Esophageal pressure signal filtering method and filtering device

Abstract

The invention provides an esophageal pressure signal filtering method and an esophageal pressure filtering device. The method comprises the following steps: acquiring an esophageal pressure signal of a patient; constructing a reference signal according to the esophageal pressure signal; updating filtering parameters of an adaptive filter according to the reference signal to obtain an updated adaptive filter; and filtering a heartbeat incisura signal in the esophageal pressure signal according to the updated adaptive filter. According to the filtering method for the esophageal pressure signal, the noise signal in the esophageal pressure signal can be effectively filtered out.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a filtering method and a filtering device for an esophageal pressure signal. Background Art

[0002] During clinical mechanical ventilation, doctors can insert a special esophageal pressure catheter with an air bag into the patient's esophagus to obtain the changes in the patient's esophageal pressure. The effectiveness of using esophageal pressure (Pes) as an approximate substitute for intra plural pressure (Ppl) has been widely recognized by doctors around the world. Doctors can use the above method to approximately obtain the patient's intra plural pressure, thereby calculating the patient's transpulmonary pressure, chest wall compliance, lung compliance, transpulmonary driving pressure and other parameters. At this point, the impact of the gas delivery pressure on the patient's lungs and chest wall during mechanical ventilation can be accurately obtained, and the patient's ventilation process can be further described and evaluated.

[0003] To obtain accurate esophageal pressure waveforms through an esophageal pressure catheter, the esophageal pressure catheter must first be placed in the correct position of the patient's esophagus. After the esophageal pressure catheter is placed, the position of the balloon in the esophagus is close to the heart, and the patient's heartbeat will also affect the pressure measured by the esophageal pressure catheter. Therefore, the esophageal pressure waveform is often interfered by the patient's heartbeat, which is usually called a cardiac notch or cardiac artifact (cardiogenic oscillation). The presence of the cardiac notch has a certain interference with the obtained esophageal pressure value, especially when the amplitude of the patient's cardiac notch is large. The noise signal brought by the cardiac notch may also cover the patient's own esophageal pressure signal, resulting in inaccurate esophageal pressure values, affecting the subsequent further calculation of the patient's lung physiological parameters. In addition, when the doctor needs to evaluate and identify the patient's inspiratory effort through the swing of esophageal pressure, the presence of the cardiac notch will affect the doctor's judgment, making it difficult for the doctor to accurately evaluate the patient's inspiratory effort, as well as to identify and analyze the patient's human-machine confrontation.

[0004] In summary, the esophageal pressure signal will be interfered by the heartbeat notch signal, so esophageal pressure filtering is very necessary and important.

[0005] At present, scholars have not done much research on esophageal pressure filtering methods. For example, Schuessler used the esophageal pressure signal after high-pass filtering and the R wave signal extracted from ECG (electrocardiogram) to construct an adaptive denoiser. Graβhoff uses the template subtraction method, which requires the introduction of ECG / EMG (electromyogram) signals to analyze the noise template, but both methods require a monitor to be installed next to the ventilator, which is inconvenient to use in clinical practice. Summary of the Invention

[0006] The present invention mainly provides a filtering method for esophageal pressure signals. This filtering method can perform real-time filtering on the heartbeat notch signals in the esophageal pressure signals without connecting other devices to ventilation devices (such as ventilators, anesthesia machines, etc.), thereby meeting the filtering requirements for esophageal pressure signals in ventilation devices. The present invention also provides a corresponding filtering device.

[0007] According to a first aspect, an embodiment provides a filtering method for esophageal pressure signals, including the steps of:

[0008] Obtain the esophageal pressure signal of a patient;

[0009] Construct a reference signal according to the esophageal pressure signal;

[0010] Update the filtering parameters of an adaptive filter according to the reference signal to obtain an updated adaptive filter;

[0011] Filter out the heartbeat notch signals in the esophageal pressure signal according to the updated adaptive filter.

[0012] According to a second aspect, an embodiment provides a filtering method for esophageal pressure signals, including the steps of:

[0013] Obtain the esophageal pressure signal and blood oxygen signal of a patient;

[0014] Construct a reference signal according to the blood oxygen signal;

[0015] Update the filtering parameters of an adaptive filter according to the reference signal to obtain an updated adaptive filter;

[0016] Filter out the heartbeat notch signals in the esophageal pressure signal according to the updated adaptive filter.

[0017] According to a third aspect, an embodiment provides a filtering method for physiological signals, including the steps of:

[0018] Obtain the physiological signal of a patient that changes over time;

[0019] Construct a reference signal that is close to the frequency-domain distribution or time-domain distribution of the noise signal in the physiological signal according to the physiological signal;

[0020] Update the filtering parameters of an adaptive filter according to the reference signal to obtain an updated adaptive filter;

[0021] Filter out the noise signal in the physiological signal according to the updated adaptive filter.

[0022] According to a fourth aspect, an embodiment provides a filtering device for esophageal pressure signals, including:

[0023] A sensor interface that receives the esophageal pressure value of a patient monitored by a pressure sensor;

[0024] A processor, which is signal - connected to the sensor interface, receives the esophageal pressure value and generates an esophageal pressure signal, constructs a reference signal according to the esophageal pressure signal, updates the filtering parameters of an adaptive filter according to the reference signal to obtain an updated adaptive filter, and filters out the heartbeat notch signal in the esophageal pressure signal according to the updated adaptive filter.

[0025] According to the fifth aspect, an embodiment provides a ventilator, comprising:

[0026] An air source interface connected to an external air source;

[0027] A breathing circuit that connects the air source interface and the respiratory system of a patient to input the gas provided by the air source to the patient and receive the gas exhaled by the patient;

[0028] A breathing assistance device that provides breathing support power to control the output of the gas provided by the air source to the patient, collects and re - uses or discharges the gas exhaled by the patient to the external environment, and the breathing assistance device includes a machine - controlled ventilation module and / or a manual ventilation module; and

[0029] The filtering device for the esophageal pressure signal in the above - mentioned fourth aspect.

[0030] According to the sixth aspect, an embodiment provides a ventilation device, comprising:

[0031] A memory for storing programs;

[0032] A processor for implementing the method described in the first aspect by executing the programs stored in the memory.

[0033] According to the seventh aspect, an embodiment provides a medical device, comprising:

[0034] A memory for storing programs;

[0035] A processor for implementing the method described in the second aspect or the third aspect by executing the programs stored in the memory.

[0036] According to the eighth aspect, an embodiment provides a computer - readable storage medium, comprising programs that can be executed by a processor to implement the methods described in the first aspect to the third aspect.

[0037] The above-mentioned filtering method for esophageal pressure signals constructs a reference signal based on the esophageal pressure signal itself during ventilation, and filters the esophageal pressure signal according to the reference signal and an adaptive filter. Only the esophageal pressure signal needs to be collected. Therefore, during the filtering process of the esophageal pressure signal, no other external devices need to be introduced except for the ventilation device used for mechanical ventilation. Moreover, with the adaptive filtering method, a good filtering effect on the esophageal pressure signal can be achieved. Description of the Drawings

[0038] Figure 1 It is a schematic structural diagram of an adaptive noise canceller;

[0039] Figure 2 It is a schematic structural composition diagram of a ventilator in an embodiment;

[0040] Figure 3 It is a waveform diagram of an esophageal pressure signal in an embodiment;

[0041] Figure 4 It is a schematic diagram of the time interval between wave peaks within one cycle in an esophageal pressure signal in an embodiment;

[0042] Figure 5 It is a frequency spectrum diagram of an esophageal pressure signal in an embodiment;

[0043] Figure 6 It is a frequency spectrum diagram of a reference signal in an embodiment;

[0044] Figure 7 It is a frequency spectrum diagram of a reference signal in another embodiment;

[0045] Figure 8 It is a schematic structural diagram of the filtering process of an adaptive filter in an embodiment;

[0046] Figure 9 It is a waveform diagram and a frequency spectrum diagram of a filtered esophageal pressure signal in an embodiment;

[0047] Figure 10 It is a flowchart of a filtering method for an esophageal pressure signal in an embodiment;

[0048] Figure 11 It is a flowchart of an update method for an adaptive filter in an embodiment;

[0049] 10. Gas source interface;

[0050] 20. Respiratory assist device;

[0051] 30. Respiratory circuit;

[0052] 30a. Inspiratory passage; 30b. Expiratory passage; 31. Carbon dioxide absorber; 32. Check valve; 33. Patient interface;

[0053] 40. Sensor interface;

[0054] 50. Memory;

[0055] 60. Processor;

[0056] 70. Display;

[0057] 80. Patient. Detailed implementation manners

[0058] The present invention will be further described in detail below in conjunction with the accompanying drawings through specific implementation manners. Similar elements in different implementation manners adopt related similar element numbers. In the following implementation manners, many detailed descriptions are provided to enable a better understanding of the present application. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, and methods. In some cases, some operations related to the present application are not shown or described in the specification, which is to avoid the core part of the present application being overwhelmed by excessive descriptions. For those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the descriptions in the specification and the general technical knowledge in the art.

[0059] In addition, the features, operations, or characteristics described in the specification can be combined in any appropriate manner to form various implementation manners. At the same time, the steps or actions in the method description can also be reordered or adjusted in an obvious manner by those skilled in the art. Therefore, the various sequences in the specification and the drawings are only for clearly describing a certain embodiment and do not mean that they are necessary sequences, unless it is stated that a certain sequence must be followed.

[0060] The serial numbers assigned to the components in this article, such as "first", "second", etc., are only used to distinguish the described objects and do not have any sequential or technical meanings. The "connection" and "coupling" mentioned in this application, unless otherwise specified, both include direct and indirect connections (couplings).

[0061] The adaptive filter referred to in the present invention refers to an algorithm or device that, based on the estimation of the statistical characteristics of the input and output signals, automatically adjusts the filter coefficients by adopting a specific algorithm to achieve the best filtering characteristics. The structure of a typical adaptive noise canceller applying the adaptive filtering method is as Figure 1As shown, this structure requires two input signals: (1) a basic signal s + n, which is considered to be the superposition of the target signal s + n and the noise signal n; (2) a reference signal n0, which is correlated with the noise signal n but uncorrelated with the target signal s. The function of this noise canceller is to obtain an output signal approximating the target signal s from the input basic signal s + n The specific filtering process is as follows:

[0062] Input the reference signal n0 into the adaptive filter to obtain the output signal of the filter Input the basic signal s + n into the noise canceller and subtract the output signal of the filter from the basic signal s + n In this way, the output signal of the noise canceller is obtained At the same time, obtain the error signal ε between the basic signal s + n and the output signal of the filter, and then continuously reduce the error signal ε by optimizing the parameters of the adaptive filter and other methods according to the adaptive filtering algorithm. Commonly used adaptive filtering algorithms include the Least Mean Square (LMS) algorithm, the Recursive Least Square algorithm, the Kalman filter algorithm (Recursive Least Square, RLS), and the neural network algorithm, etc., as well as improved algorithms derived from these algorithms, such as the normalized LMS, variable step size LMS, fast RLS, square root RLS, and RLS algorithm based on QR decomposition, etc. Under the rules of the above filter algorithms, the filter automatically optimizes the filter parameters and / or structure according to the filtering result of the previous moment to adapt to the unpredictable and time-varying statistical characteristics of the signal and noise. The final effect is that the filter output signal is the optimal estimate of the noise signal n under this rule, and the output signal is also approximately the target signal s. The adaptive filtering algorithm does not require prior information about the signal and noise, has a small computational amount, is suitable for real-time processing, and is widely used in many engineering problems

[0063] In the present invention, the fundamental wave refers to several types of waveforms that can be superimposed and changed to obtain other waveforms, and these several types of waveforms are called fundamental waves

[0064] In the present invention, the fundamental frequency, also known as the base frequency, refers to the lowest frequency in a complex wave. And any periodic waveform can be decomposed into a fundamental frequency sine wave plus many sine waves of higher frequencies. The higher frequencies are integer multiples (N, only an integer) of the base frequency, and the sine waves of higher frequencies can be considered as the harmonics referred to in the present invention

[0065] Please refer to Figure 2 , Figure 2 ​The following is a schematic diagram of the structural composition of a ventilator (or it can also be called a ventilation device). The ventilator includes a gas source interface 10, a respiratory assistance device 20, a respiratory circuit 30, a sensor interface 40, a memory 50, a processor 60, and a display 70.

[0066] The gas source interface 10 is used to connect to a gas source (not shown in the figure), and the gas source is used to provide gas. This gas can usually be oxygen, air, etc. In some embodiments, the gas source can be a compressed gas cylinder or a central gas supply source, and the ventilator is supplied with gas through the gas source interface 10. The types of gas supplied include oxygen O2, air, etc. The gas source interface 10 may include conventional components such as a pressure gauge, a pressure regulator, a flow meter, a pressure reducing valve, and a proportional control protection device, which are respectively used to control the flow rates of various gases (such as oxygen and air). The gas input through the gas source interface 10 enters the respiratory circuit 30 and forms a mixed gas with the gas originally in the respiratory circuit 30.

[0067] The respiratory assistance device 20 is used to provide power for the non - autonomous breathing of the patient 80 and maintain the patency of the airway, that is, to drive the gas input through the gas source interface 10 and the mixed gas in the respiratory circuit 30 into the respiratory system of the patient 80, and drain the gas exhaled by the patient 80 into the respiratory circuit 30, thereby improving ventilation and oxygenation and preventing hypoxia of the patient 80's body and the accumulation of carbon dioxide in the patient 80's body. In a specific embodiment, the respiratory assistance device 20 usually includes a mechanical ventilation module, and the air flow pipeline of the mechanical ventilation module is connected to the respiratory circuit 30. In the state where the patient 80 has not recovered autonomous breathing during the operation, the mechanical ventilation module is used to provide the power for the patient 80's breathing. In some embodiments, the respiratory assistance device 20 further includes a manual ventilation module, and the air flow pipeline of the manual ventilation module is connected to the respiratory circuit 30. During the induction stage before intubating the patient 80 during the operation, it is usually necessary to use the manual ventilation module to assist the patient 80's breathing. When the respiratory assistance device 20 includes both a mechanical ventilation module and a manual ventilation module, the mechanical or manual ventilation mode can be switched through a mechanical or manual switch (such as a three - way valve) so as to connect the mechanical ventilation module or the manual ventilation module to the respiratory circuit 30, thereby controlling the breathing of the patient 80. Those skilled in the art should understand that according to specific needs, the ventilator may only include a mechanical ventilation module or a manual ventilation module.

[0068] The breathing circuit 30 includes an inhalation passage 30a, an exhalation passage 30b, and a carbon dioxide absorber 31. The inhalation passage 30a and the exhalation passage 30b communicate to form a closed circuit, and the carbon dioxide absorber 31 is provided on the pipeline of the exhalation passage 30b. The mixed gas of fresh air introduced by the gas source interface 10 is input from the inlet of the inhalation passage 30a and provided to the patient 80 through the patient interface 33 provided at the outlet of the inhalation passage 30a. The patient interface 33 can be a face mask, a nasal cannula, or an endotracheal tube. In a preferred embodiment, a one-way valve 32 is provided on the inhalation passage 30a, and the one-way valve 32 opens during inhalation and closes during exhalation. A one-way valve 32 is also provided on the exhalation passage 30b, and the one-way valve 32 closes during inhalation and opens during exhalation. The inlet of the exhalation passage 30b communicates with the patient interface 33. When the patient 80 exhales, the exhaled gas enters the carbon dioxide absorber 31 through the exhalation passage 30b, and the carbon dioxide in the exhaled gas is filtered by the substance in the carbon dioxide absorber 31. The gas after filtering the carbon dioxide is recycled into the inhalation passage 30a. In some embodiments, a flow sensor and / or a pressure sensor are further provided in the breathing circuit 30 for detecting the gas flow and / or the pressure in the pipeline respectively.

[0069] The sensor interface 40 is used to receive various breathing information of the patient 80 collected by the sensor. For example, the esophageal pressure value of the patient 80 collected by the sensor under the state of machine-assisted ventilation. Specifically, the sensor interface 40 is connected to the signal output end of the pressure sensor. In one embodiment, the sensor interface 40 can just be a connector between the sensor output end and the subsequent circuit (such as the processor 60) without processing the signal. The sensor interface 40 can also be integrated into the processor 60 as an interface for the processor 60 to access the signal. In another embodiment, the sensor interface 40 can include an amplification circuit, a filtering circuit, and an A / D conversion circuit for respectively amplifying, filtering, and performing analog-to-digital conversion processing on the input analog signal. Of course, those skilled in the art should understand that the connection relationship among the amplification circuit, the filtering circuit, and the A / D conversion circuit can vary according to the specific design of the circuit, and one of the circuits can also be reduced. For example, the amplification circuit or the filtering circuit can be reduced, thereby reducing its corresponding function. In addition, the sensor interface 40 can also access the information collected by other sensors, such as the flow information and / or the pressure information of the breathing circuit 30.

[0070] As a specific product, the pressure sensor for monitoring the esophageal pressure value can be a part of the ventilator or an external accessory independent of the ventilator.

[0071] The memory 50 can be used to store data or programs, such as data collected by various sensors, data generated by the processor 60 through calculations, or image frames generated by the processor 60. The image frames can be 2D or 3D images. Alternatively, the memory 50 can store a graphical user interface, one or more default image display settings, and programming instructions for the processor 60. The memory 50 can be a tangible and non-transitory computer-readable medium, such as flash memory, RAM, ROM, EEPROM, etc.

[0072] The processor 60 is used to execute instructions or programs, control various control valves in the respiratory assistance device 20, the gas source interface 10, and / or the respiratory circuit 30, or process the received data to generate required calculation or judgment results, or generate visual data or graphics, and output the visual data or graphics to the display 70 for display. In this embodiment, the processor 60 is signal-connected to the sensor interface 40 to obtain the esophageal pressure signal of the patient 80, and construct a reference signal according to the esophageal pressure signal. For example, taking the esophageal pressure signal as the basic signal, a reference signal with a frequency domain distribution or a time domain distribution close to that of the heart rate notch signal is constructed. The reason for constructing a reference signal with a frequency domain distribution or a time domain distribution close to that of the heart rate notch signal is to obtain a signal with an energy distribution close to that of the heart rate notch signal.

[0073] Taking the construction of a reference signal with a frequency domain distribution close to that of the heart rate notch as an example, the construction process of the reference signal will be described below:

[0074] First, obtain the fundamental frequency of the heart rate notch signal in the esophageal pressure signal. Two ways to obtain the fundamental frequency will be described below.

[0075] The first way is to identify the time domain characteristics and / or frequency domain characteristics of the heart rate notch signal from the esophageal pressure signal, and calculate the fundamental frequency of the heart rate notch signal. Figure 3 The waveform diagram of a typical esophageal pressure signal is shown. The processor 60 can identify the waveform within at least one cycle of the esophageal pressure signal. For example Figure 3 The embodiment shown has multiple cycles. The processor 60 can select one of the cycles and calculate the time interval between two adjacent wave peaks within this cycle. This time interval is the cycle of the heart rate notch signal, and the reciprocal of this time interval can be used as the fundamental frequency of the heart rate notch signal.

[0076] In some embodiments, the average value of the time intervals between multiple consecutive wave peaks within one cycle can also be calculated (for example Figure 4 the dotted line in shows the time intervals between three consecutive wave peaks within one cycle), and this average value is used as the average value of the cycle of the heart rate notch signal. Then, the fundamental frequency of the heart rate notch signal can be obtained according to this average value, thus making the calculation of the heart rate notch signal more accurate.

[0077] The second method is to obtain the frequency distribution information of the heartbeat notch signal based on the short-time spectrum of the esophageal pressure signal calculation, so as to obtain the fundamental frequency of the heartbeat notch signal. For example, Figure 5 The spectrum diagram of a typical esophageal pressure signal is shown. The processor 60 can identify that the frequencies of the heartbeat notch signal are mainly distributed at the abscissas 1, 2, 3, 4, and 5. Therefore, 1 Hz can be used as the fundamental frequency of the heartbeat notch signal.

[0078] Both of the above two methods can obtain the fundamental frequency of the heartbeat notch signal. Either one of them or a combination of both can be used. In other embodiments, other feasible methods can also be adopted.

[0079] In some embodiments, after obtaining the fundamental frequency of the heartbeat notch signal, an external device can be introduced to verify the fundamental frequency of the heartbeat notch signal. For example, a monitor or an electrocardiogram device is used to obtain the pulse parameter and / or heart rate parameter of the patient 80, and the fundamental frequency of the heartbeat notch signal is verified according to the heart rate parameter and / or pulse rate parameter. For example, when the difference between the two does not exceed a preset threshold, it can be considered that the fundamental frequency of the heartbeat notch signal obtained by the ventilator is credible.

[0080] In some embodiments, the blood oxygen signal of the patient 80 can also be obtained. For example, the pulse parameter and / or heart rate parameter of the patient 80 is obtained through a monitor or an electrocardiogram device, and then the pulse parameter and / or heart rate parameter is used as the fundamental frequency of the heartbeat notch signal. In addition to the ventilator, this method also requires other medical devices or monitoring devices.

[0081] After determining the fundamental frequency of the heartbeat notch signal, a relevant reference signal can be constructed according to the fundamental frequency. The reference signal can be obtained by constructing a fundamental wave with the same frequency as the fundamental frequency of the heartbeat notch signal. The fundamental wave can be any one of a sine wave, a triangular wave, a square wave, a sawtooth wave, and a Gaussian function. The reference signal can be obtained according to the fundamental wave. For example, the fundamental wave can be directly used as the reference signal. The construction process of the fundamental wave will be illustrated by taking a sine wave as an example below. The construction method is as follows:

[0082] Noise_ref = A * sin(2 * π * HeartRate / (SamplingRate * 60) * t + B).

[0083] Among them, Noise_ref is the reference signal, HeartRate is the heart rate of the patient 80, Samplingrate is the sampling rate of the above ventilator when obtaining the esophageal pressure signal of the patient 80, A is the amplitude, and B is the phase. The amplitude, phase, and the frequency of the sine wave on the right side can be adjusted to meet the construction requirements.

[0084] According to the above formula, a sine wave with a frequency equal to the fundamental frequency of the heart rate notch signal can be constructed. For example, based on the fundamental frequency of the heart rate notch signal obtained from the esophageal pressure signal of Figure 3 , a reference signal is constructed. Figure 6 is the signal spectrogram after Fourier transform of the constructed reference signal. It can be seen that the frequency domain energy of this signal is mainly concentrated on the fundamental frequency of the heart rate notch signal. Therefore, the above method successfully constructs a suitable reference signal.

[0085] It should be noted that the above process of constructing the reference signal is an explanation of an implementation manner of the present invention. The method for constructing the reference signal of the present invention is not limited to the examples given.

[0086] In some embodiments, the reference signal may include not only the signal component of the fundamental frequency energy of the heart rate notch, but also the signal components at the respective harmonic frequencies of the heart rate notch. For example, the reference signal can be constructed in the following manner:

[0087] Noise_ref = A1*sin(2*π*HeartRate / (SamplingRate*60)*t + B1)+

[0088] A2*sin(2*π*HeartRate*2 / (SamplingRate*60)*t + B2)+

[0089] A3*sin(2*π*HeartRate*3 / (SamplingRate*60)*t + B3)+...

[0090] This formula can represent that the fundamental wave is superimposed with its 2nd to Nth harmonics to obtain the reference signal Noise_ref, where 2 ≤ N.

[0091] Among them, A2*sin(2*π*HeartRate*2 / (SamplingRate*60)*t + B2) contains the energy of the heart rate notch signal at the second harmonic.

[0092] A3*sin(2*π*HeartRate*3 / (SamplingRate*60)*t + B3) contains the energy of the heart rate notch signal at the third harmonic. The meanings of the symbols in this formula have been explained above and will not be elaborated here. Figure 7 is the signal spectrogram after Fourier transform of the reference signal constructed according to the above formula.

[0093] Through the above construction method, when filtering the esophageal pressure signal, the harmonic energy of the heart rate notch signal can be filtered out, so as to better filter the esophageal pressure.

[0094] The constructed reference signal includes the harmonic energy of the nth order of the incisura signal, which can be determined according to actual usage requirements. For example, it can be fixed that the reference signal is constructed to include the third harmonic energy of the incisura signal, or constructed to include the fifth harmonic energy of the incisura signal to meet the requirements of most filtering scenarios.

[0095] After obtaining the reference signal, the processor 60 updates the filtering parameters of the adaptive filter according to the obtained reference signal, thereby obtaining an updated adaptive filter. The classical noise canceller mentioned above is a way to update the filtering parameters of the adaptive filter using the reference signal. In addition, the filtering parameters of the adaptive filter can also be updated in the Figure 8 way shown below, which will be described in detail below.

[0096] As Figure 8 shown, the esophageal pressure signal is used as the basic signal, and on this basis, the reference signal is superimposed. The superimposed signal is used as the first input signal, and an adaptive filter is used to filter the first input signal to obtain an output signal. Then, the error signal between the esophageal pressure signal and the output signal is calculated. Finally, according to the error signal, the filtering parameters of the adaptive filter are updated using an adaptive algorithm. For example, any one or several of the step size, order, and coefficients of the adaptive filter can be updated, thereby obtaining an updated adaptive filter. The first input signal is the result of superimposing noise on the basic signal. Therefore, the result of adaptive filtering is to make the adaptive filter obtain an attenuation gain coefficient at the noise frequency, thereby reducing the noise signal in the first input signal. After obtaining the error signal, the filtering parameters of the adaptive filter can be updated using various algorithms mentioned above when introducing the noise canceller to minimize the error signal.

[0097] After obtaining the adaptive filter, the esophageal pressure signal can be filtered using the adaptive filter. One filtering method is as Figure 8 shown. The esophageal pressure signal is used as the second input signal, and the updated adaptive filter is used to filter the second input signal, thereby filtering out the incisura signal in the esophageal pressure signal. The filtering result is as Figure 9 shown, where the dark solid line part is the spectrogram and waveform diagram after filtering, and the light solid line is the spectrogram and waveform diagram before filtering. It can be clearly seen that the noise signal in the esophageal pressure signal is effectively filtered out.

[0098] Based on the above ventilator, the filtering method of the esophageal pressure signal will be described below.

[0099] As Figure 10 shown is a flowchart of a filtering method for the esophageal pressure in an embodiment, including the steps:

[0100] Step 100: Obtain the esophageal pressure signal of patient 80.

[0101] For example, when a ventilation device (such as an anesthesia machine and / or a ventilator) ventilates patient 80, the esophageal pressure signal of patient 80 is obtained through the sensors of the ventilation device itself.

[0102] Step 200: Construct a reference signal based on the esophageal pressure signal.

[0103] This reference signal can be close to the frequency-domain distribution of the heartbeat notch signal in the esophageal pressure signal, or close to the time-domain distribution of the heartbeat notch signal in the esophageal pressure signal. The reason for constructing a reference signal with a similar frequency-domain or time-domain distribution is to obtain a signal with an energy distribution close to that of the heartbeat notch signal.

[0104] In some embodiments, taking the esophageal pressure signal as the basic signal, constructing a reference signal close to the frequency-domain distribution of the heartbeat notch signal specifically includes the following steps:

[0105] Step 210: Obtain the fundamental frequency of the heartbeat notch signal in the esophageal pressure signal based on the esophageal pressure signal.

[0106] For example, identify the time-domain characteristics and / or frequency-domain characteristics of the heartbeat notch signal from the esophageal pressure signal, and calculate the fundamental frequency of the heartbeat notch signal. Figure 3 The waveform diagram of a typical esophageal pressure signal is shown. With the help of the ventilator that obtains the esophageal pressure signal, the waveform within at least one cycle of the esophageal pressure signal can be identified. For example, Figure 3 the embodiment shown has multiple cycles. Select one of the cycles and calculate the time interval between two adjacent wave peaks within this cycle. This time interval is the period of the heartbeat notch signal, and the reciprocal of this time interval can be used as the fundamental frequency of the heartbeat notch signal.

[0107] In some embodiments, the average value of the time intervals between multiple consecutive wave peaks within one cycle can also be calculated (for example, Figure 4 the dotted line in shows the time intervals between three consecutive wave peaks within one cycle), take this average value as the average value of the period of the heartbeat notch signal, and then the fundamental frequency of the heartbeat notch signal can be obtained based on this average value. Thus, the calculation of the heartbeat notch signal is made more accurate.

[0108] In some embodiments, the frequency distribution information of the heartbeat notch signal can also be obtained by calculating the short-time spectrum of the esophageal pressure signal, so as to obtain the fundamental frequency of the heartbeat notch signal. For example, Figure 5 the spectrum diagram of a typical esophageal pressure signal is shown. With the help of an external machine, it can be recognized that the frequency of the heartbeat notch signal is mainly distributed at abscissas 1, 2, 3, 4, and 5. Therefore, 1 Hz can be used as the fundamental frequency of the heartbeat notch signal.

[0109] Both of the above two methods can obtain the fundamental frequency of the heartbeat notch signal. Either one or a combination of the two can be used. In other embodiments, other feasible methods can also be adopted.

[0110] In some embodiments, after obtaining the fundamental frequency of the heartbeat notch signal, an external device can be introduced to verify the fundamental frequency of the heartbeat notch signal. For example, a monitor or an electrocardiogram device is used to obtain the pulse parameter and / or heart rate parameter of patient 80, and the fundamental frequency of the heartbeat notch signal is verified according to the heart rate parameter and / or pulse rate parameter. For example, when the difference between the two does not exceed a preset threshold, it can be considered that the fundamental frequency of the heartbeat notch signal obtained by the ventilator is credible.

[0111] Step 220, construct a reference signal according to the fundamental frequency of the heartbeat notch signal.

[0112] In this step, a reference signal can be obtained by constructing a basic wave with the same frequency as the fundamental frequency of the heartbeat notch signal. The basic wave can be any one of a sine wave, a triangular wave, a square wave, a sawtooth wave, and a Gaussian function. A reference signal can be obtained according to the basic wave. For example, the basic wave can be directly used as the reference signal. Taking a sine wave as an example below, the construction process of the basic wave is illustrated as follows, and the construction method is as follows:

[0113] Noise_ref = A * sin(2 * π * HeartRate / (SamplingRate * 60) * t + B).

[0114] Where Noise_ref is the reference signal, HeartRate is the heart rate of patient 80, Samplingrate is the sampling rate of the above ventilator when acquiring the esophageal pressure signal of patient 80, A is the amplitude, B is the phase, and the amplitude, phase, and the frequency of the sine wave on the right side can all be adjusted to meet the construction requirements.

[0115] A sine wave with a frequency equal to the fundamental frequency of the heartbeat notch signal can be constructed according to the above formula. For example, according to the fundamental frequency of the heartbeat notch signal obtained from the Figure 3 esophageal pressure signal, a reference signal is constructed. Figure 6 is the signal frequency spectrum diagram after the Fourier transform of the constructed reference signal. It can be seen that the frequency domain energy of the signal is mainly concentrated on the fundamental frequency of the heartbeat notch signal. Therefore, the above method successfully constructs a suitable reference signal.

[0116] It should be noted that the above process of constructing the reference signal is an explanation of an implementation manner of the present invention, and the method for constructing the reference signal of the present invention is not limited to the examples given.

[0117] In some embodiments, the reference signal may include, in addition to the signal component containing the fundamental frequency energy of the heart rate notch, signal components at the respective harmonic frequencies of the heart rate notch. For example, the reference signal can be constructed in the following manner:

[0118] Noise_ref = A1*sin(2*π*HeartRate / (SamplingRate*60)*t + B1)+

[0119] A2*sin(2*π*HeartRate*2 / (SamplingRate*60)*t + B2)+

[0120] A3*sin(2*π*HeartRate*3 / (SamplingRate*60)*t + B3)+...

[0121] This formula can represent that the fundamental wave is superimposed with its 2nd to Nth harmonics to obtain the reference signal Noise_ref, where 2 ≤ N.

[0122] Among them, A2*sin(2*π*HeartRate*2 / (SamplingRate*60)*t + B2) contains the energy of the heart rate notch signal at the second harmonic.

[0123] A3*sin(2*π*HeartRate*3 / (SamplingRate*60)*t + B3) contains the energy of the heart rate notch signal at the third harmonic. The meanings of the symbols in this formula have been explained above and will not be elaborated here. Figure 7 It is the signal frequency spectrum diagram after Fourier transform of the reference signal constructed according to the above formula.

[0124] Through the above construction method, when filtering the esophageal pressure signal, the harmonic energy of the heart rate notch signal can be filtered out, so as to better filter the esophageal pressure.

[0125] The number of harmonics of the heart rate notch signal included in the constructed reference signal can be determined according to actual usage requirements. For example, the reference signal can be fixed to be constructed to include the third harmonic energy of the heart rate notch signal, or to include the fifth harmonic energy of the heart rate notch signal to meet the requirements of most filtering occasions.

[0126] Step 300, update the filtering parameters of the adaptive filter according to the reference signal to obtain an updated adaptive filter.

[0127] The classical noise canceller mentioned above is a way to update the filtering parameters of the adaptive filter using the reference signal. In addition, it can also be Figure 11The process shown updates the filtering parameters of the adaptive filter, including the steps:

[0128] Step 310, superimpose the esophageal pressure signal and the reference signal, and use the superimposed signal as the first input signal.

[0129] Step 320, filter the first input signal using an adaptive filter to obtain an output signal.

[0130] Step 330, calculate the error signal between the esophageal pressure signal and the output signal.

[0131] Step 340, according to the error signal, use an adaptive algorithm to update the filtering parameters of the adaptive filter to obtain an updated adaptive filter.

[0132] Step 340 is to change the filtering parameters of the adaptive filter in the direction of minimizing the error signal according to the above various algorithms. For example, change any one or several of the step size, order, and coefficients of the adaptive filter, so as to obtain an updated adaptive filter.

[0133] Step 400, filter out the heartbeat notch signal in the esophageal pressure signal according to the updated adaptive filter.

[0134] In some embodiments, use the esophageal pressure signal as the second input signal, and use the updated adaptive filter to filter the second input signal, so as to filter out the heartbeat notch signal in the esophageal pressure signal. The filtering result is as Figure 9 shown, where the dark solid line part is the spectrogram and waveform diagram after filtering, and the light solid line is the spectrogram and waveform diagram before filtering. It can be clearly seen that the noise signal in the esophageal pressure signal is effectively filtered out.

[0135] In addition to the esophageal pressure signal, other physiological signals can also filter out noise signals in a similar manner as above. Specifically, first obtain the physiological signal of patient 80 changing with time; then construct a reference signal whose frequency domain distribution or time domain distribution is close to the noise signal in the physiological signal according to the physiological signal; then update the filtering parameters of the adaptive filter according to the reference signal to obtain an updated adaptive filter; finally, filter out the noise signal in the physiological signal according to the updated adaptive filter. It should be understood that in the embodiments of the present invention, the user can decide when to enable and stop the filtering method of the adaptive filter by himself, and does not need to always use the filtering method of the adaptive filter, and other filtering methods can also be used in combination.

[0136] In the above embodiments, no other external devices need to be introduced except for the respiratory device used for mechanical ventilation; moreover, the frequency-domain distribution of the reference signal of the adaptive filter is close to that of the heartbeat notch signal, and the frequency-domain energy distribution conforms to that of the heartbeat notch, resulting in better real-time filtering effect.

[0137] The above uses specific examples to illustrate the present invention, which is only used to help understand the present invention and is not intended to limit the present invention. For those of ordinary skill in the art, according to the idea of the present invention, the above specific embodiments can be changed.

Claims

1. A filtering method for esophageal pressure signals, characterized in that , including the steps: Obtain the esophageal pressure signal of the patient, and obtain the blood oxygen signal of the patient; Construct a reference signal according to the blood oxygen signal; Update the filtering parameters of the adaptive filter according to the reference signal to obtain an updated adaptive filter; Filter out the heartbeat notch signal in the esophageal pressure signal according to the updated adaptive filter.

2. The filtering method according to claim 1, wherein The constructing the reference signal according to the blood oxygen signal includes: Use the blood oxygen signal as the basic signal to construct a reference signal whose frequency domain distribution is close to that of the heartbeat notch signal.

3. The method according to claim 1 or 2, characterized in that, Constructing the reference signal includes: Obtain the fundamental frequency of the heartbeat notch signal in the blood oxygen signal according to the blood oxygen signal; Construct the reference signal according to the fundamental frequency of the heartbeat notch signal.

4. The method according to claim 3, wherein Obtaining the fundamental frequency of the heartbeat notch signal in the blood oxygen signal according to the blood oxygen signal includes at least one of the following methods: Use the pulse parameter and / or heart rate parameter in the blood oxygen signal as the fundamental frequency of the heartbeat notch signal; Identify the time domain feature and / or frequency domain feature of the heartbeat notch signal from the blood oxygen signal, and calculate the fundamental frequency of the heartbeat notch signal; Obtain the frequency distribution information of the heartbeat notch signal by calculating the short-time spectrum of the blood oxygen signal, so as to obtain the fundamental frequency of the heartbeat notch signal.

5. The method according to claim 4, wherein It further includes: Obtain the pulse parameter and / or heart rate parameter in the blood oxygen signal through a monitor; or obtain the pulse parameter and / or heart rate parameter in the blood oxygen signal through an electrocardiogram device.

6. The method according to claim 4, wherein The calculating the fundamental frequency of the heartbeat notch signal includes: Identify the waveform within at least one period of the blood oxygen signal; Calculate the time interval between two adjacent wave peaks within one period of at least one period of the blood oxygen signal; Calculate the fundamental frequency of the heartbeat notch signal according to the time interval.

7. The method according to claim 4, characterized in that The calculating the fundamental frequency of the heartbeat notch signal includes: Identify the waveform within at least one period of the blood oxygen signal; Calculate at least two time intervals between multiple adjacent wave peaks within at least one period of the blood oxygen signal; Calculate the fundamental frequency of the heartbeat notch signal according to the average value of at least two time intervals.

8. The method according to claim 3, wherein The method further includes: Obtain the pulse rate parameter and / or heart rate parameter of the patient, and verify the fundamental frequency of the heartbeat notch signal according to the heart rate parameter and / or pulse rate parameter.

9. The method according to claim 3, characterized in that, The constructing the reference signal according to the fundamental frequency of the heartbeat notch signal includes: Construct a basic wave with the same frequency as the fundamental frequency, and the basic wave includes any one of a sine wave, a triangular wave, a square wave, a sawtooth wave, and a Gaussian function; Obtain a reference signal according to the basic wave.

10. The method according to claim 9, wherein The obtaining a reference signal according to the basic wave includes: Use the basic wave as the reference signal; or Calculate the 2nd to Nth harmonics of the basic wave; Superimpose the basic wave and its 2nd to Nth harmonics to obtain a reference signal, where N is greater than or equal to 2.

11. The method according to claim 10, wherein N is equal to 3 or 5.

12. The method according to claim 1, wherein The method for updating the filtering parameters of the adaptive filter includes: Superimpose the esophageal pressure signal and the reference signal, and use the superimposed signal as the first input signal; Filter the first input signal using the adaptive filter to obtain an output signal; Calculate an error signal between the esophageal pressure signal and the output signal; Update the filtering parameters of the adaptive filter according to the error signal using an adaptive algorithm to obtain the updated adaptive filter.

13. The method according to claim 12, wherein The updating of the filtering parameters of the adaptive filter using the adaptive algorithm includes: Updating at least one of the step size, order, and coefficients of the adaptive filter.

14. The method according to claim 1, characterized in that The filtering of the heartbeat notch signal from the esophageal pressure signal according to the updated adaptive filter includes: Using the esophageal pressure signal as a second input signal; Filter the second input signal using the updated adaptive filter to filter out the heartbeat notch signal from the esophageal pressure signal.

15. A medical device, characterized in that, Comprising: A memory for storing programs; A processor for implementing the method according to claims 1-14 by executing the programs stored in the memory.

16. A computer-readable storage medium, characterized in that, Including a program that can be executed by a processor to implement the method according to claims 1-14.

Images