) | 2021-10-15 | 2021-10-15 | Method and apparatus for smart respiratory monitoring by electrocardiogram, breath acoustics and thoracic acceleration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116761544A true CN116761544A (en) | 2023-09-15 |
Family
ID=78819564
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202180089672.2A Pending CN116761544A (en) | 2021-10-15 | 2021-10-15 | Method and device for intelligent respiration monitoring by electrocardiogram, respiration acoustics and chest acceleration |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN116761544A (en) |
| WO (1) | WO2023062420A1 (en) |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6918878B2 (en) | 2003-06-13 | 2005-07-19 | Ge Medical Systems Information Technologies, Inc. | Methods and systems for monitoring respiration |
| US7733224B2 (en) | 2006-06-30 | 2010-06-08 | Bao Tran | Mesh network personal emergency response appliance |
| EP2142095A1 (en) | 2007-05-02 | 2010-01-13 | Earlysense Ltd. | Monitoring, predicting and treating clinical episodes |
| US10463340B2 (en) | 2009-10-15 | 2019-11-05 | Masimo Corporation | Acoustic respiratory monitoring systems and methods |
| US9066680B1 (en) | 2009-10-15 | 2015-06-30 | Masimo Corporation | System for determining confidence in respiratory rate measurements |
| US9492138B2 (en) | 2012-10-15 | 2016-11-15 | Rijuven Corp | Mobile front-end system for comprehensive cardiac diagnosis |
| US20150313484A1 (en) | 2014-01-06 | 2015-11-05 | Scanadu Incorporated | Portable device with multiple integrated sensors for vital signs scanning |
| US20190000349A1 (en) * | 2017-06-28 | 2019-01-03 | Incyphae Inc. | Diagnosis tailoring of health and disease |
| US10881330B2 (en) | 2018-08-20 | 2021-01-05 | Thomas Jefferson University | Acoustic sensor and ventilation monitoring system |
| EP3698715A1 (en) * | 2019-02-19 | 2020-08-26 | Koninklijke Philips N.V. | A sleep monitoring and position therapy system and method |
-
2021
- 2021-10-15 CN CN202180089672.2A patent/CN116761544A/en active Pending
- 2021-10-15 WO PCT/IB2021/059526 patent/WO2023062420A1/en active Search and Examination
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023062420A1 (en) | 2023-04-20 |
| WO2023062420A9 (en) | 2023-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2021203784B2 (en) | Multivariate residual-based health index for human health monitoring | |
| JP5887057B2 (en) | Device for measuring and predicting patient respiratory stability | |
| US20230210401A1 (en) | System and method for non-invasively determining an internal component of respiratory effort | |
| Bozkurt et al. | Detection of abnormal respiratory events with single channel ECG and hybrid machine learning model in patients with obstructive sleep apnea | |
| US20220167856A1 (en) | Lung function monitoring from heart signals | |
| EP2874539A1 (en) | A method and system for determining the state of a person | |
| CN103892812A (en) | Matching degree analysis technology | |
| JP2023518805A (en) | Devices for predicting, identifying and/or managing pneumonia or other medical conditions | |
| WO2023087512A1 (en) | Depression objective screening system, method, and device | |
| EP3203900A1 (en) | Weaning readiness indicator, sleeping status recording device, and air providing system applying nonlinear time-frequency analysis | |
| CN105266764B (en) | A kind of traditional Chinese medical science pectoral qi assessment device | |
| US11291406B2 (en) | System for determining a set of at least one cardio-respiratory descriptor of an individual during sleep | |
| Orphanidou et al. | Telemetry-based vital sign monitoring for ambulatory hospital patients | |
| US20230007922A1 (en) | Method and System for Estimating Physiological Parameters Utilizing a Deep Neural Network to Build a Calibrated Parameter Model | |
| CN116761544A (en) | Method and device for intelligent respiration monitoring by electrocardiogram, respiration acoustics and chest acceleration | |
| US20220401037A1 (en) | Ml-based anomaly detection and descriptive root-cause analysis for biodata | |
| JP4745236B2 (en) | System and method for estimating signal artifacts | |
| CN111840931A (en) | Double-level respiratory function monitoring and intervention equipment | |
| TWM593242U (en) | Detection device for apnea based on chest respiratory signal | |
| JPWO2020196323A1 (en) | Programs, information processing methods and information processing equipment | |
| AU2011203044B2 (en) | Systems and methods for respiratory event detection | |
| Bashi et al. | Diagnosis of obstructive apnea disease AHI in chemical warfare veterans based on HRV signals analysis using the ANFIS neural network | |
| CN212880869U (en) | Double-level respiratory function monitoring and intervention equipment | |
| TWI772086B (en) | A method of monitoring apnea and hypopnea events by using fully convolutional networks | |
| WO2024047621A1 (en) | Method and system for estimating physiological parameters utilizing a deep neural network to build a calibrated parameter model |
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 |