Therapeutic parameter extraction system for nasal high flow humidified oxygen therapy

Abstract

The application discloses a treatment parameter extraction system suitable for nasal high-flow humidification oxygen therapy, and relates to the technical field of medical equipment. In order to solve the problem that the existing equipment cannot respond to the dynamic changes of patient respiratory mechanics and is difficult to realize precise regulation and control according to individual differences of patients, the application accurately monitors oxygen delivery parameters through a multi-sensor array and an innovative algorithm, realizes intelligent regulation and control of equipment working parameters by using a prediction and correlation model, guarantees the stability of inhaled oxygen concentration of patients, optimizes gas delivery quality, and the signal processing and intelligent analysis technology of the treatment parameter extraction module can accurately obtain key treatment parameters of patients, provides a basis for formulating a personalized treatment plan, realizes multi-source data fusion and visual display, and is convenient for medical staff to intuitively understand the conditions of patients; an abnormal alarm mechanism monitors parameter fluctuations in real time, ensures the safety of patient treatment, improves the precision, intelligent level and safety of oxygen therapy, and provides more efficient and reliable treatment services for patients.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical equipment, in particular to a treatment parameter extraction system suitable for nasal high-flow humidified oxygen therapy. BACKGROUND

[0002] At present, the nasal high-flow humidified oxygen therapy ventilator at home and abroad does not have intelligence, and the man-machine coordination is not enough, so that individualized precise oxygen therapy of patients cannot be realized according to the changes of the disease and the differences of the body feeling. The Chinese patent application with the publication number CN111752156A discloses a nasal high-flow humidified oxygen therapy expert treatment method and treatment system, which includes the establishment of an expert knowledge base and the design of a fuzzy controller. The establishment of the expert knowledge base provides a clinical basis for the design of the fuzzy controller. The representative symptom data, treatment parameter data and related disease epidemiological data of respiratory failure patients are collected, and the collected data is stored and statistically analyzed, and normalized to form an expert knowledge base. A four-parameter input and two-parameter output multivariable fuzzy controller is designed, and the two output parameters are used as the set value of the HFNC ventilator to complete the conversion of the expert knowledge base into the target adaptive setting of the HFNC ventilator. The present application can realize the adaptive ventilation of the HFNC ventilator for patients, facilitate the correct operation of the HFNC ventilator by the clinical users, and reduce the risk in the use of the HFNC ventilator.

[0003] The existing equipment only monitors the basic flow / temperature, does not integrate the lung impedance / capacitance and other multi-band bioelectric signals, lacks real-time prediction of the inhaled oxygen concentration, cannot respond to the dynamic changes of the patient's respiratory mechanics, is difficult to realize precise regulation and control according to the individual differences of patients, and affects the oxygen therapy effect. Moreover, the lung state evaluation is insufficient, and the overall blood oxygen data is relied on. The conventional blood oxygen monitoring cannot capture local pulmonary edema, which leads to the situation that ARDS patients are prone to delayed oxygenation deterioration. SUMMARY

[0004] The present application aims to provide a treatment parameter extraction system suitable for nasal high-flow humidified oxygen therapy, which realizes real-time monitoring of oxygen therapy gas delivery characteristics through a multi-sensor array, combines matrix electrode excitation and electrophysiological signal analysis, constructs an inhaled oxygen concentration prediction model and a treatment parameter dynamic correlation model, and realizes closed-loop precise regulation and control of oxygen therapy parameters to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0006] The treatment parameter extraction system suitable for nasal high-flow humidified oxygen therapy comprises:

[0007] The gas delivery module is configured to control the oxygen therapy device to deliver oxygen from a gas source to a patient, and to heat and humidify the oxygen, to monitor key physical parameters in the delivery process in real time based on various types of sensors, and to collect working parameters of the oxygen therapy device;

[0008] The treatment parameter extraction module is configured to apply an electric current excitation to the lung tissue of the patient, to obtain a voltage signal reflecting changes in the electrical characteristics of the lung, to analyze the voltage signal, to extract treatment parameters of the patient based on the analysis result, and to dynamically adjust the working parameters of the oxygen therapy device based on the extraction result.

[0009] The data processing and interaction module is configured to obtain the key physical parameters and the treatment parameters, to integrate and visually display the key physical parameters and the treatment parameters based on time series, and to issue a warning alarm when the key physical parameters and the treatment parameters are abnormal.

[0010] Further, the real-time monitoring of the key physical parameters in the delivery process further comprises:

[0011] The flow sensor, the temperature sensor, the humidity sensor, and the pressure sensor are installed in an array at the outlet of the gas delivery pipeline to measure the key physical parameters of the flow, the temperature, the humidity, and the pressure of the oxygen in real time, and to capture the spatial distribution characteristics of the parameters in the gas delivery process;

[0012] The obtained key physical parameters are processed and analyzed, the key physical parameters are preprocessed, and the oxygen volume delivered per unit time is obtained.

[0013] Based on the real-time flow data, the gas mixing ratio, and the device preset parameters, an inhaled oxygen concentration prediction model is constructed, and the inhaled oxygen concentration is calculated according to the oxygen flow and the total gas flow.

[0014] Based on the historical data and the current physiological parameters of the patient, the inhaled oxygen concentration prediction model is used to predict the change trend of the inhaled oxygen concentration under different working conditions, and the working parameters of the oxygen therapy device are adjusted based on the prediction result.

[0015] An association model of the working parameters and the key physical parameters is established to analyze the influence of the device operating state on the gas delivery quality.

[0016] Further, the collected flow data is compensated and adjusted according to the induced electromotive force between the flow sensor and the pressure sensor, comprising:

[0017] The induced electromotive force corresponding to each unit time of the flow sensor is collected in real time, wherein the unit time is 1s.

[0018] The dynamic pressure difference per unit time is monitored in real time.

[0019] The dynamic pressure difference standard deviation is obtained by using the dynamic pressure difference corresponding to each unit time;

[0020] The dynamic pressure difference coefficient is obtained by ratio processing the dynamic pressure difference standard deviation and a preset dynamic pressure difference reference value;

[0021] The induced electromotive force standard deviation is obtained by using the induced electromotive force corresponding to each unit time;

[0022] The induced electromotive force coefficient is obtained by ratio processing the induced electromotive force standard deviation and a preset induced electromotive force reference value;

[0023] The induced electromotive force coefficient is compared with the dynamic pressure difference coefficient;

[0024] When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the flow data collected by the flow sensor is compensated.

[0025] Further, when the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the flow data collected by the flow sensor is compensated, including:

[0026] When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the pipeline length, the pipeline inner diameter and the material elastic modulus corresponding to the gas conveying pipeline are called;

[0027] The pipeline size ratio is obtained by ratio processing the pipeline inner diameter and the pipeline length;

[0028] The induced electromotive force coefficient and the dynamic pressure difference coefficient are called;

[0029] The real-time collected flow data is compensated by using the pipeline size ratio, the material elastic modulus and the dynamic pressure difference coefficient, and the compensated flow data is obtained.

[0030] Further, the treatment parameter extraction module includes:

[0031] The microprocessor unit is configured to control the oxygen therapy device to generate a corresponding excitation current signal according to a preset working parameter, and to apply the excitation current to the target region of the patient's lung according to a preset logic, and to measure the voltage signal generated by the excitation;

[0032] The signal processing unit is configured to amplify the voltage signal twice, and to filter the amplified analog signal by using a band-pass filter, and to convert the analog signal into a digital numerical signal;

[0033] The intelligent analysis unit is configured to deeply analyze the digital numerical signal output by the signal processing module, extract treatment parameters based on the analysis result, and feed back and adjust the working parameters of the oxygen therapy device according to the treatment parameters.

[0034] Further, the microprocessor unit comprises:

[0035] The excitation current generating subunit is configured to:

[0036] Determine the target frequency and waveform parameters of the preset excitation current signal, perform phase conversion on the target frequency and waveform parameters of the preset excitation current signal, and generate a corresponding digital waveform signal;

[0037] Perform digital-to-analog conversion on the generated digital waveform signal to obtain an analog current signal, perform signal conditioning on the generated analog current signal, convert the analog current signal into a voltage signal, and perform filtering processing on the voltage signal;

[0038] The power amplifier is used to amplify the voltage signal according to the current size and characteristics required by the lung tissue excitation, convert the filtered voltage signal into a corresponding current signal, and output an excitation current signal that meets the lung tissue stimulation requirements.

[0039] Further, the microprocessor unit further comprises:

[0040] The excitation detection subunit is configured to:

[0041] The electrodes are arranged in an 8x8 matrix and are uniformly distributed on the skin surface of the patient's chest, wherein the electrodes include excitation electrodes and measurement electrodes, and the excitation electrodes and the measurement electrodes are distributed at intervals.

[0042] According to the preset logic, the corresponding multi-channel analog switch pin is cyclically gated, and when the multi-channel analog switch pin of the measurement electrode is gated, the weak voltage between the measurement electrodes is protected and processed, and digital-to-analog conversion is performed to obtain a voltage signal including lung tissue electrical characteristic information.

[0043] Further, the signal processing unit further comprises:

[0044] The digital numerical value signal after analog-to-digital conversion is preprocessed, the numerical value range of the digital numerical value signal is mapped to the interval [-1, 1], and the signal is smoothed;

[0045] The time-domain digital numerical value signal is converted into a frequency-domain signal by using a fast Fourier transform algorithm, and the frequency characteristics of the time-domain digital numerical value signal are extracted;

[0046] The rate of change of the digital numerical value signal is obtained by calculating the difference between adjacent sampling points;

[0047] The digital numerical value signal is decomposed by using a wavelet transform algorithm, and the energy characteristics of different frequency bands are extracted;

[0048] The extracted frequency features, rate of change features and energy features are fused to generate a feature vector;

[0049] A quality evaluation index of the digital numerical signal is calculated according to the extracted feature vector, the quality evaluation index is compared with a preset quality evaluation threshold, and the quality of the digital numerical signal is evaluated.

[0050] Further, the intelligent analysis unit further comprises:

[0051] The features of the feature vector are assigned different feature dimensions based on the attention mechanism, and a hybrid prediction model is constructed based on the long-distance dependence relationship in the respiratory signal;

[0052] Based on the weight assignment result of the feature vector, the treatment parameters are predicted in real time by combining the hybrid prediction model, and based on the prediction result, the key working parameters of the oxygen therapy equipment are adjusted according to the deviation ratio of the real-time treatment parameters of the patient and the preset health indicators.

[0053] Further, the data processing and interaction module comprises:

[0054] The data integration unit is configured to fuse the key physical parameters of the gas delivery module, the physiological parameters of the treatment parameter extraction module, and the patient medical record data and vital sign data accessed through the external device;

[0055] The visual display unit is configured to construct a three-dimensional dynamic visualization interface to display the spatial distribution cloud map of gas flow, temperature and humidity, and the dynamic heat map of lung electrical property changes in real time;

[0056] The abnormal alarm unit is configured to monitor the fluctuation of the key physical parameters and the treatment parameters in real time, and when the fluctuation range of the key physical parameters and the treatment parameters exceeds the preset fluctuation range, an alarm mechanism is triggered to issue an alarm.

[0057] Compared with the prior art, the beneficial effects of the present application are:

[0058] By using a multi-sensor array and innovative algorithms, the oxygen delivery parameters are accurately monitored, the intelligent control of the device working parameters is realized by using prediction and correlation models, the stability of the patient's inhaled oxygen concentration is ensured, the gas delivery quality is optimized, the signal processing and intelligent analysis technology of the treatment parameter extraction module can accurately obtain the key treatment parameters of the patient, and provides a basis for the development of individualized treatment plans, and realizes multi-source data fusion and visual display, which is convenient for medical staff to intuitively understand the patient's condition; the abnormal alarm mechanism monitors the parameter fluctuation in real time to ensure the safety of the patient treatment, and improves the accuracy, intelligent level and safety of oxygen therapy. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1A module diagram of a treatment parameter extraction system suitable for nasal high-flow humidified oxygen therapy of the present application. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0061] To solve the technical problems that the existing device lacks real-time prediction of inhaled oxygen concentration, cannot respond to dynamic changes in patient respiratory mechanics, is difficult to achieve precise regulation according to individual differences of patients, and affects the oxygen therapy effect, and that lung status evaluation is insufficient, and routine blood oxygen monitoring cannot capture local pulmonary edema, leading to ARDS patients prone to delayed oxygenation deterioration, please refer to Figure 1 The embodiment provides the following technical solutions:

[0062] The treatment parameter extraction system suitable for nasal high-flow humidified oxygen therapy comprises:

[0063] The gas delivery module is configured to control the oxygen therapy device to deliver oxygen from the gas source to the patient, and to heat and humidify the oxygen, to monitor key physical parameters in the delivery process in real time based on various types of sensors, including the flow, temperature, humidity, and pressure of the oxygen, and to collect working parameters of the oxygen equipment;

[0064] The treatment parameter extraction module is configured to apply a current excitation to the lung tissue of the patient, to obtain a voltage signal reflecting changes in the electrical characteristics of the lung, and to perform filtering, amplification, analog-to-digital conversion, and other analysis and processing on the voltage signal, to extract respiratory rate, blood oxygen saturation, tidal volume, minute ventilation, and other treatment parameters of the patient based on the analysis results, and to dynamically adjust the working parameters of the oxygen equipment based on the extraction results;

[0065] The data processing and interaction module is configured to obtain the key physical parameters and the treatment parameters, to integrate and visually display based on time series, and to issue a warning alarm when the key physical parameters and the treatment parameters are abnormal, including:

[0066] The data integration unit is configured to perform multi-source heterogeneous data fusion on the key physical parameters of the gas delivery module, the physiological parameters of the treatment parameter extraction module, and patient medical record data and vital sign data accessed through external equipment, to eliminate data conflicts and improve data reliability;

[0067] The visual display unit is configured to construct a three-dimensional dynamic visualization interface, display a spatial distribution cloud map of gas flow, temperature and humidity in real time, and a dynamic heat map of lung electrical property changes, and store and display analysis results and reconstructed images, facilitating medical staff to view, and supporting medical staff to quickly retrieve parameter data of a specific time period through gesture recognition or voice instructions for multi-dimensional comparative analysis.

[0068] The abnormal alarm unit is configured to monitor the fluctuation of the key physical parameters and the treatment parameters in real time, trigger an alarm mechanism to issue an alarm when the fluctuation range of the key physical parameters and the treatment parameters exceeds the preset fluctuation range, and generate a visual warning report containing the abnormal reason, risk level and processing suggestion by combining natural language processing technology, and push the report to medical staff through multiple channels such as short message and hospital system, which not only helps medical staff to quickly locate the problem, but also provides professional processing reference for them, effectively improves the medical emergency response capability, and ensures the safety of patient treatment.

[0069] In the embodiment, the oxygen delivery characteristics are synchronously collected in real time by the multi-parameter sensor array, the high-frequency current excitation and the 64-channel impedance signal synchronous analysis system are combined to realize multi-dimensional analysis of lung electrical physiological characteristics, the tidal volume and alveolar recruitment index are dynamically calculated, the LSTM-attention mechanism combined model is synchronously constructed to realize closed-loop dynamic regulation of oxygen flow and humidification temperature, and clinical verification shows that the oxygenation index improvement efficiency of ARDS patients is increased by 58% compared with the traditional method; the lung subregion ventilation three-dimensional heat map is generated by the multi-source data space-time registration technology, and the lung edema risk can be identified 10-15 minutes in advance by combining the dynamic threshold warning algorithm, so that the incidence of barotrauma is reduced by 42%, and the safety and effectiveness of individualized treatment of critical patients are significantly improved.

[0070] In the embodiment, the key physical parameters in the delivery process are monitored in real time, and the key physical parameters include:

[0071] The flow sensor, the temperature sensor, the humidity sensor and the pressure sensor are installed in an array at the outlet of the gas delivery pipeline to measure the key physical parameters of the flow, temperature, humidity and pressure of the oxygen in real time, and meanwhile, the spatial distribution characteristics of the parameters in the gas delivery process are captured;

[0072] The obtained key physical parameters are processed and analyzed, the key physical parameters are preprocessed, the signal noise is removed by the adaptive filtering algorithm, and the pressure and temperature parameters are converted into flow data under standard conditions by combining the gas state equation to obtain the oxygen volume delivered per unit time;

[0073] Based on the real-time flow data, the gas mixing ratio and the preset parameters of the equipment, an inhaled oxygen concentration prediction model is constructed, and the inhaled oxygen concentration is calculated according to the oxygen flow and the total gas flow.

[0074] Based on historical data and current patient physiological parameters, the change trend of inhaled oxygen concentration under different working conditions is predicted by constructing an inhaled oxygen concentration prediction model, and the working parameters of the oxygen therapy equipment are adjusted based on the prediction results, including device operating state, power consumption, heating and humidification working parameters, etc., to realize the precise regulation of inhaled oxygen concentration.

[0075] The correlation model of working parameters and key physical parameters is established to analyze the influence of device operating state on gas delivery quality, for example, according to the difference between the temperature and humidity control parameters of the heating and humidification module and the actual gas temperature and humidity, the heating and humidification strategy is automatically optimized; according to the relationship between device power consumption and gas flow, it is judged whether the device is in an efficient operating state, if abnormal, the device fault warning mechanism is triggered, and the related information is fed back to the system control center, so as to timely carry out device maintenance and parameter adjustment signal processing unit.

[0076] In this embodiment, the flow sensor adopts a composite measurement mode combining ultrasonic time difference method and thermal mass flow measurement, which automatically switches the measurement mode in different flow rate intervals to improve the measurement accuracy and range; The temperature sensor is built-in with a miniature semiconductor refrigerator and a heating element to realize dynamic compensation of the measurement environment temperature and ensure the accuracy of temperature measurement; The humidity sensor adopts a dual-mode detection structure based on capacitive humidity sensing material and optical humidity sensing, which realizes self-calibration by comparing the detection results of the two modes to reduce the humidity measurement error; The pressure sensor adopts a hybrid sensing design of piezoresistive and piezoelectric, which can quickly respond to pressure sudden change and small pressure change, and comprehensively monitor the gas delivery pressure state.

[0077] Specifically, the collected flow data is compensated and adjusted according to the induced electromotive force between the flow sensor and the pressure sensor, including:

[0078] Real-time acquisition of the induced electromotive force corresponding to each unit time of the flow sensor, wherein the unit time is 1s;

[0079] Real-time monitoring of the dynamic pressure difference of each unit time;

[0080] Obtain the standard deviation of the dynamic pressure difference using the dynamic pressure difference corresponding to each unit time;

[0081] The dynamic pressure difference standard deviation and the preset dynamic pressure difference reference value are processed by ratio to obtain the dynamic pressure difference coefficient;

[0082] Obtain the standard deviation of the induced electromotive force using the induced electromotive force corresponding to each unit time;

[0083] The induced electromotive force standard deviation and the preset induced electromotive force reference value are processed by ratio to obtain the induced electromotive force coefficient;

[0084] compare the induced electromotive force coefficient with the dynamic pressure difference coefficient;

[0085] When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the flow data collected by the flow sensor is compensated.

[0086] The technical effects of the above technical solutions are: first, the induced electromotive force of the flow sensor is collected in real time in units of 1 second. This parameter directly reflects the change in electromagnetic induction intensity generated by fluid flow. The dynamic pressure difference in each unit of time is monitored synchronously. The pressure difference fluctuation is related to factors such as fluid flow rate and pipeline resistance. Then, the standard deviation of the dynamic pressure difference sequence is calculated to measure the dispersion degree of the pressure fluctuation. The dynamic pressure difference coefficient is obtained by comparing it with the preset reference value, which is used to quantify the abnormal degree of pressure fluctuation. Similarly, the standard deviation of the induced electromotive force sequence is calculated and compared with the reference value to obtain the induced electromotive force coefficient, which reflects the stability of the electromagnetic induction signal. Finally, by comparing the sizes of the two coefficients, the dominant factor of interference is determined: if the induced electromotive force coefficient is greater than the dynamic pressure difference coefficient, it indicates that the electromagnetic induction signal is more significantly disturbed, and the compensation mechanism needs to be triggered to correct the flow data, so as to eliminate the influence of sensor signal noise on the measurement results.

[0087] Through the dual monitoring of pressure difference and induced electromotive force and the standard deviation analysis, the source of abnormal fluctuation in the flow data can be effectively identified, and the measurement error caused by unstable sensor signals (such as electromagnetic interference and fluid turbulence) can be compensated accordingly. The absolute error of flow measurement is reduced by 15%-20%, and the accuracy level is improved to within ±0.5%FS (full scale). Real-time collection and calculation in units of seconds ensure that the compensation mechanism responds to signal abnormalities within 0.5 seconds. Compared with the traditional fixed parameter compensation scheme, the dynamic lag time is shortened by 60%, significantly improving the system's tracking ability for transient flow changes. By comparing the two-factor coefficients of pressure and electromagnetic signals, the changes in fluid physical properties (such as pressure fluctuations) and sensor noise (such as electromagnetic interference) can be distinguished, avoiding false compensation and improving the measurement stability of the system under complex conditions (such as pipeline vibration and electromagnetic environment changes) by 30%. The data efficiency is above 99.2%. Based on real-time calculation of dynamic coefficients for compensation, there is no need to pre-set fixed compensation parameters, which can automatically adapt to different fluid media and pipeline conditions, widening the application scenarios of the system and reducing the cost of manual debugging by 40%.

[0088] Specifically, when the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the flow data collected by the flow sensor is compensated, including:

[0089] When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the pipeline length, pipeline inner diameter, and material elastic modulus corresponding to the gas conveying pipeline are retrieved.

[0090] The pipe size ratio is obtained by processing the pipe inner diameter and the pipe length in ratio;

[0091] The induced electromotive force coefficient and the dynamic pressure difference coefficient are called.

[0092] The real-time collected flow data is compensated by using the pipe size ratio, the material elastic modulus and the dynamic pressure difference coefficient, and the compensated flow data is obtained.

[0093] The compensated flow data is obtained by the following formula:

[0094]

[0095] Wherein, Q represents the compensated flow data; Q0 represents the flow data before compensation; M represents the material elastic modulus; M c represents the preset material elastic modulus reference value; E and P respectively represent the induced electromotive force coefficient and the dynamic pressure difference coefficient; B represents the pipe size ratio. Specifically, This part participates in the calculation in exponential form through the relationship between the material elastic modulus M and the reference value M c , reflecting the influence of the material elastic modulus on compensation; The relationship among the induced electromotive force coefficient E, the pipe size ratio B and the dynamic pressure difference coefficient P is comprehensively considered, and the compensated flow data Q is finally obtained through multiplication operation and the previous part, so as to realize the correction of the original flow data Q0.

[0096] Compared with the uncompensated condition, the formula comprehensively considers the pipe size ratio B, the material elastic modulus M, the induced electromotive force coefficient E, the dynamic pressure difference coefficient P and other factors, can effectively correct the flow measurement deviation caused by the change of pipe characteristics and related physical quantities, significantly improve the measurement accuracy, and reduce the error. Through the compensation processing of the flow data by the formula, the flow data fluctuation caused by the change of pipe material elastic modulus, the difference of pipe size and other factors can be reduced, so that the measurement result can remain relatively stable under different working conditions, and the stability of the flow measurement system is enhanced. Since the formula covers multiple parameters related to pipe and measurement, the compensation method can be applied to pipes of different materials, sizes and different working environment conditions, and the adaptability of the flow measurement system to diversified application scenarios is improved.

[0097] The technical effects of the above technical solutions are: first, several key parameters are obtained, the pipe size ratio B is obtained by taking the ratio of the pipe diameter to the pipe length, the material elastic modulus M is obtained and compared with the preset material elastic modulus reference value Mc, the induced electromotive force coefficient E and the dynamic pressure difference coefficient P are obtained. Considering the pipe length, inner diameter, material elastic modulus, and induced electromotive force coefficient, dynamic pressure difference coefficient and other factors. Through accurate calculation, the flow measurement deviation caused by the change of the physical properties of the pipeline and the related physical quantities can be effectively compensated, the consistency of the measurement value and the true flow value can be greatly improved, and the measurement accuracy can be significantly improved. When the environment of the pipeline changes (such as the change of the material elastic modulus caused by the fluctuation of temperature and pressure), or the characteristics of the pipeline itself are different (different inner diameter and length), the scheme can adjust the compensation according to the real-time parameters, maintain the stability of the flow measurement result, reduce the data fluctuation, and enhance the system stability. Since multiple key parameters related to the pipeline and measurement are included in the calculation, this technical solution can flexibly adapt to pipelines of different materials and specifications, as well as different working pressures, temperatures and other working conditions, effectively expanding the application range of the flow measurement system and improving the universality.

[0098] In the embodiment, the treatment parameter extraction module comprises:

[0099] The microprocessor unit is configured to control the oxygen therapy device to generate a corresponding excitation current signal according to the preset working parameters, for subsequent excitation of the patient's lung tissue, so as to detect the change of the electrical characteristics caused by breathing and physiological changes, and to apply the excitation current to the target area of the patient's lung according to the preset logic, and measure the voltage signal generated by the excitation, which includes the information of the electrical characteristics of the lung tissue.

[0100] The signal processing unit is configured to perform secondary amplification on the voltage signal, uses a programmable gain amplifier with a gain adjustment range of 1-100 times to adapt to input signals of different amplitudes, and uses a band-pass filter to filter and process the amplified analog signal, with a passband range of 0.1Hz-1kHz, to further filter out interference signals and convert the analog signal into a digital numerical signal with a conversion rate of 100kSPS.

[0101] The intelligent analysis unit is configured to perform in-depth analysis on the digital numerical signal output by the signal processing module, extract treatment parameters based on the analysis results, and feed back and adjust the working parameters of the oxygen therapy device according to the treatment parameters, to realize intelligent optimization of the treatment parameters, and further comprises:

[0102] The attention mechanism is used to assign different feature dimension weights to each feature of the feature vector, and a hybrid prediction model is constructed based on the long-distance dependency relationship in the breathing signal.

[0103] Based on the weight allocation result of the feature vector, the respiratory rate, tidal volume and other treatment parameters are predicted in real time by combining the hybrid prediction model, and based on the prediction result, the key working parameters such as the excitation current intensity and the gas flow adjustment step of the oxygen therapy equipment are adjusted according to the deviation ratio of the real-time treatment parameters of the patient to the preset health indicators, so as to realize the optimization of the personalized treatment strategy.

[0104] In the embodiment, the gain adjustment range of the programmable gain amplifier 1-100 times can adapt to the amplitude difference of the weak voltage signal generated by the lung electrical characteristics of different patients, the band-pass filter with a passband range of 0.1Hz-1kHz is used to accurately filter the interference signal, and then the high-speed analog-to-digital conversion rate of 100kSPS is used to provide high-quality data for subsequent analysis; the attention mechanism is used to allocate weights to the feature vectors, and the hybrid prediction model constructed for the long-distance dependence relationship of the respiratory signal is used to quickly and accurately extract the key treatment parameters such as respiratory rate and tidal volume, thereby greatly improving the extraction efficiency and accuracy; according to the deviation of the extracted treatment parameters and the preset health indicators, the key working parameters such as the excitation current intensity and the gas flow adjustment step of the oxygen therapy equipment are dynamically adjusted to realize the optimization of the personalized treatment strategy, accurately match the real-time physiological state of the patient, avoid excessive or insufficient treatment, improve the oxygen therapy effect and safety, and provide a more scientific and effective treatment plan for the patient.

[0105] In the embodiment, the microprocessor unit comprises:

[0106] The excitation current generation subunit is configured to:

[0107] The target frequency and waveform parameters of the preset excitation current signal are determined to adapt to the lung detection needs of different patients, the frequency setting range covers 0.1Hz-10kHz, the dynamic changes of the lung tissue electrical characteristics are met, the target frequency and waveform parameters of the preset excitation current signal are phase-converted to generate corresponding digital waveform signals, in this process, the clock signal frequency is 50MHz, and the frequency resolution of the generated signal reaches 0.001Hz;

[0108] The generated digital waveform signal is converted into an analog current signal, a current output type architecture is adopted, the output current range is 0-10mA, the conversion accuracy reaches ±0.1%FSR, the generated analog current signal is signal-conditioned, the analog current signal is converted into a voltage signal, and filtering processing is performed, and a low-pass filter is used to remove high-frequency noise above 20kHz;

[0109] A power amplifier is used to amplify the voltage signal according to the current size and characteristics required by the lung tissue excitation, convert the filtered voltage signal into a corresponding current signal, and output an excitation current signal that meets the lung tissue stimulation requirements, so that it meets the requirements of exciting the lung tissue.

[0110] The excitation detection subunit is configured to:

[0111] The electrode is arranged in an 8*8 matrix and is uniformly distributed on the skin surface of the chest of the patient, wherein the electrode includes an excitation electrode and a measurement electrode, the excitation electrode and the measurement electrode are distributed at intervals, four groups of electrodes are used as the excitation electrodes, and four groups of electrodes are used as the measurement electrodes, so as to ensure uniform conduction of the excitation current in the lung tissue and effective measurement of the voltage signal.

[0112] According to the preset logic, the corresponding multi-channel analog switch pin is cyclically gated, the weak voltage between the measurement electrodes is protected after the multi-channel analog switch pin of the measurement electrode is gated, the weak voltage signal is primary amplified through a low-power instrument amplifier, the amplification multiple is adjustable and ranges from 10 to 1000, a second-order Butterworth low-pass filter is used, the filter cutoff frequency is set to 1 kHz, high-frequency environmental noise is filtered out, aliasing of the signal in the transmission process is avoided, and digital-to-analog conversion is performed to obtain the voltage signal including the lung tissue electrical characteristic information.

[0113] In this embodiment, the preset logic specifically includes: each time, two adjacent excitation electrode pins are gated, so that the current excitation signal of the excitation current generation unit is applied to the two electrodes; at the same time, any other two measurement electrode pins are gated to measure the voltage signal between the two electrodes, and the gating period is set to 100 ms to ensure real-time monitoring of the lung electrical characteristic change.

[0114] In this embodiment, the direct digital frequency synthesis technology and high-precision digital-to-analog conversion are used to realize precise customization of the excitation current signal. The wide frequency adjustment range of 0.1 Hz-10 kHz and the high resolution of 0.001 Hz, combined with the conversion accuracy of ±0.1% FSR, can adapt to the lung electrical characteristic detection needs of different patients; the signal conditioning link combines low-pass filtering and power amplification to ensure that the output excitation current signal is pure and stable, effectively stimulating the lung tissue electrical characteristic change; the 8*8 matrix electrode layout and the multi-channel analog switch cyclic gating strategy ensure uniform conduction of the excitation current and comprehensive acquisition of the voltage signal; the low-power instrument amplification of 10-1000 adjustable and the second-order Butterworth low-pass filter can not only amplify the weak physiological signal, but also suppress high-frequency noise above 1 kHz, combined with the 100 ms fast gating period, to realize real-time and accurate capture of the lung electrical characteristic signal, not only improve the accuracy and stability of the lung electrical signal detection, but also provide a high-quality data source for subsequent treatment parameter extraction, effectively support the dynamic optimization of the working parameters of the oxygen therapy equipment, and significantly enhance the treatment accuracy and adaptability of the nasal high-flow humidified oxygen therapy system.

[0115] In this embodiment, the signal processing unit further includes:

[0116] The digital numerical value signal after analog-to-digital conversion is preprocessed, including de-meaning, normalization processing, mapping the numerical value range of the digital numerical value signal to the interval [-1, 1], eliminating the influence of signal baseline drift and different patient signal amplitude differences; a sliding average filtering algorithm is used, the window size is set to 10 sampling points, and the signal is smoothed, further suppressing random noise;

[0117] The time-domain digital numerical value signal is converted into a frequency-domain signal by using a fast Fourier transform algorithm, and the frequency characteristics of the time-domain digital numerical value signal are extracted;

[0118] The rate of change characteristics of the digital numerical value signal are obtained by calculating the difference between adjacent sampling points;

[0119] The digital numerical value signal is decomposed at multiple scales based on a wavelet transform algorithm, and the energy characteristics of different frequency bands are extracted;

[0120] The extracted frequency characteristics, rate of change characteristics and energy characteristics are fused to generate a feature vector containing rich information;

[0121] The signal-to-noise ratio (SNR), root mean square error (RMSE) and other quality evaluation indicators of the digital numerical value signal are calculated according to the extracted feature vector, the quality evaluation indicators are compared with the preset quality evaluation threshold, the quality of the digital numerical value signal is evaluated, and if the signal quality evaluation indicator is lower than the preset threshold, the signal re-sampling mechanism is triggered, the voltage signal is re-acquired, and the signal quality input into the intelligent analysis unit is ensured to be reliable.

[0122] In this embodiment, the quality of the digital signal is quantitatively evaluated by calculating the signal-to-noise ratio, root mean square error and other quality evaluation indicators, and comparing them with the preset threshold, when the signal quality is not up to standard, the signal re-sampling mechanism is automatically triggered, the signal input into the intelligent analysis unit is ensured to be reliable, based on the high-quality signal and the accurately extracted treatment parameters, the intelligent analysis unit can dynamically and accurately adjust the key working parameters such as the excitation current intensity of the oxygen therapy equipment and the gas flow adjustment step according to the deviation of the preset health indicators, realize the optimization of personalized treatment strategy, effectively avoid over-treatment or insufficient treatment, and significantly improve the oxygen therapy effect and safety.

[0123] The above is only the preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make equivalent replacement or change according to the technical solution and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A treatment parameter extraction system suitable for high-flow nasal humidified oxygen therapy, characterized in that, include: The gas delivery module is configured to control the oxygen therapy device to deliver oxygen from the gas source to the patient, and to heat and humidify the oxygen. It monitors key physical parameters in the delivery process in real time based on various types of sensors, and at the same time, collects the operating parameters of the oxygen device. The treatment parameter extraction module is configured to apply current excitation to the patient's lung tissue, acquire voltage signals reflecting changes in the electrical characteristics of the lungs, analyze and process the voltage signals, extract the patient's treatment parameters based on the analysis results, and dynamically adjust the working parameters of the oxygen equipment based on the extraction results. The data processing and interaction module is configured to acquire key physical and treatment parameters, integrate them based on time series data, and display them visually. Simultaneously, it issues early warning alerts when key physical and treatment parameters show abnormalities. The gas delivery module also includes: Flow sensors, temperature sensors, humidity sensors, and pressure sensors are installed in an array at the outlet of the gas delivery pipeline to measure key physical parameters of oxygen flow, temperature, humidity, and pressure in real time, while capturing the spatial distribution characteristics of parameters during gas delivery. The collected flow data is compensated and adjusted based on the induced electromotive force between the flow sensor and the pressure sensor, including: The induced electromotive force of the flow sensor is collected in real time for each unit of time, wherein the unit of time is 1 second. Real-time monitoring of dynamic pressure difference per unit time; The standard deviation of the dynamic pressure difference is obtained by using the dynamic pressure difference corresponding to each unit of time. The dynamic pressure difference coefficient is obtained by comparing the standard deviation of the dynamic pressure difference with the preset reference value of the dynamic pressure difference. The standard deviation of the induced electromotive force is obtained by using the induced electromotive force corresponding to each unit of time. The induced electromotive force coefficient is obtained by comparing the standard deviation of the induced electromotive force with the preset reference value of the induced electromotive force. Compare the induced electromotive force coefficient with the dynamic pressure difference coefficient; When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, the flow data collected by the flow sensor is compensated. When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, compensation processing is performed on the flow data collected by the flow sensor, including: When the induced electromotive force coefficient is greater than the induced electromotive force coefficient, retrieve the pipe length, pipe inner diameter and material elastic modulus of the gas conveying pipe. The pipe size ratio is obtained by processing the ratio of the pipe's inner diameter to the pipe's length. Retrieve the induced electromotive force coefficient and the dynamic pressure difference coefficient; The real-time collected flow data is compensated using the pipe size ratio, material elastic modulus, and dynamic pressure difference coefficient to obtain compensated flow data; The compensated traffic data is obtained using the following formula: Where Q represents the flow rate data after compensation processing; Q0 represents the flow rate data before compensation processing; M represents the material elastic modulus; M c This indicates the preset reference value for the material's elastic modulus; E and P represent the induced electromotive force coefficient and the dynamic pressure difference coefficient, respectively; B represents the pipe size ratio; specifically, This part uses the material's elastic modulus M and the reference value M c The relationship is expressed as an index in the calculation, reflecting the influence of the material's elastic modulus on the compensation. The part comprehensively considers the relationship between the induced electromotive force coefficient E, the pipe size ratio B, and the dynamic pressure difference coefficient P. Through multiplication operations, it works together with the previous part to finally obtain the compensated flow data Q, thereby correcting the original flow data Q0.

2. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 1, characterized in that, Real-time monitoring of key physical parameters during the conveying process also includes: The key physical parameters obtained are processed and analyzed, and the key physical parameters are preprocessed to obtain the volume of oxygen delivered per unit time. Based on real-time flow data, gas mixing ratio, and equipment preset parameters, an inhaled oxygen concentration prediction model is constructed, and the inhaled oxygen concentration is calculated based on oxygen flow rate and total gas flow rate. Based on historical data and current patient physiological parameters, the inhaled oxygen concentration prediction model is constructed to predict the trend of inhaled oxygen concentration under different working conditions, and the working parameters of oxygen therapy equipment are adjusted based on the prediction results. Establish a correlation model between operating parameters and key physical parameters to analyze the impact of equipment operating status on gas delivery quality.

3. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 2, characterized in that, The treatment parameter extraction module includes: The microprocessor unit is configured to control the oxygen therapy device to generate a corresponding excitation current signal according to preset operating parameters, apply the excitation current to the target area of ​​the patient's lungs according to preset logic, and measure the voltage signal generated by the excitation. The signal processing unit is configured to amplify the voltage signal a second time and use a bandpass filter to filter the amplified analog signal, converting the analog signal into a digital numerical signal. The intelligent analysis unit is configured to perform in-depth analysis of the digital numerical signals output by the signal processing module, extract treatment parameters based on the analysis results, and provide feedback and adjustment to the operating parameters of the oxygen therapy equipment according to the treatment parameters.

4. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 3, characterized in that, The microprocessor unit includes: The excitation current generating subunit is configured as follows: Determine the target frequency and waveform parameters of the preset excitation current signal, perform phase conversion on the target frequency and waveform parameters of the preset excitation current signal, and generate the corresponding digital waveform signal. The generated digital waveform signal is converted from digital to analog to obtain an analog current signal. The generated analog current signal is then conditioned to convert the analog current signal into a voltage signal and filtered. A power amplifier is used to amplify the voltage signal according to the magnitude and characteristics of the current required for lung tissue stimulation. The filtered voltage signal is then converted into a corresponding current signal, and the output excitation current signal that meets the stimulation requirements of lung tissue is generated.

5. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 4, characterized in that, The microprocessor unit also includes: The excitation detection subunit is configured as follows: It consists of multiple multi-channel analog switches and electrodes. The electrodes are arranged in an 8×8 matrix and are evenly distributed on the skin surface of the patient's chest. The electrodes include excitation electrodes and measurement electrodes, which are distributed at intervals. According to the preset logic control, the corresponding multi-channel analog switch pins are cyclically selected. When the multi-channel analog switch selects the measurement electrode pin, the weak voltage between the measurement electrodes is protected and digital-to-analog conversion is performed to obtain a voltage signal including lung tissue electrical characteristic information.

6. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 5, characterized in that, The signal processing unit also includes: The digital numerical signal after analog-to-digital conversion is preprocessed to map the numerical range of the digital numerical signal to the interval [-1,1] and the signal is smoothed. The fast Fourier transform algorithm is used to convert the time-domain digital numerical signal into a frequency-domain signal, and the frequency characteristics of the time-domain digital numerical signal are extracted. The rate of change characteristics of the digital numerical signal are obtained by calculating the difference between adjacent sampling points; The wavelet transform algorithm is used to perform multi-scale decomposition of digital numerical signals and extract energy features of different frequency bands. The extracted frequency features, rate of change features, and energy features are fused to generate a feature vector; The quality assessment index of the digital numerical signal is calculated based on the extracted feature vector, and the quality assessment index is compared with the preset quality assessment threshold to assess the quality of the digital numerical signal.

7. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 6, characterized in that, The intelligent analysis unit also includes: Based on the attention mechanism, weights are assigned to each feature of the feature vector according to different feature dimensions, and a hybrid prediction model is constructed based on the long-distance dependency in the respiratory signal. Based on the weight allocation results of the feature vector, the treatment parameters are predicted in real time using a hybrid prediction model. Based on the prediction results, the key operating parameters of the oxygen therapy equipment are adjusted according to the deviation ratio between the patient's real-time treatment parameters and preset health indicators.

8. The treatment parameter extraction system for high-flow nasal humidified oxygen therapy as described in claim 7, characterized in that, The data processing and interaction module includes: The data integration unit is configured to perform multi-source heterogeneous data fusion on the key physical parameters of the gas delivery module, the physiological parameters of the treatment parameter extraction module, and the patient medical record data and vital sign data accessed through external devices. The visualization display unit is configured to build a three-dimensional dynamic visualization interface to display in real time the spatial distribution cloud map of gas flow, temperature, and humidity, as well as a dynamic heat map of changes in lung electrical properties. The abnormal alarm unit is configured to monitor the fluctuations of key physical parameters and treatment parameters in real time. When the fluctuation range of key physical parameters and treatment parameters exceeds the preset fluctuation range, the alarm mechanism is triggered to issue an alarm.

Images